Ectap Winter - 2020

Theoretical and Applied Economics
Volume XXVII, No. 4(625)/2020, Winter
Contents
The European Union in the face of the prospect of achieving major objectives
in the conditions of the pandemic and economic-financial crisis
Gabriela Victoria Anghelache, Ștefan Daniel Armeanu, Constantin Anghelache ................................................. 5
What determines the portfolio investment flows

to Central and Eastern European Countries in the European Union 2001-2017?
Donny Tang ........................................................................................................................................................ 21
Analysis of the correlation between taxation indicators

and economic growth at the macroeconomic level in European Union and Romania
Cătălina Motofei .................................................................................................................................................. 43
Measuring Adobe Company performance

from the perspective of a Complex Adaptive System
Nora Chiriță, Andrei Vișan, Mihaela Popescu..................................................................................................... 55
Welfare enhancing uncertainty

Sravaitri Chaudhuri ............................................................................................................................................. 73
Business process modeling. Using Unified Modeling Language

to streamline the design of the TO-BE system within a company
Ionuț Nica, Ștefan Ionescu .................................................................................................................................. 89
Static and dynamic analysis of intra-industry trade of BRICS countries

Kuldeep Kumar Lohani ..................................................................................................................................... 107
Fossil fuel consumption, economic growth and CO2 emissions.

Causality evinced from the BRICS world
Rochna Arora, Dr. Baljit Kaur ........................................................................................................................... 131
Methods used in risk financing

Ana Maria Popescu, Ștefan Virgil Iacob, Alina Eliza Dabija ............................................................................. 143
Is there a relationship between economic welfare and innovation performance?

Evidence from selected European countries
Lejla Terzić........................................................................................................................................................ 159
Analyzing Romania GDP: Final consumption, gross investment, and net exports influence
compared to previously published models
Alexander I. Villanueva ..................................................................................................................................... 169
2 Contents
Analysis of the natural movement of the population

under the spectrum of the health crisis
Constantin Anghelache, Mădălina-Gabriela Anghel, Ștefan Virgil Iacob .......................................................... 177
On the relationship between economic growth and government debt for Bulgaria.
Test of the Reinhart-Rogoff hypothesis
Yu Hsing ........................................................................................................................................................... 187
Testifying the role of regulatory environment in trade facilitation:

Impact on intra-regional trade in South-Asia
Tanya Gandhi, Shahid Ahmed .......................................................................................................................... 195
India and trade blocs: A gravity model analysis

Arjun Singh, Dr. S.P. Padhi .............................................................................................................................. 217
Predicting the volatility in stock return of emerging economy:

An empirical approach
Aastha Khera, Dr. Miklesh Prasad Yadav ........................................................................................................ 233
Portfolio optimization with VaR approach:

A comparative analysis for Japan, London, New York and India
Parul Bhatia, Priya Gupta ................................................................................................................................. 245
Income-dependent impacts of financial development

and human capital on economic growth. A non-stationary panel analysis
Mohsen Mohaghegh, A.S. Valipour .................................................................................................................. 263
Monetary policy and exchange rate pass-through in India

Bhavish Sharma................................................................................................................................................ 275
Impact of medium of instruction in education on economic

growth and development
Durbhita V Vaishnav. ........................................................................................................................................ 289
Institutions and economic growth: A comparative analysis of developing

and developed countries based on institutionalized social technologies index
Danish Ahmed Siddiqui, Qazi Masood Ahmed ................................................................................................. 309
Exports, imports and economic growth in India:

An empirical analysis
K. Krishna Reddy .............................................................................................................................................. 323
BEPS – A challenge for the increasing budget revenues

in the process of digitalization
Ionuț Mișa, Meral Kagitci .................................................................................................................................. 331
Implementation of public policies.

The compatibility of the model of public policy with the target space of the policy
Luminița Gabriela Popescu ............................................................................................................................... 345
Fiscal sustainability in Romania

Ioan-Radu Petrariu, Lucian Constantin Vîlcu, Iulian Cornel Lolea, Liana Vladu .............................................. 357
The distinctivities of the complexity 3
Note: The authors are responsible for the content of their articles and for obtaining necessary permissions.
Mircea Dinu Data base indexation:

Tel.: (+4) 031.432.96.02 EconLit
Fax: (+4) 021.210.73.10 http://www.aeaweb.org
E-mail: comenzi@edecon.ro Research Papers in Economics (RePEc)
http://www.ideas.repec.org
http://econpapers.repec.org
Directory of Open Access Journals (DOAJ)
http: //www.doaj.org
EBSCO Publishing
http://www.ebscohost.com
International Consortium for the Advancement
of Academic Publication (ICAAP)
http://www.icaap.org
Cabell’s Directories
http://www.cabells.com
CNCSIS B+
www.economieteoreticasiaplicata.ro; www.ectap.ro
Reception of texts: economia.ta@edeconomica.com
ISSN 1841-8678 (Print)

ISSN 1844-0029 (Online)
4 Contents
Volume XXVII (2020), No. 4(625), Winter, pp. 5-20
The European Union in the face of the prospect of achieving

major objectives in the conditions of the pandemic
and economic-financial crisis
Gabriela Victoria ANGHELACHE

Bucharest University of Economic Studies, Romania
gabriela.anghelache@gmail.com
Ștefan Daniel ARMEANU
darmeanu@yahoo.com
Constantin ANGHELACHE
“Artifex” University of Bucharest, Romania
actincon@yahoo.com
Abstract. The Member States of the European Union aim individually and at the same time in a
complex way to achieve superior results, in order to ensure sustainable economic growth, which
would be the main factor in improving living conditions. In this sense, that of the economic growth
of the European Union and of all the Member States, it is important to initiate economic objectives
with the participation of as many Member States as possible. At the same time, the individual
development of each stratum is important so as to ensure harmonized growth in order to pursue the
second objective, which is to bring the level of economic growth closer to all Member States.
Member States' individual investments, as well as those harmonized in the context of economic
objectives, are the steps by which this challenge could be achieved, could be met. There are enough
contradictions in the European Union, in the sense that some states that have entered the last
tranche or the last stage in the European Union, have yet to set a series of individual objectives to
ensure the elimination of dysfunctions and consequently, creating a harmonization of economic
developments.
The question is whether the European Union is an appropriate framework, so that actors and
partners can look with confidence and confidence to achieve major goals. The example of Brexit is
one that brings a new possibility of insecurity, in the sense that the United Kingdom leaves a gap in
terms of individual participation in the Community budget. Under these conditions, it is likely that
the annual and multiannual budget will be slightly smaller and the lack of Great Britain in the
entourage of the Member States will be able to be covered only by a sustained activity on multiple
levels by the Community countries.
6 Gabriela Victoria Anghelache, Ștefan Daniel Armeanu, Constantin Anghelache
We have in mind the destructive effects that the economy of the member states will have, globally
and individually, in the conditions of triggering and propagating the pandemic and economic-
financial crisis.
There are a number of objectives in the European Union's strategy that must be pursued in full
accordance with the national strategies of the Member States. Only a harmonized policy can ensure
a development closer to the aspirations and wishes of the Member States. The European Union is
also manifested in a series of contradictions that exist between some Member States, which arise
between what each Member State pursues and the global policy promoted by the European Union.
In this regard, in this article we will seek to identify those points that may be positive aspects of
harmonization of efforts or that are discordant and should be eliminated in order to reach a
possibility of correlating the intentions and measures taken, so that it can be achieve some of the
challenges facing the European Union. It should also be noted that between the United States and
the European Union in economic, trade, tourism there are some different positions that can have an
effect on the development of the European Union and its Member States within it.
Keywords: European Union, economic objectives, challenges, economic growth, harmonization of

economic growth.
JEL Classification: F02, F40.

The European Union in the face of the prospect of achieving major objectives 7
Introduction
The Member States of the European Union aim individually and also through the Union's
strategy to achieve economic growth as consistent as possible by capitalizing on the internal
potential of each country, but especially by initiating large-scale economic objectives to be
achieved bring as many Member States as possible.
The European Union has directives in this regard, suggesting that economic objectives must
consist in the initiation of economic objectives based on advanced technology, which will
ensure consistent economic growth. These are priority areas in which this individual and
communal research of the Member States needs to be directed. Thus, in the field of energy
there are limitations in terms of current resources, but which must on the basis of an
extensive study and research program lead to the discovery of new resources, to the
expansion of existing ones, so that in the context of coal resources, oil and gas are depleted
over time, to provide a replacement based on other energy sources, so as to ensure the
potential of the European Union.
Individual investment by Member States can play an important role in this area, but at the
same time other objectives may be to work together between Member States, to ensure
more consistent funding and to meet the conditions facing all mankind. I said about energy
resources, food resources, the exploitation of land fertility in countries that have a penchant
for this possibility of evolution, is another goal. There is also high potential in the field of
transport and other areas. The efforts of several Member States must be combined in order
to increase the technological potential, adjust the industry and other branches of national
economies, the technological conditions offered by the present, but especially the future
based on extensive research, which can be triggered.
There are a number of objectives in the European Union's strategy that must be pursued in
full accordance with the interests and strategies of the Member States. This relationship
between the European Union and the Member States is unequivocal in the sense that the
Member States must also adjust their investment and development strategies according to
the conditions and guidelines that the European Union launches in this regard. We can point
out that only a harmonized policy in the field of multi-level research can provide potential
in increasing the prospect of initiating objectives, let's say quite important, which will lead
to the growth of the economies of all states.
We must also take into account the possibility of cooperation between the European Union
as an entity of all the Member States and the other states that have great potential, with
which we can work in multiple directions to achieve the objectives we were talking about.
At this point in the study, to the question Is the creation of major objectives in the European
Union an objective that can be a challenge for the Member States of the European Union?
Yes but only through a coherent, well-developed European Union strategy and at the same
time through the widest possible participation of the Member States, both in the
development of their own research and innovation and in the participation of to the
development of research and innovation within the European Union.
Literature review
In the development strategy of the European Union, the directives highlight the need to
make better use of resources, technological development, the prospect of initiating major
objectives, so that member countries achieve better and consequent results and
consequently the European Union is in line with this pace of progress and ensuring
additional income in order to achieve the goal of ensuring a higher standard of living. The
issue of these challenges for Member States to start new projects is addressed extensively
in the literature. Analyses, interpretations and attempts are also made to identify the current
situation, especially in which the European Union and the Member States, as well as all the
countries of the world, are facing a major pandemic coronavirus crisis, which will have
particularly serious effects in the context of it will be combined with the effects of the
economic and financial crisis, it will produce great disturbances and especially reductions
of the individual economic development.
A number of authors, respectively Anghelache (2013) made a comprehensive analysis of
the evolution of the national and world economy under the effect of the crisis that was in
2008-2010, foreshadowing the prospect of new delicate situations in the current context of
the coronavirus pandemic crisis and especially in the perspective of the economic and
financial crisis be long lasting. Anghelache and Anghel (2017) and Anghelache et al.
(2016b) they focused on using the possibilities of the evolution of the member states, in a
concrete case of Romania after ten years from the accession, in the conditions in which
they still felt the effects of the economic-financial crisis from 2008-2010.
A number of models have also been suggested that can be used to identify the development
prospects of a country's economy, but also of the European Union economy as a whole.
Anghelache et al. (2016a) they extended an analysis on the platform used in the financial
projects that can be carried out in the individual interest of the engaged countries, but also
of the European Union. Barbosa and Faria (2011) stresses that innovation must be a key,
factor in the technical and scientific development and economic consequences of the
European Union. Berezin and Diez-Medrano (2008) in their work they focus on the need
for a legitimate policy to support European integration. Chalmers, A.W. (2013), addresses
the issue of information strategy as a need to support the practical application of advanced
technologies. Dachs and Pyka (2010) it refers to an answer to be given to the question of
where the innovation is headed. Farole et al. (2011) refers to the need for political cohesion
in the European Union, so that regional development is a priority objective, through the
implementation of projects aimed at reducing the differences that exist between Member
States. Goldberg and Pavcnik (2007) extends its analysis to the process of globalization,
which at the present time of the two mentioned crises seems to be out of the picture, but
which must nevertheless be considered as a variable to ensure further economic growth.
Lane (2006) addresses the issue of the effects of the European Monetary Union, which
creates advantages for some Member States and makes other Member States somewhat
difficult, from which point of view there are some difficulties for the participation of states
with limited financial resources or low technological potential in large European projects,
Lima and Cardenete (2007) considers the role and effects of European funds which are
likely to strike a balance in the development of the Member States and to lead to their
greater participation in the initiation of major projects. Tosun (2014) and Voigt and
Moncada-Paternò-Castello (2012) addresses issues relating to the importance of the
absorption of regional funds by the Member States, as well as to the prospect of intensive
growth of the economies of the Member States of the European Union. The question arises
as to whether the structure of the European Union can be considered in order to take action
through some projects to ensure the approximation of the Member States' level of
development.
Methodology
In the study we aimed to use a series of research methods, to extract the essence from the
data series we took from the European Union, to analyse from the future perspective the
proposals for evolution and development of the European Union in accordance with the
approved strategy by 2030. The term appears longer, but no longer in a perspective in which
the pillars of a complex research program are put in place, solutions can be obtained based
on which to initiate large-scale economic and social objectives, which will have the effect
on the one hand, economic growth of all Member States and the European Union at the
same time, but also of social and consequent conditions of increasing incomes and
improving the quality of life.
We used the methodology of the comparative study that put us in possession of ideas for
future evolution, but taking into account the concrete conditions that are currently encoun-
tered. We have also made extensive use of induction and deduction, exploring the resources
available to the Member States of the European Union, trying to identify the major needs that
Member States and the European Union as a whole strictly need, but also more broadly in
terms of regarding the exchange and extension of research with other non-EU states.
We have widely used data series, graphical representations, evolution based on dynamic
models, but also spectral models that can highlight at the inauguration of projects if they
are seasonal, cyclical and how they can influence to some extent more or less the evolution
of the development of the Member States of the European Union.
From a methodological point of view, we used interpolation and extrapolation syntheses of
data series, as well as estimating the evolution perspectives of some fields of activity using
statistical-econometric methods that led us to obtain parameters used to estimate the future
perspective.
As I mentioned, an important thing is the research, innovation and complex development
of the European Union, but also the individual of all the Member States associated in this
context in the perspectives offered by the European Union.
Data, results and discussions

Any economic objective that is set in motion must continue to benefit from new funding
opportunities in the first place. That is why the theory of the existence of two speeds of
development of the European Union in the field of research and innovation and
consequently of development, can actually be combated by major projects that can be
initiated by Member States. Of course, going and adapting the research topic to the current
conditions, we specify that the objective is a challenge for the Member States of the
European Union, but it is not easy, unfortunately becoming more and more difficult in the
current context determined by two major events. Firstly, Brexit will have an effect on the
prospects of launching major targets as the United Kingdom of Great Britain and Northern
Ireland had a rather delicate word to say in initiating such targets. On the other hand, the
sanitary pandemic crisis (coronavirus) is likely to hamper the prospect of initiating major
objectives, primarily due to diminished funding resources of individual member countries,
but also of the states included in the European Union. Of course, without a doubt, for a
period of time, the health crisis will evolve in conjunction with the economic and financial
crisis that is in progress and with negative effects for the future. However, the problem is
even more acute for the Member States to join forces, for the European Union's leading
institutions (the Council, the European Parliament) to review the development strategy,
adapt it to current conditions and identify additional development resources during the
period next. Undoubtedly, the development of research and innovation, the progress of
multilevel progress with the combined efforts of all Member States, is all and perhaps the
basis for these evolutionary opportunities.
Starting from what I mentioned in the introduction, it follows that the European Union has
now entered a cone of uncertainty due to the effects of Brexit and especially the crisis
caused by the coronavirus pandemic, which will be followed by an economic and financial
crisis perhaps unprecedented.
Based on studies of data series and individual developments in the Member States, it is
anticipated that in 2020 there will be many falls in the area of total and structured growth.
A number of phenomena such as unemployment and inflation, which are detrimental to
economic growth, will have a much more consistent effect in the coming period. We
anticipate on the basis of the results obtained by regression functions that the level of
unemployment may be more than double at the end of 2020, of course, with a not very
uniform distribution among the Member States. The countries that entered the European
Union after 2004 and left with a lower level of development economy, which do not have
sufficient resources for major investments and thus do not have the opportunity to absorb
unemployment will be most affected. Inflation, rising prices, whether it is the consumer
price index or the harmonized price index, will have a destructive effect on the economy
of most Member States.
The problem of retraining remains only theoretical, as the economies of the member
countries are restructuring their ability to create jobs in another field, being positive and
only if they could resume work in all areas of the economy gradually.
Simultaneously with the reduction of production potential, the reduction of agricultural

incomes as a result of a subsidy not exactly at the level of needs, as a result of the cyclicality
in a year 2020 which is affected by an unprecedented crisis, the reduction of tourism and
other activities. as a result of strategies to stop the infection of the population with the
COVID virus 19, will lead to additional costs and thus reduce the individual and global
sources of the European Union in terms of improving research, innovation and
development and consequently even the initiation of large-scale objectives. There may be
some possibilities to reconsider the role of the European Union in the context in which each
Member State will primarily want to secure financial resources, to attract more financial
resources in order to return to macrostability. Macrostability will be seriously affected by
the effects of the economic and financial crisis. That is why the EBRD (European Bank for
Reconstruction and Development), the International Monetary Fund, the World Bank
should reformulate their strategies a little in order to be able to provide sources of financing
for major projects. Here an important role belongs to the European Union, which through
its own strategy ensures the perspective that at least for the common objectives that can be
identified by the Member States and triggered, to obtain cheaper, more attractive sources
of financing in order to make progress wanted.
We are currently witnessing, at least in the geographical area of Eastern and Central
Europe, differentiated interest rates that Member States obtain from the banking system,
either nationally or in Europe. Thus, Romania pays an interest of 3%, Bulgaria of 0.2%,
Poland of 0.77%, Serbia of 0.8%, Hungary of 1.2%, these are figures taken a little behind
and maybe more changed over time. These conditions do not ensure equality between
Member States wishing to participate in projects and to finance their projects in part, as
appropriations are no longer a major source of funding for projects in some countries.
Therefore, we believe that the European Union should make a complex analysis of this
situation in the program of projects and objectives to be achieved to address the issue of
uniform funding, from own funds or from attractive sources, which should then be provided
to Member States. Member States as an attractive source of funding. Of course, the
European Union can participate directly even by allocating resources from the annual and
multiannual budget to finance major investment projects.
The multiannual financial framework 2014-2020 and 2021-2027 must provide sources of
funding so as to ensure the economic growth of the Member States. There is a need for
projects in transport infrastructure, environmental projects, projects on the rehabilitation of
historical and cultural sites, expansion of utility infrastructure in urban areas, but especially
in rural areas. Projects on human resource development must be initiated through the active
involvement of the population in the labor market, the development of skills in this regard
to ensure the most consistent development in each country. Research projects need to be
launched and some objectives set out to reduce the gap between village and city, especially
in some Eastern European countries. The implementation of e-government solutions or ICT
infrastructure investments must be considered.
The previous financial framework 2007-2013 took place in parallel with a global financial
crisis and the effects were seen. In the perspective of 2020-2027 or 2030, it seems that we
will again have to bear the effect of another economic and financial crisis that will be
reflected in the evolution of the Member States, not only in the European Union, but in the
world, which will have a combined effect of reducing economic growth.
The current state must allow the completion of the research, development, innovation
project 2014-2020 and ensure the perspective of their evolution. Graph 1, regarding the
European funds allocated to the Member States as a percentage of the Gross Domestic
Product until 2019, shows that the stars also had different allocated funds available.
Graph 1. European funds allocated to Member States as a percentage of Gross Domestic Product by 2019
Source: Eurostat (data are processed by the authors).
Most of the funds are aimed at improving living standards and living standards, so that by
2025 it will increase by an average of up to 25% in most Member States. It is difficult, but
action must be taken. European co-financing remains an essential element for the
development of the Gross Domestic Product, so for the economic growth in each country,
but also for the provision of resources to be used by the Member States in initiating other
projects that ensure on the one hand the approach of the development level, at the top of
the developed countries in this respect, but also individual and complex advantages for all
Member States. Table 1 presents some statistical data on the multiannual financial
framework (MFF) in the period 2013 to 2019.
Table 1. Multiannual financial framework (MFF) in the period 2013-2019
ECE Member Population GDP Billion GDP per capita EU funds EU funds per EU funds for
State (millions) Euro Euro Billions Euro capita Euro GDP % Euro
Bulgaria 7.2 42.00 5833 6.67 927 15.9%
Croatia 4.24 42.96 10129 1.27 299 3%
Czech Republic 10.51 154.94 14700 26.30 2502 17%
Estonia 1.32 19.53 14849 3.40 2588 17.4%
Hungary 9.88 103.00 10458 24.92 2523 24.1%
Latvia 1.99 24.06 12099 4.53 2278 18.8%
Lithuania 2.94 36.29 12329 6.78 2302 18.7%
Polonia 38.48 403.08 10474 67.19 1746 16.7%
Romania 19.95 150.66 7553 19.18 961 12.7%
Slovakia 5.42 75.21 13875 11.65 2149 15.5%
Slovenia 2.06 37.25 18067 4.10 1989 11%
It is found that Member States had different participations, received differentiated funds
and the implementation of the MFF program until 2015, failed for the Eastern European
states of the European Union to ensure a guarantee of economic growth.
In terms of the degree of contracting and reimbursement of funds, there are differences
between Member States depending on the objectives they have set for themselves, but
especially in terms of the future perspective.
Currently, the official data from the 2014-2020 multiannual program highlight
discrepancies in the evolution of states and access to, use and attract new funds to ensure
the perpetuation of major objectives. It is important to note that in this region of the
European Union, European funds continue to be an important segment of public existence.
A comparative analysis of the states of Eastern and Central Europe with those of Western
Europe that are much more economically and industrially developed reveals large
discrepancies. Eastern European countries benefit from European funds, but there are still
other conditions that often make it difficult to analyse the results of the funds used.
The following graph shows the difference between the degree of contracting and the degree
of reimbursement, making it clear that the states that approached and the possibility of
reimbursement of funds had evolution possibilities clearly superior to other states that
encountered great difficulties in this respect.
Graph 2. Degree of contracting and degree of reimbursement of European funds
In the case of Eastern European countries, operational programs were launched late due to
lack of funding. In the absence of co-financing, beneficiaries submitted requests for
reimbursement more often than necessary, which led to delays in the implementation of
those programs. And from the point of view of the public procurement process, despite
some managerial deficiencies, there were also financing deficiencies, which should be the
basis for creating the foundation in order to relaunch the respective projects.
There is a lack of expertise on the part of managing authorities so that some evaluations
have been delayed and other aspects have also been a brake on the harmonious evolution
of the initiation of common programs and objectives.
At the level of managing authorities, the following causes may explain the low rate of
absorption and utilization of resources available for the initiation of projects to deepen
research in areas of utmost importance to the economy. Thus, there have been fluctuations
in specialized staff, there are still not enough concrete guides and methodologies, programs
and projects are sometimes not sufficiently anchored in the realities of that objective, in
many cases technical assistance has been deficient, the mechanism of operation of the
projects was hampered by many other possibilities that led to an even lower rate of
absorption of fixed assets.
The level of absorption increased after the enlargement of the European Union from 2004
and then 2007 and continued along the way, but the efforts to ensure that the targets set by
the European Union were met were delayed. Thus, action still needs to be taken to reduce the
period of analysis of reimbursement requests, the allocation of funds from the Community
budget needs to be increased, these funds need to be channelled into those areas where
Member States can make individual contributions and ensure development consistency.
The indicators at the level of the main targets are still quite far in the field of transport
infrastructure, road infrastructure in the first place, environmental investments, renewable
energy projects, the achievement of economic and socio-cultural objectives. Social
infrastructure has not been rehabilitated to the extent of the objectives of the directive aimed
at increasing the income of the population and consequently the quality of life.
In the field of complex pre-university, university and post-graduate education, there are
still difficulties that must underlie large-scale European Union projects, which must ensure
not only the Bologna program but also the raising of the quality of complex education in
all Member States.
The development of human capital requires the creation of new jobs and these can only be
achieved at the expense of research projects, which ensure the creation of new jobs, which
ensure the prospect of obtaining additional income and consequently to achieve with
increasing proximity to the level of states with a high quality level of education.
We must not forget that this pandemic coronavirus crisis has highlighted a number of issues
regarding the standard of health. In this field, national projects must be undertaken, with
the support of the European Union, in order to bring this medical capacity (hospital
capacity, quality of medication capacity) to a level as close as possible to the requirements
of today's life. The coronavirus pandemic is just an acute fact that can deepen, can continue,
but which in turn must be substantially controlled.
We are in the third millennium and this pandemic can be repeated even further through
other developments that can be even more harmful, destructive. In the period 2020 and the
following years, the member countries of the European Union need operational programs
approved by the European Commission and completed to meet the 23 mentioned
conditionality’s and three other conditionality’s still on the way. The Infrastructure
Operational Program and the Competitive Operational Program must be financed in the
Member States of the European Union, so as to eliminate all shortcomings encountered
during this period.
In the following table we have highlighted the stage of absorption of funds in the period
2014-2020 by Romania, specifying that the same trend is manifested for the following
periods.
Table 2. Absorption of funds in the period 2014-2020 by Romania
Source: Ministry of European Funds.
The field of research and innovation must play an important role in the European Union's
strategy for the immediate future. The year 2020 can be considered one in which not all the
objectives that the European Union and the Member States have set for themselves can be
met due to the impact that the health crisis will have, combined with the economic and
financial crisis that will break out.
Some countries, such as Germany, Denmark and others, are heavily funding the research
system and have thus easily overcome the economic and financial crisis of 2008-2010 and
are likely to do so now to overcome the effects of the coronavirus crisis and the ensuing
economic and financial crisis. It should not be neglected that other countries and especially
those in Central and Eastern Europe due to limited resources will encounter difficulties in
proposing and initiating new programs.
Research and development policy in the field of technology and innovation has so far been
achieved through important projects in sensitive areas of economic life in each country.
Each country has its own priorities that must be combined with the priorities of the
European Union in the 2020 strategy and the period up to 2027. These priorities of the
European strategy must mean an increase based on intelligence, sustainability and
inclusion, as well as cooperation Member States.
The European Union's target of allocating 3% of Gross Domestic Product in all Member
States to investment in research, development and innovation is an important step that must
be supported in order to achieve the objectives that currently constitute individual priority
objectives, and complexes of the whole union.
All countries must make a significant contribution to finding those forms of increasing
economic competitiveness. The National Strategy for Research, Development and
Innovation (SNCDI) must be ensured through funding contracts, which must be supported
by all countries by the European Union as a whole.
It must be borne in mind that in the economic field, growth potential depends on a number
of factors. The structural analysis of each national economy of the member countries of the
union leads us to the conclusion that each country has a strength in the intra-community
market and hence the possibilities in the interest of each country, but also in the European
Union is to develop those branches, those areas that are beneficial. Of course, there have
been in the past, but from completely different criteria, a policy of graphic-economic
zoning of the states of the European Union. I do not refer to opinions that have a deep
political character, but to some they were even worthy of consideration.
For example, we can consider as priority area of the European Union the agricultural area,
of the countries that have a special natural potential and which, through national and
European Union subsidies, can lead to massive growth, ensuring the internal market, the
community market and for the countries respectively the increase of the international
potential.
The Gross Domestic Product in each country from a factorial point of view has a different
structure on resources and utilities. These resources can be the elements that underlie the
economic growth of each country. Combined economic growth would have the effect of
improving living conditions, raising the national standard of living that must become
European.
The PHYSICS nuclear infrastructure project is one of the utmost importance that was given
attention in the financial year 2014-2020 and will have to be continued, as it offers a high
potential of all states contributing to this project and of course, to increase European
performance. In other words, the European Union was created, consolidated, developed,
precisely in order, on the basis of honest cooperation, to ensure the individual economic
growth of each country and consequently the development of major investment projects
and finally ensuring the increase of the quality of life.
It is worth considering that this DANUBIUS program is a European one, through the
involvement of several countries, but also in the general European interest.
ELI-NP is selected and considered the most important scientific forum in the field of
nuclear physics in Europe.
A major infrastructure is emerging in the long-term plan for nuclear physics in Europe.
This infrastructure is LI-PHI which comprises two components:
 a laser system of great importance, with two laser bridges capable of reaching very high
intensities;
 an intense beam that can ensure extremely high performance.
This infrastructure will create a new European laboratory with a wide range of fields of
science, from basic physics to innovative nuclear physics and auto-physics with material
applications that depend on a management that must be true. Of course, this project is a
major one in the opinion of the authors for the European Union, but it will be able to evolve,
to take a special shape in the context in which the European Union that integrated it as a
program of the European Union will make the necessary efforts to reach a peak of physical
engineering with extraordinary results for the large population.
The DANUBIUS-RI project will be a pan-European research infrastructure dedicated to the
interdisciplinary studies of the appropriate system. The research will have to be a symbiotic
one between the European scientific forces and those on other meridians, so as to ensure
effective cooperation so that the final results benefit all Member States.
This DANUBIUS-RI program has funding arrangements that have been granted so far by a
number of four countries Germany, Italy, Romania and Spain. It is a political commitment
of seven other states, namely Bulgaria, Greece, Hungary, Ireland, Moldova, the
Netherlands and the United Kingdom of Great Britain and Northern Ireland. There are also
expressions of interest and support from the scientific community that have been launched
by these countries and in sixteen other European countries (Austria, Czech Republic,
Finland, France, Lithuania, Norway, Poland, Serbia, Switzerland, Turkey, Ukraine,
Azerbaijan, China, India, Morocco and the USA).
Points of interest are also expressed by other prestigious international organizations, as a
process of developing renewable energy resources will be a huge resource for humanity
given that non-renewable resources are particularly significant but in a negative sense.
Crude oil, natural gas and coal resources are quantified at the level of international
explorations as sufficient for 43 years for oil, up to 100 years for coal and natural gas.
Under these conditions, there is a very serious question of how humanity will be able to
regenerate new resources, ensure national consumption and produce substitutes that will
give essence and perspective to the evolution of humanity. After all, the European Union's
challenges in initiating projects with global validity are not their own. They are also
considered by the rest of humanity.
A number of leading researchers from around the globe have expressed in papers, some
quite deep the perspective that humanity will reach a terminus. Probably the millennial
evolution of mankind will lead to new scientific discoveries, will lead to new adjustments
in world relations, but all must be based on the correlation between population growth and
the level of resources producing goods and services in the interest of the whole mankind.
The environment, the soil and the subsoil, the riches that are in this natural environment,
the scientific capacity to capitalize on resources, human intelligence, the robotization of
economic processes and all these together are nothing but factors that must underlie an
evolution, we dare to -we call it special, of the correlated development of the population
and of the resources.
In this way, a number of projects have been discussed, some objectives that need to be
considered, and perhaps even stand before the European Union. They are challenges which
in the European Union, in the opinion of the authors, despite other views expressed in
relation to the cohesion of the European Union, must be the support for future
developments within this European Union.
The study shows that no state has joined and is not trying to join the European Union in the
future, except from a simplified point of view and the individual advantages it would gain
within the European Union. This is a principle that must underpin the leadership, through
its institutions, of the European Union in order to unite and bring together research efforts
aimed at meeting the challenges ahead.
We are going over the effects that will be quite pronounced of the sanitary crisis
(coronavirus) and of the financial-economic one that will follow and we believe that
nevertheless by coagulating the resources and the interest of the member countries special
results can be reached.
To the title question, we now answer that yes, the European Union is the appropriate
framework in which the Member States and the Union as a whole can set themselves and
fight for this goal, which should be pursued with interest and skill, so that coagulation can
be achieved the interest of Member States to increase the European Union's individual and
concentrated economic and social needs.
Conclusions
The study that formed the basis of this article, quite extensive and somewhat controversial
due to the last two events – Brexit and the health crisis (coronavirus) combined with the
economic and financial crisis – lead to a series of conclusions. First of all, the European
Union needs challenges in order to achieve a higher level of individual economic growth
of each stratum and of the European Union as a whole.
Secondly, it is easy to identify the number of challenges or the combination of challenges
that must underpin the strategies to be pursued by the European Union. Member States
need to align with the European strategy, taking into account the effects of a crisis, such as
the one we are going through at the time of writing, with a somewhat changed idea of
overcoming these difficulties and moving towards net results for the benefit of the Member
States and the European Community as a whole.
Another conclusion is that without cohesion and objective analysis of the situation of the
European budget, the situation of the level of development of each country, the general
interest of the European Union, but combined with the individual interest of each country,
the forces cannot be coagulated to overcome these challenges.
In other words, the economic and financial crisis will trigger great difficulties for most
states, especially those that joined the European Union after 2004, in terms of financial
resources. Therefore, the European Union must make a special effort to reconsider the
multiannual budget for the period 2021-2027 so that the correlation between the
contribution of the Member States and the benefit obtained by them through the funds
allocated to each state is at least neutral, not producing losses. We are thinking first of all
of the countries with a lower level of development that have entered the European Union
with thoughts and hopes of being better.
A final conclusion is that the European Union is a community of states based on the
principles of independence of each, but combined with the common effort to produce a
better world with superior results that will gradually but continuously affect the quality of
life, the standard of living of the population of this European entity. We do not want to, we
resort to the material principle namely that only making progress in economic development
and in the results obtained can coagulate that European communion that militates to meet
the challenges that lie ahead.
References
Anghelache, C., 2013. România 2013. Starea economică sub povara efectelor crizei, Economica,
Bucharest.
Anghelache, C. and Anghel, M.G., 2017. România – membră a Uniunii Europene. Zece ani de la
aderare, Economica, Bucharest.
Anghelache, C., Soare, D.V. and Dumitrescu, D., 2016a. IT&C Platform Used in Projects Financed
from European Union Funds, Romanian Statistical Review, Supplement, 6, pp. 59-67.
Anghelache, C., Anghel, M.G., Diaconu, A., Badiu, A. and Niţă, G., 2016b. Modele utilizate în
analiza absorbţiei fondurilor comunitare, Proceedings of the International Symposium
“Romania in the European Union – Methods and Models of Macroeconomic Analysis and
Prognosis”, “Artifex” University Bucharest, 19-20 mai 2016, Artifex, Bucur 4.
Barbosa, N. and Faria, A.P., 2011. Innovation across Europe: How important are institutional
differences?. Research Policy, 40 (9), pp. 1157-1169.
Berezin, M. and Diez-Medrano, J., 2008. Distance matters: Place, political legitimacy and popular
support for European integration. Comparative European Politics, 6 (2), pp. 1-32, 278-292.
Cincera, M. and Veugelers, R., 2013. Young leading innovators and the EU's R&D intensity gap.
Economics of Innovation and New Technology, 22 (2), pp. 177-198.
Chalmers, A.W., 2013. Trading information for access: informational lobbying strategies and interest
group access to the European Union. Journal of European Public Policy, 20 (1), pp. 39-58.
Dachs, B. and Pyka, A., 2010. What drives the internationalisation of innovation? Evidence from
European patent data. Economics of Innovation and New Technology, 19 (1), pp. 71-86.
Farole, T., Rodríguez-Pose, A. and Storper, M., 2011. Cohesion Policy in the European Union:
Growth, Geography, Institutions. Journal of Common Market Studies, 49 (5), pp. 1089-1111.
Goldberg, P.K. and Pavcnik, N., 2007. Distributional effects of globalization in developing
countries. Journal of Economic Literature, 45 (1), pp. 39-82.
Lane, P.R., 2006. The real effects of European monetary union, The Journal of Economic
Perspectives, 20 (4), pp. 47-66.
Lima, C. and Cardenete, A., 2007. The effects of European funds on a regional economy: an applied
general equilibrium analysis, Applied Economics Letters, 14 (11), pp. 851-855.
Tosun, J., 2014. Absorption of Regional Funds: A Comparative Analysis. Journal of Common
Market Studies, 52 (2), pp. 371-387.
Voigt, P. and Moncada-Paternò-Castello, P., 2012. Can Fast Growing R&D-Intensive Smes Affect
the Economic Structure of the Eu Economy?: A Projection to the Year 2020. Eurasian Business
Review, 2 (2), pp. 96-128.
What determines the portfolio investment flows

to Central and Eastern European Countries
in the European Union 2001-2017?
Donny TANG
Temple University, Philadelphia, USA
dnytng@gmail.com
Abstract. This study estimates whether the Central and Eastern European (CEE) financial market
integration and higher trade flows can explain the foreign portfolio investment inflows since EU
accession during 2001-2017. The results suggest that the stock market development has facilitated
the foreign portfolio equity flows during the European Union (EU) and crisis periods. But it only
has positive effect on the foreign portfolio debt flows during the EU period. In contrast, the
banking sector development has very weak effect on the foreign portfolio investment. The higher
bank development has only increased the foreign portfolio equity flows during the EU period.
Keywords: economic integration; European Union; portfolio investment; financial market.
JEL Classification: F15; F36; G15.

22 Donny Tang
1. Introduction
The European Union (EU) countries deepened their financial integration by forming the
Economic and Monetary Union (EMU) in 1992. The EMU reduced exchange rate
uncertainty through euro adoption and facilitated monetary policy convergence. They
played a crucial role in accelerating the financial market integration among the eurozone
countries. Their financial market efficiency substantially improved because of the higher
competition. More importantly, the EMU further boosted capital flows among the
eurozone countries and the rest of the world (Lane and Milesi-Ferretti, 2008). In
particular, the high level of capital flows had helped Central and Eastern Europe (CEE)
countries to achieve high economic growth for the past two decades.
The objective of this study is to identify the main determinants of foreign portfolio
investments in the CEE countries during 2001-2017. First, this study examines whether
the CEE financial market development has facilitated the higher foreign portfolio
investment inflows. Most of these countries have established more efficient financial
markets since EU and EMU accession. The EU has contributed to the equity and bond
market integration (De Santis and Gerard, 2009). The euro launch triggered by the EMU
has further increased the depth and liquidity of the eurozone financial markets (Giofre,
2012). The larger stock markets and more efficient banking sectors have positive impact
on the foreign portfolio investment inflows (Broto et al., 2011; Schmitz, 2011). This
study analyzes whether the higher CEE financial market development due to EU and
EMU integration has boosted the foreign portfolio investment inflows since 2004.
Moreover, this study also examines the financial crisis impact on the foreign portfolio
investment inflows during 2010-2017. Compared to the developed countries, the
developing countries experienced quicker rebound in capital inflows after the 2008
financial crisis (Calderon and Kubota, 2019). The countries’ macroeconomic conditions
and institutional qualities were crucial factors for foreign capital inflows during the
recovery period (McQuade and Schmitz, 2017). This study focuses on whether the CEE
countries experienced quick rebound in foreign portfolio investments during the recovery
period 2010-2017.
Second, this study examines whether the trade flows have contributed to the higher
foreign portfolio investments in the CEE countries. In particular, they have maintained
very stable export growth because of their access to western EU countries since EU
accession. The high trade flows have boosted foreign portfolio investments in the CEE
countries because high trade transactions have made the trading partners more familiar
with the CEE investment environment (Lane and Milesi-Ferretti, 2008). The EU trade
integration has helped the CEE countries to boost more foreign portfolio investments
from western EU countries. The growing CEE trade flows with the rest of the world have
also contributed to more portfolio investment inflows. This study analyzes whether the
higher trade flows have promoted the foreign portfolio investments in the CEE countries.
This study contributes to the literature in two respects. First, this is the first study to
examine the impact of financial market development on foreign portfolio investments in
the CEE countries. The stock market development has positive relationship with foreign
portfolio investments in developing countries. Those with much larger and more
What determines the portfolio investment flows to Central and Eastern European Countries 23
developed stock markets would have more stable level of portfolio investment inflows.
The banking sector development also has similar relationship with foreign portfolio
investments. The high level of bank credit flows would result in more stable level of
portfolio investment inflows (Broto et al., 2011). The financial deregulations that reduce
transaction costs would increase the availability of various financial instruments. This
would make countries’ banking system more appealing to global capital flows. This in
turn would make them more likely to hold a larger amount of foreign portfolio
investments (Araujo et al., 2015). This supports the argument that developing countries
such as the CEE countries with financially open economies would receive net capital
inflows while developed countries experience net capital outflows (Reinhardt et al.,
2013). Due to EMU accession, the CEE countries received the higher level of foreign
portfolio investments from western EU countries during 2008-2015. If the results of this
study confirm the positive impact of financial market development on foreign portfolio
investments, the CEE countries should further improve their financial market efficiency.
This study would also provide valuable suggestions on the long-term policies for CEE
financial market development. Second, this is the first study to identify the major factors
that facilitated the increase in foreign portfolio investments in the CEE countries after the
2008 financial crisis and 2010 eurozone debt crisis. The main drivers of capital flows
during the recovery period 2010-2012 included the developing countries’ macroeconomic
fundamentals, institutional qualities and policies. In contrast, the financial openness
which made the developing countries vulnerable to common global shocks played little or
no role in affecting the capital flow volatility in the CEE countries (Fratzscher, 2012).
Based on these arguments, the macroeconomic conditions would determine the rebound
in foreign portfolio investments during the recovery period. Consistent with the
expectation, the developing countries such as the CEE countries experienced much
quicker rebound in foreign portfolio investments than banking flows and foreign direct
investments after the crisis (Milesi-Ferretti and Tille, 2011). As seven of the CEE
countries joined the eurozone during 2008-2015, their integrated financial markets such
as equity markets have become more important in the world market since the mid-1990s
as their financial market expansion has continued in terms of size and liquidity
(Fratzscher, 2002). It is important to identify the major factors that can boost the foreign
portfolio investment to the pre-crisis level in these countries. If the results of this study
confirm the importance of macroeconomic conditions on foreign portfolio investments,
the CEE countries should develop more effective economic policies to withstand financial
crisis. This study would provide important suggestions for implementing good economic
policies in the long run.
The rest of the article is organized as follows. The next section reviews the previous
studies on foreign portfolio investments. Section 3 describes the analytical framework
and estimation model for the determinants of foreign portfolio investments in the CEE
countries. Section 4 presents the results and discusses their significance. Section 5
provides important implications for the long-term policies to attract more foreign
portfolio investments. Section 6 concludes.
24 Donny Tang
2. Literature review
Most of the previous studies have confirmed the positive impact of the EMU on foreign
portfolio investment. An earlier study argued that the EMU which eliminated the
exchange rate risk and reduced transaction costs for cross-border capital flows boosted
the intra-eurozone portfolio investments (Haselmann and Herwartz, 2010). Another study
found that eurozone investors reallocated more portfolio investment flows to other
eurozone countries than investors from other countries. The EMU substantially eased the
access of eurozone investors to the entire eurozone markets (De Santis and Gerard, 2009).
Due to EMU accession, the deepened financial integration contributed to the growing
importance of eurozone equity markets in the world since the mid-1990s. The more
integrated eurozone financial markets made themselves more attractive places for foreign
investments (Fratzscher, 2002). Moreover, the EMU contributed to strong convergence
among the eurozone equity portfolios. The convergence in bilateral investment barriers
facilitated by the euro adoption mainly consolidated the portfolio convergence among the
eurozone countries (Giofre, 2012). A more recent study reached similar conclusion. The
developing countries with financially open economies experienced net capital inflows
while more developed countries experienced net capital outflows. This result still held
true after controlling for various determinants of current account (Reinhardt et al., 2013).
Previous studies have analyzed the effect of financial market development on foreign
portfolio investments. The more developed stock markets and banking sectors facilitated
asset trade among local residents which thereby reduced the need for foreign portfolio
investments. But financial market development might be spurred by foreign investments
in domestic financial system (Lane and Milesi-Ferretti, 2008). The higher financial
market development measured by financial market capitalization strongly affected foreign
portfolio investment inflows (Mandilaras and Popper, 2009). A more recent study
confirmed the direct relationship between stock market development and foreign portfolio
investments. The smaller stock market size was associated with higher volatility of
portfolio investment flows. As the stock markets became more developed, the portfolio
investment flows became more stable. The banking sector development has similar
relationship with volatility of portfolio investment flows. The more developed banking
sectors in terms of higher banking credit and deposit flows were associated with more
stable portfolio investment flows (Broto et al., 2011). Moreover, the banking sector
reforms were associated with high capital inflows. Due to EU accession, the EU countries
with banking sector reforms received more foreign portfolio investment inflows
(Schmitz, 2011). The EU membership since 2004 substantially improved the financial
sector quality in the CEE countries by upgrading their legal, regulatory, and supervisory
frameworks comparable to those in western EU countries (Von Hagen and Zhang, 2014).
More related studies have explored the impact of trade flows on foreign portfolio
investments. The high trade linkage helped foreign investors to gain more information
over host countries, which increased their willingness to increase foreign portfolio
investments in these countries (Lane and Milesi-Ferretti, 2008; De Santis and Gerard,
2009). The higher bilateral trade flows had positive impact on bilateral foreign capital
flows. When two countries traded more with each other, they also held higher shares of
each other’s foreign assets which provided a better hedge for output risks (Peter, 2012).
Similar study confirmed the positive trade effect on foreign portfolio investments. The
countries that had close trade ties with the United States invested more portfolio
investments in the United States (Forbes, 2010). Another study concluded that trade flows
and financial capital flows reinforced each other. Hence, countries could boost portfolio
investment flows by better coordinating trade policies and capital liberalization policies
(Aviat and Coeurdacier, 2007).
Finally, a related strand of literature has tried to identify the main factors that can
minimize the impact of financial crisis on foreign portfolio investments. First, the
countries with high debt burden suffered more decline in capital inflows during the crisis.
In contrast, the countries with low dependence on external finance were less affected by
the capital flow volatility (Milesi-Ferretti and Tille, 2011). A later study posited that the
countries’ macroeconomic conditions, institutional qualities, and policies were the major
determinants of capital flows during the crisis recovery period 2009-2010 (Fratzscher,
2012). A similar study suggested that the eurozone countries that were severely affected
by the 2010 eurozone debt crisis should replace their financial supervision institutions by
forming supranational institutions capable of managing and resolving financial crises
(Sapir, 2011). This study would extend the literature to focus on foreign portfolio
investment patterns in developing countries. It examines whether the CEE financial
market development has played a crucial role in facilitating the foreign portfolio
investments after EU accession.
3. Econometric specification
3.1. Analytical framework
This study will identify the main determinants of foreign portfolio investments in the
CEE countries during 2001-2017. The model derives from the gravity model which
primarily explains the bilateral trade flows by distance between trading countries and
gross domestic products (Tinbergen, 1962). The gravity model has also been applied in
the analysis of foreign capital flows. This study modifies the gravity model by adding the
CEE financial development and conventional variables such as trade flow. It analyzes
whether the higher financial market development and trade flows can explain the foreign
portfolio investments in the CEE countries.
First, the euro launch triggered by the EMU creation has deepened the financial market
integration among the eurozone countries (Giofre, 2012). The more integrated financial
markets would reduce transaction costs and hence further boost cross-border trade in
financial assets (Lane, 2000). More importantly, the larger stock markets would lead to
more stable foreign portfolio investment inflows (Broto et al., 2011). The euro launch has
also increased the depth and liquidity of the eurozone financial markets. The CEE stock
markets have further expanded due to their stock market integration with western EU
countries. The larger and more liquid CEE stock markets have made themselves more
attractive places for foreign portfolio investments. The modified model examines whether
the higher CEE stock market development can explain the foreign portfolio investment
26 Donny Tang
inflows. The stock market capitalization is a good measure to predict capital flow volume
(Portes and Rey, 2005). The CEE stock market development is measured by the stock
market capitalization (MktCap).
Similar to the stock market development, the CEE banking sector development also has
positive relationship with foreign portfolio investment inflows. The more developed
banking sectors measured by their bank credits and deposits would lead to more stable
foreign portfolio investment inflows (Broto et al., 2011). Given high economic growth
potential, developing countries would experience higher inflows from developed
countries (Schmitz, 2011). Due to EU accession, the CEE countries with bank
liberalization have received more foreign portfolio investments. More importantly, EU
accession has increased the CEE banking sector efficiency through upgrading their legal,
regulatory, and supervisory frameworks comparable to those in western EU countries.
Besides, the high foreign bank presence in these countries has substantially improved the
quality of their banking sectors (Hagen and Zhang, 2014). The modified model examines
whether the CEE banking sector development can affect the foreign portfolio investment
inflows. The CEE banking sector development is measured by the domestic bank credits
to private sector (BankCred).
Finally, the trade flows can influence the foreign portfolio investments in the CEE
countries. The higher trade flows among trading countries would boost their bilateral
investment flows (Pericoli et al., 2013). Two countries that trade more with each other
tend to hold higher shares of each other’s financial assets. The reason is that these assets
provide a better hedge for output risks (Peter, 2012). The higher trade transactions would
allow trading partners to know more about their investment environments, which would
increase their willingness to make investments in these countries (De Santis and Gerard,
2009; Lane and Milesi-Ferretti, 2008). The higher trade flows in the CEE countries would
attract more foreign portfolio investments from their major trading partners. The modified
model examines whether the higher trade flows would increase the foreign portfolio
investment inflows. The trade flows between the CEE countries and their trading partners
are measured by the sum of exports and imports as a percentage of CEE GDP (Trade).
3.2. Empirical model
This study modifies the gravity model to identify the main determinants of foreign portfolio
investments in the CEE countries during 2001-2017. The model estimates whether the CEE
deeper financial market integration and higher trade flows can explain the foreign portfolio
investments since EU accession. The regression equations are given as:
log(Portfolioit) = α + β1 log(MktCapit) + β2 log(BankCredit) + β3 log(Tradeit) +
+ β4 log(FDIit) + β5 log(Manufit) + β6 log(GDPPCit) + β7 log(NFAit) +
+ β8 log(GovDebtit) + β9 log(HiTechit) + β10 log(PopSizeit) + β11 log(FinFreeit) +
+ β12 (Savingit) + εit (1)
log(Portfolioit) = α + β1 log(MktCapit) + β2 log(DomCredit) + β3 log(Tradeit) +
+ β4 log(FDIit) + β5 log(Manufit) + β6 log(GDPPCit) + β7 log(NFAit) +
β8 log(GovDebtit) + β9 log(HiTechit) + β10 log(PopSizeit) + β11 log(FinFreeit) +
+ β12 (Savingit) + εit (2)
where Portfolioit is the level of the foreign portfolio equity and debt investments in the
CEE country i in time period t (2001-2017). All variables are measured in US dollars
adjusted for inflation to the base year 2005. The CEE countries in this study refer to
eleven Central and Eastern European countries (Bulgaria, Croatia, the Czech Republic,
Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia) and two
Mediterranean countries (Cyprus and Malta). All of the CEE countries except Bulgaria,
Croatia and Romania joined the EU in 2004. Bulgaria and Romania joined the EU in
2007 and Croatia followed suit in 2013. Moreover, seven CEE countries (Slovenia,
Cyprus, Malta, Slovakia, Estonia, Latvia, and Lithuania) ultimately adopted the euro
currency between 2007 and 2015. Although the CEE countries only joined the EU and
EMU in 2004 and 2007, EU and EMU accession facilitated their financial market
integration through the elimination of currency conversion risk. The anticipated EMU
effect bolstered the growing importance of eurozone equity markets since the mid-1990s.
The more integrated CEE financial markets made themselves more attractive places for
foreign investments (Fratzscher, 2002). The main focus of this study is to examine the
impact of the CEE financial market development on foreign portfolio investments in the
CEE countries during the EU and EMU periods 2005-2017. This study also runs data
analysis for the earlier period 2001-2004 to see whether the anticipated EU and EMU
effects can explain the higher foreign portfolio investments in these countries.
The main explanatory variables in equations (1) and (2) are the stock market and banking
sector development variables (MktCap and BankCred). MktCap is the stock market
capitalization variable. It is the total value of stocks listed on the domestic market divided
by CEE GDP. It measures the stock market size relative to the economy. EU accession has
deepened the financial liberalization among the CEE countries. Their stock market
expansion due to the stock market reforms has attracted more foreign portfolio investments
(Schmitz, 2011). The larger stock market size would result in higher foreign portfolio
investment in the CEE countries. Second, BankCred is the bank credit flow variable. It is
the domestic credit provided by deposit money banks to private sectors as a share of CEE
GDP. A larger amount of bank credits indicates a higher level of financial services and
banking sector development. The foreign bank entry facilitated by EMU accession has
further boosted the bank credit supply in the CEE countries. The CEE eurozone countries
have received more foreign bank credits from western eurozone countries (Moral-Benito
and Roehn, 2016). The larger bank credit supply would lead to higher foreign portfolio
investments in the CEE countries. To test for the robustness of the results, the model would
include the banking sector variable (DomCred) which measures the domestic credits
provided by financial intermediaries to private sector as a share of CEE GDP. The variable
only measures the total credits issued to the private sectors, but excludes credits issued to
governments and public agencies. The euro launch has facilitated the banking sector
liberalization among the member countries. This has boosted the country’s holding of
foreign portfolio investment inflows (Araujo et al., 2015). Hence, the larger domestic credit
flows would result in higher foreign portfolio investments in the CEE countries.
Another related variable is the level of financial market competition that can strongly
affect the foreign portfolio investment inflows. The financial freedom (FinFree) variable
measures the financial freedom scores of the CEE countries. An efficient financial system
28 Donny Tang
ensures the availability of diversified savings, credit, payment, and investment services to
individuals and businesses. To facilitate financial market competition, government should
regulate banking institutions to ensure transparency and integrity and promote disclosure
of assets, liabilities, and risks (Miller et al., 2019). Developing countries with financially
open economies would experience net capital inflows while developed countries would
experience net capital outflows (Reinhardt et al., 2013). The EMU has accelerated the
financial liberalization in the CEE countries. The higher financial freedom would lead to
higher foreign portfolio investments in these countries.
Another major concern is whether trade flow is one of the crucial determinants for
foreign portfolio investment. Trade is the amount of CEE exports and imports divided by
CEE GDP. It reflects the trade linkages between the CEE countries and their trading
partners. The higher bilateral trade flows would result in higher bilateral capital flows
(Pericoli et al., 2013). The close trading partners tend to hold higher shares of each
other’s capitals as they provide better hedge for output risks (Peter, 2012). Another
reason why trade boosts portfolio investment is that higher trade linkages would improve
information flows over host countries and thereby increase the willingness to make
foreign portfolio investment (Lane and Milesi-Ferretti, 2008; De Santis and Gerard,
2009). EU accession has boosted the CEE trade flows with western EU countries.
Moreover, their trade flows with the rest of the world have also increased because foreign
countries exporting to the CEE countries have gained full access to the rest of the EU
markets. Higher CEE trade openness would lead to higher foreign capital flows in the
CEE countries (Schmitz, 2011). Hence, the higher trade flows would result in higher
foreign portfolio investments in these countries.
Similar to the trade variable, the foreign direct investment (FDI) variable has positive
effect on foreign portfolio investment in the CEE countries. FDI is the total amount of the
foreign direct investments in the CEE countries. Compared to portfolio investment, FDI
has a rather long-term nature and tends to be associated with domestic investment and
economic growth. It is generally less volatile and more persistent than non-FDI inflows
(Broto et al., 2011). But foreign investors would need to pay high information cost to
make FDI in host countries. The cost would increase when distance between home and
host countries increases (Guerin, 2006). The high FDI transactions would provide foreign
investors with valuable information about the CEE countries, which would substantially
reduce the information cost in making foreign portfolio investments in these countries.
Therefore, the higher FDI inflows would have positive impact on foreign portfolio
investments in the CEE countries.
Manufacturing output (Manuf) variable can have substantial effect on foreign portfolio
investment. It is the share of manufacturing output in total output. If developing countries
have higher share of manufacturing output, it would suggest their high dependence on
export revenues. The external economic shocks would make their export revenues more
volatile (Milesi-Ferretti and Tille, 2011). Since the late 1990s, the CEE countries focused
on developing manufacturing industries to boost their economic growth. However, when
their major export markets such as western EU countries experienced severe economic
slowdown, the CEE export revenues substantially decreased because of the sharp drop in
demand for their exports. Their low economic growth due to poor performance in
manufacturing industries would deter foreign portfolio investment inflows. Hence, the
declining share of manufacturing output in total output would lead to lower foreign
portfolio investments in the CEE countries.
Several conventional variables can explain the foreign portfolio investments in the CEE
countries. First, the gross domestic product per capita (GDPPC) variable measures the
level of economic development of the CEE countries. The level of economic development
would have different impacts on foreign portfolio investments. Developing countries
measured by their lower GDP per capita level would receive larger foreign portfolio
investment inflows. This is consistent with the concept of downhill net financial flows to
relatively poorer countries (Schmitz, 2011). On the other hand, developing countries
would receive much less foreign portfolio investment than developed countries because
the former have less safer investment environments. They are more likely to experience a
capital flow turnaround (Milesi-Ferretti and Tille, 2011). Therefore, the level of economic
development may have either positive or negative impact on foreign portfolio investment
in the CEE countries. Similar to the GDPPC variable, the population size (PopSize)
variable is also considered as economic size variable. It refers to the total population of
the CEE countries which include all residents regardless of their citizenship. The larger
size countries would attract more foreign capital inflows as they offer more
diversification opportunities (Papaioannou, 2009). The larger population size would
result in higher foreign portfolio investment in the CEE countries.
Second, the government debt (GovDebt) variable indicates the total government debt as a
percentage of CEE GDP. It reflects the macroeconomic fundamentals of the CEE
countries. Those countries with strong macroeconomic fundamentals would receive more
stable foreign portfolio investment as they can better insulate their financial markets from
financial crisis (Fratzscher, 2012). The stronger macroeconomic fundamentals reflected
by lower value in GovDebt would bolster foreign investors’ confidence in the CEE
countries. Hence, this would likely increase foreign portfolio investment in these
countries. A similar variable to measure country’s debt level is the net foreign assets
(NFA) variable. It measures the sum of foreign assets held by monetary authorities and
deposit money banks but excludes their foreign liabilities. It is equal to the cumulative
changes in its current account. The NFA position indicates whether the country is a net
creditor or debtor to the rest of the world. A positive NFA balance means that it is a net
lender while a negative NFA balance shows that it is a net borrower. The higher debt
level would deter foreign portfolio investment in the CEE countries.
Third, the high-technology exports measure the research and development capabilities of
CEE countries. The high-technology export proportion (HiTech) variable is equal to the
total amount of high-technology exports as a percentage of CEE GDP. High technology
exports reflect high research and development intensity such as aerospace, computers,
scientific instruments, and electrical machinery. Given the high level of research
capability, the high-technology export proportion has a significant impact on the CEE
long-term economic growth. The main positive externalities are derived from knowledge
spillovers and economies of scale (Sheridan, 2014). The local firms can learn the high
30 Donny Tang
technological content of imports and incorporate them into exports to boost their long-
term growth. The higher growth would make the CEE countries more ideal places for
foreign investors. The higher high-technology export proportion would help boost foreign
portfolio investment in the CEE countries.
Finally, the saving (Saving) variable is the gross domestic saving of the CEE countries. It
is calculated as gross domestic product less final consumption expenditure (total
consumption). Domestic saving should act as a complement rather than a substitute to
capital inflows (Verdier, 2008). Capital flows such as portfolio investments should move
to countries where they are scarce. Foreign portfolio investments are substitutes for
domestic savings as they can help finance domestic investment. Moreover, portfolio
investments can serve as a complement to domestic savings as countries with higher
savings would experience higher foreign portfolio investments. The high saving may
have positive or negative impact on foreign portfolio investment in the CEE countries.
3.3. Two-stage least squares estimation
There may be an endogeneity problem in the foreign portfolio investment, trade and FDI
variables. The higher trade and FDI flows would facilitate foreign portfolio investment in
the CEE countries because these flows would make foreign investors more familiar with
the CEE investment environment. However, the higher foreign portfolio investment
inflow would deepen the CEE ties with portfolio investors which would boost their
bilateral trade and FDI flows. To address this concern, this study uses the two-stage least
squares (2SLS) method to re-estimate the endogenous variables (Trade and FDI). The
instrumental variable (IV) would replace these variables. First, the IVs for Trade include
Inflat and Educat. Inflat is the inflation rate variable which is measured by the consumer
price index. It reflects the annual percentage change in the cost of living of average
consumers. The higher inflation rate would increase the total production costs for the
CEE countries. Hence, it would decrease exports but increase the demand for imports.
Educat is the labor force education variable. It refers to the proportion of the labor force
that has a secondary school education as a percentage of the total labor force. More
educated labor force would improve country capability to produce both low- and high-
technology export goods. It would have positive impact on exports but decrease the
demand for imports. Second, the IVs for FDI include CapForm and IntPay. CapForm is
the gross capital formation as a percentage of CEE GDP. It measures the CEE
manufacturing industry competitiveness that can help predict their future output potential.
The strong manufacturing competitiveness would help attract more FDI inflows because
higher production capabilities would improve production efficiencies of foreign
companies in the CEE countries. This in turn would boost more FDI in these countries.
IntPay is the interest payments on government debt including long-term bonds, loans, and
other debt instruments to domestic and foreign residents. The higher interest payments
would indicate that countries have huge debt burden. This would slow their overall
economic growth and deter FDI inflows. To test for the robustness of the results, the tax
(Tax) variable is also used as IV for FDI. Tax refers to the taxes on income, profits, and
capital gains that are levied on the actual or presumptive net income of individuals,
corporate profits, and capital gains on assets. This variable measures the taxes collected
as a share of total CEE taxes. The amount of taxes represents tax burden on foreign
investors. The higher tax would result in lower corporate profits for foreign investors,
thereby discouraging FDI in the CEE countries. Finally, the endogeneity problem may
also exist in the GDPPC variable. While more foreign portfolio investment inflows
would accelerate economic development, the CEE countries with higher economic
development would attract more foreign portfolio investment inflows. The IV for
GDPPC is the lagged GDPPC of the CEE countries. Since it may take at least a year
before the current GDPPC would have impact on foreign portfolio investment, the
GDPPC is lagged by a year to measure this impact.
3.4. Data sources
The data on foreign portfolio investment are taken from the International Monetary Fund
(IMF)’s Coordinated Portfolio Investment Survey. The government debt data are drawn
from the IMF’s Historical Public Debt Database. The missing data in 2016 and 2017 are
obtained from the Trading Economics Database. The financial freedom data are found in
the 2019 Index of Economic Freedom compiled by Heritage Foundation. The data for the
explanatory variables and the IVs are all drawn from the IMF’s World Development
Indicators. The missing data on market capitalization are obtained from CEIC’s Global
Databases.
4. Empirical results
4.1. Financial development effects on foreign portfolio investment
This study identifies the main determinants of foreign portfolio investments in the CEE
countries during 2001-2017. It examines whether the CEE financial market integration
and trade flows can explain the portfolio investments since EU accession. To better
estimate the integration effects, this study conducts three subperiod estimations. The
estimation results for the subperiod 2001-2004 would suggest whether the CEE financial
market reforms since the late 1990s have affected the foreign portfolio investments. The
results for the subperiod 2005-2009 would indicate whether EU accession has helped the
CEE countries to attract more foreign portfolio investments. Finally, the results for the
subperiod 2010-2017 would reveal whether the financial crisis of 2008 and eurozone debt
crisis of 2010 have adversely affected the foreign portfolio investments since 2010.
To test for the robustness of the results, equations (1) and (2) include BankCred and
DomCred as the bank development variables respectively to see if the results remain the
same. As shown in Tables 1 and 3, the results for equation (1) include BankCred as the
bank development variable. Table 1 uses IntPay as IV, whereas Table 3 uses Tax as IV.
As presented in Tables 2 and 4, the results for equation (2) include DomCred as the bank
development variable. Table 2 uses IntPay as IV, whereas Table 4 uses Tax as IV. The
results in all four tables for the foreign portfolio equity and debt flows are shown in
columns (1) to (3) and columns (4) to (6), respectively. The overall results suggest that
the stock market development has boosted the foreign portfolio equity flows to the CEE
countries during the entire period. As presented in columns (1) to (3) of Tables 1 and 2,
32 Donny Tang
the coefficients on MktCap are positive and statistically significant in all subperiods
(2001-2004, 2005-2009, and 2010-2017). In fact, the positive stock market development
effect has further increased since EU accession. As shown in columns (2) and (3) of
Tables 3 and 4, the magnitude of the coefficient has increased from 0.730 in 2005-2009
to 1.034 in 2010-2017. This suggests that the positive stock market effect on the portfolio
equity flows has appeared since the EU period 2005-2009 and persisted throughout the
financial crisis and recovery periods 2010-2017.
The stock market development has only boosted the foreign portfolio debt flows during
the EU period. As seen in column (5) of Tables 1 to 4, the coefficients on MktCap only
become positive and statistically significant in 2005-2009. There is very mixed evidence
of the stock market effect during the crisis and recovery periods. The negative
coefficients in column (6) of Tables 1 and 2 only remain marginally significant over
2010-2017, whereas the coefficients in Tables 3 and 4 are not statistically significant at
all. The overall results indicate that the CEE stock market development has facilitated the
foreign portfolio equity flows during 2005-2017. But it only has the same effect on the
foreign portfolio debt flows right after EU accession 2005-2009. This confirms the
impact of EU financial integration on the foreign portfolio equity flows to the CEE
countries. The EMU which eliminated the exchange rate risk has eased the access of
eurozone investors to the entire eurozone markets. The eurozone investors have
reallocated more portfolio investments to the other eurozone countries (De Santis and
Gerard, 2009). The results also support the prediction of neoclassical theory that
developing countries with financially open economies would receive higher capital
inflows (Reinhardt et al., 2013). As the EMU has deepened the CEE stock market
integration with western EU countries, the more developed CEE stock markets as
reflected by their larger size have stabilized the foreign portfolio investment inflows
(Broto et al., 2011). The CEE countries with larger stock market sizes have boosted both
of the foreign portfolio equity and debt inflows during the EU period.
Finally, it is noteworthy that the stock market development has stronger positive effect on
the foreign portfolio equity flows than debt flows. The stock market development has
briefly boosted the portfolio debt flows during 2005-2009. But it has strong effect on the
portfolio equity flows during 2005-2017. This can be explained by the strong
convergence of eurozone equity markets due to the EMU creation. The euro adoption has
contributed to the integration in portfolio equity markets (Giofre, 2012). This can explain
why the larger stock markets have a longer positive effect on the foreign portfolio equity
flows than debt flows during the study period.
In contrast to the stock market development, the CEE banking sector development has
very weak effect on the foreign portfolio investment in the CEE countries. The higher
bank development has only increased the foreign portfolio equity flows during the EU
period. As seen in column (2) of Tables 1 to 4, the coefficients on BankCred and
DomCred are only statistically significant over 2005-2009. This result is highly expected
because of the banking sector reforms implemented by the CEE countries since the late
1990s. Besides, the EMU has led to more foreign bank entry into the CEE countries. The
bank deregulation has promoted bank competition which has resulted in lower transaction
costs and greater variety of financial instruments. This in turn has increased cross-border
capital flows including portfolio investment flows (Araujo et al., 2015). Moreover, their
shift toward market-based banking systems has attracted more foreign portfolio equity
flows because these reforms have contributed to the larger and efficient CEE banking
sectors (Schmitz, 2011). Second, the lack of the bank development effect on the foreign
portfolio equity flows during 2010-2017 can be attributed to the outbreak of the financial
crisis in 2008 and eurozone debt crisis in 2010. The risk-averse foreign investors from
western EU countries rebalanced their portfolio investments toward non-EU countries
because of the low correlation in their financial markets (Vermeulen, 2013). This explains
the lack of bank development on the foreign portfolio equity flows during 2010-2017.
Finally, the bank development has no impact on the foreign portfolio debt flows during
the entire study period. The reason is that more efficient CEE banking sectors have
facilitated high transactions of financial assets among the local residents. This has
substantially reduced their need for external finance. As a result, foreign capitals have
played a less crucial role in providing major financing to these countries (Lane and
Milesi-Ferretti, 2008). This may account for the lack of bank development effect on the
foreign portfolio debt flows to the CEE countries.
Another related issue is whether financial freedom has any impact on foreign portfolio
investments. As noted in Tables 1 to 4, the financial freedom in the CEE countries has
promoted the foreign portfolio debt flows during the entire period 2001-2017. In contrast,
it has no effect on the foreign portfolio equity flows. To a certain extent, the overall
results are consistent with the prediction of neoclassical theory that among financially
open economies, developing countries would experience net capital inflows, whereas
developed countries would experience net capital outflows (Reinhardt et al., 2013). Both
EU and EMU accession that eliminated exchange rate risk have facilitated the financial
liberalization among the CEE countries since the 2000s. The CEE financial market
openness has boosted the foreign portfolio debt inflows for the entire study period.
4.2. Other explanatory variables affecting foreign portfolio investment
Another important concern of this study is whether trade flows would increase the foreign
portfolio investment in the CEE countries. The high trade flows only have the expected
positive effect on the foreign portfolio equity flows during the crisis and recovery
periods. As shown in column (3) of Tables 3 and 4, the coefficients on Trade are only
statistically significant during 2010-2017. In comparison, the trade flows have no effect
on the foreign portfolio debt flow at all. As seen in Tables 1 and 2, the coefficients on
Trade are not statistically significant for the entire period. The overall results suggest that
the CEE countries have further expanded trade flows with their trading partners including
non-EU countries during the crisis and recovery periods. Since trade transactions have
increased trading partners’ information about the CEE investment environment, high
trade flows have made foreign investors more willing to make portfolio investments in
these countries (Lane and Milesi-Ferretti, 2008; De Santis and Gerard, 2009). This
confirms that trade flows in goods have reinforced financial asset flows (Aviat and
Coeurdacier, 2007). This explains why the high trade flows have boosted the foreign
portfolio equity flows to the CEE countries during 2010-2017.
34 Donny Tang
Foreign direct investment has a positive effect on the foreign portfolio equity flows but
has no impact on the foreign portfolio debt flows. As presented in column (3) of Tables 1
and 2, the coefficients on FDI are only statistically significant over 2010-2017. The result
indicates that the higher foreign direct investment inflows have only boosted the foreign
portfolio equity inflows during the crisis and recovery periods 2010-2017. This is
contrary to the argument that foreign direct investment would be positively related to
foreign portfolio investment. Foreign direct investments are long-term investment
commitment. They are considered as a good proxy for information costs (Guerin, 2006).
After foreign investors obtain valuable information about the CEE countries through FDI
activities, they would experience lower information cost in making portfolio investments
in these countries. However, the results of this study do not support this argument. This
may be due to the fact that more non-EU countries have increased their shares of total
FDI in the CEE countries. While the major western EU countries such as the Netherlands,
Germany, France and Luxembourg have accounted for half of the total FDI, countries
from other parts of Europe and the rest of the world have dominated the rest of the FDI
(Szabo, 2019). In particular, the United States, Japan, South Korea, and China have
further increased their FDI in the CEE countries. Unlike western EU countries, these
countries may consider non-FDI factors to make foreign portfolio investment in the CEE
countries. This may explain the lack of FDI effect on the foreign portfolio investment in
these countries.
The high-technology export proportions have very mixed effect on the foreign portfolio
equity and debt flows. The export proportions have the expected positive impact on the
foreign portfolio equity flows during 2001-2009. As noted in columns (1) and (2) of Tables 1
and 2, the coefficients on HiTech are positive and statistically significant in 2001-2004
and 2005-2009. But the export proportions have the opposite effect on the foreign
portfolio debt flows during 2005-2017. As seen in columns (5) and (6) of Tables 3 and 4,
the coefficients become negative and statistically significant in 2005-2009 and 2010-
2017. It suggests that the export proportions have discouraged the foreign portfolio debts
flows since EU accession in 2004. The level of innovation capabilities is considered as
one of the crucial determinants for facilitating high and sustainable economic growth
(Lloyd-Ellis and Roberts, 2002). The advanced research and innovation capabilities as
reflected by high-technology export proportions would make the CEE countries more
favorable places for foreign portfolio investments. But the results of this study only
confirm this argument for the foreign portfolio equity flows.
Contrary to the expectation, the government debt has a positive effect on the foreign
portfolio debt flows. As presented in columns (4) to (6) of Tables 1 to 4, the coefficients
on GovDebt are positive and statistically significant for the entire period 2001-2017. As
seen in column (3) of Tables 1 and 2, the same effects are only found in the foreign
portfolio equity flows during the crisis and recovery periods 2010-2017. The overall results
are not consistent with the argument that countries with strong macroeconomic
fundamentals would receive more foreign portfolio investment as they can better insulate
their financial markets from financial crisis (Fratzscher, 2012). Countries with worsening
public finance would face huge capital outflows during crisis (Milesi-Ferretti and Tille,
2011). The possible explanation for this surprising result is that the CEE countries have relied
on foreign capitals to sustain their economic growth for the past two decades. Therefore, they
have run much larger debts to achieve rapid convergence with western EU countries in
output and living standards (Schmitz, 2011). This can explain why the larger amount of
government debts has boosted the foreign portfolio debt flows to the CEE countries.
5. Implications for the long-term policies to boost foreign portfolio investment inflows
The results provide very important implications for attracting more foreign portfolio
investments. First, the results indicate that the CEE banking sector development has only
played a minor role in boosting the foreign portfolio investment. To strengthen the bank
effect, the CEE countries should continue their banking sector reforms which have started
since EU accession. The quality of their legal institutions should be improved to create
more favorable environment for foreign banks. In particular, the protection of property
rights needs to be strictly enforced to facilitate cross-border bank flows (Papaioannou,
2009). As foreign investors are protected against expropriation and contract repudiation,
they would make larger bank investments in the CEE countries. Besides, the
improvement in their institutional quality through lower sovereign risks can help them to
better insulate from financial crisis (Fratzscher, 2012). The well-developed banking
sectors would improve investment environment for foreign portfolio investors. Besides
the banking sector reforms, the CEE countries should pursue much deeper bank
integration with western EU countries. The developed EU countries have made huge
investments in the CEE banking sectors since the late 1990s. The high foreign bank
presence during the 2000s contributed to the higher CEE economic growth (Schnabela
and Seckinger, 2019). The eurozone debt crisis of 2010 highlighted the importance of
replacing national bank supervision institutions by supranational institutions. The closer
EU-wide bank supervision in each EU member state can protect against financial crisis
caused by serious bankruptcy in member states (Sapir, 2011). The more integrated bank
supervisory regime would allow western EU banks to allocate bank capitals more
efficiently in the CEE countries. This can help them to attract more foreign portfolio
investments from western EU countries.
Second, the results suggest that the CEE stock market development has strong impact on
facilitating the foreign portfolio investment in the CEE countries. The higher stock
market development has accelerated the portfolio equity market integration among the
eurozone countries. The euro adoption triggered by the EMU formation has reduced
market imperfections such as high transaction costs (Haselmann and Herwartz, 2010).
The EMU has created a single market in financial services and integrated money and
credit markets. More importantly, the high depth and liquidity of a single eurozone
financial market have boosted capital flows with non-member countries (Lane and
Milesi-Ferretti, 2008). To attract more foreign portfolio investment in the long run, the
CEE countries should deepen the stock market integration with western eurozone
countries. As the CEE stock markets would receive more capital flows from developed
eurozone countries, their stock market size would further expand. They can have better
access to more external financing including foreign portfolio investment. To achieve the
deeper stock market integration, the CEE countries should continue stock market reforms
36 Donny Tang
to establish proper institutional and corporate governance framework. Better market

regulatory and supervisory mechanisms would boost their appeal to foreign investors. In
particular, they should improve the legislation for shareholders’ protection which can
alleviate information problems especially monitoring costs. This would substantially
reduce information constraints that can hinder portfolio diversification in eurozone
countries (Giofre, 2017). The well-developed CEE stock markets can help boost foreign
portfolio investments from both eurozone and non-eurozone countries.
Finally, the CEE countries should further control global capital inflows to better insulate
themselves from financial crisis. The economic policy uncertainty in Germany, France,
and Italy would lead to volatile environment which can facilitate risk spillovers among
the eurozone financial markets. In fact, the risk spillovers triggered by the 2008 financial
crisis occurred among the eurozone and United States financial markets (Apergisad et al.,
2019). To avoid similar crisis in future, they should mitigate economic uncertainty to
reduce the risk of destabilization of eurozone financial market performance (Bernal et al.,
2016). Foreign investors that are less familiar with the eurozone countries are more likely
than domestic investors to exit the eurozone financial markets during crisis period
(Galstyan and Lane, 2013). The financial market integration has made the CEE
economies very vulnerable to external shocks in western EU countries. In the long run,
the CEE countries should establish institutions to monitor foreign capital inflows especially
from their major allies. They should further control the debt level due to huge capital
inflows from the eurozone countries. More stringent legislations should be made to
maintain stable level of foreign capital inflows including portfolio investments. They
should impose capital control which can reduce the risk associated with sudden reversal of
capital inflows (Cardarelli et al., 2010). Furthermore, the CEE financial institutions should
closely monitor the allocation of foreign portfolio investments to productive investments.
This can reduce the risk of massive foreign capital withdrawal in case of financial crisis
outbreak. As the CEE countries adopt appropriate capital control policy, the stable foreign
portfolio investment inflow can contribute to their long-term economic growth.
6. Conclusion
This study identifies the main determinants of foreign portfolio investments in the CEE
countries during 2001-2017. It focuses on whether the CEE financial market integration
and trade flows can explain the foreign portfolio investments since EU accession. First,
the results suggest that the stock market development has facilitated the foreign portfolio
equity flows during the EU and crisis periods. But it only has positive effect on the
foreign portfolio debt flows during the EU period. The overall results confirm the positive
EMU effect on the CEE stock market integration with western EU countries. The larger
CEE stock markets have attracted more foreign portfolio investments. In contrast to the
stock market development, the banking sector development has very weak effect on the
foreign portfolio investment. The higher bank development only has increased the foreign
portfolio equity flows during the EU period. There is no bank development effect on the
foreign portfolio debt flows during the entire period. The CEE bank reforms which have
emphasized the market-based banking systems have created the larger and deeper banking
sectors. Besides, EMU accession has led to more foreign bank entry into the CEE countries.
This has resulted in the higher bank competition in these countries. All these have
facilitated the higher foreign portfolio equity inflows. Moreover, the reason for the lack of
bank development impact on the foreign portfolio debt flows is that the higher CEE bank
development has reduced their need for external finance. More developed banking sectors
have facilitated high transactions of financial assets among the local residents.
References
Apergisad, N., Christoub, C. and Kynigakisc, I., 2019. Contagion across US and European
Financial Markets: Evidence from the CDS Markets. Journal of International Money and
Finance 96, pp. 1-12.
Araujo, Pedro de, Mykhaylova, O. and Staveley-O’Carroll, J., 2015. Financial Liberalization and
Patterns of International Portfolio Holdings. Empirical Economics 49(1), pp. 213-234.
Aviat, A. and Coeurdacier, N., 2007. The Geography of Trade in Goods and Asset Holdings.
Journal of International Economics 71(1), pp. 22-51.
Bernal, O., Gnabo, J.-Y. and Guilmin, G., 2016. Economic Policy Uncertainty and Risk Spillovers
in the Eurozone. Journal of International Money and Finance 65, pp. 24-45.
Broto, C., Díaz-Cassou, J., and Erce, A., 2011. Measuring and Explaining the Volatility of Capital
Flows to Emerging Countries. Journal of Banking and Finance 35(8), pp. 1941-1953.
Calderon, C. and Kubota Megumi. 2019. Ride the Wild Surf: An Investigation of the Drivers of
Surges in Capital Inflows. Journal of International Money and Finance 92, pp. 112-136.
Cardarelli, R., Elekdag, S. and Kose, M.A., 2010. Capital Inflows: Macroeconomic Implications
and Policy Responses. Economic Systems 34(4), pp. 333-356.
Forbes, K. 2010. Why Do Foreigners Invest in the United States? Journal of International
Economics 80(1): pp. 3-21.
Fratzscher, M. 2002. Financial Market Integration in Europe: On the Effects of EMU on Stock
Markets. International Journal of Finance and Economics 7(3), pp. 165-193.
Fratzscher, M. 2012. Capital Flows, Push versus Pull Factors and the Global Financial Crisis.
Journal of International Economics 88(2), pp. 341-356.
Galstyan, V. and Lane, P.R., 2013. Bilateral Portfolio Dynamics during the Global Financial
Crisis. European Economic Review 57(1), pp. 63-74.
Giofre, M., 2012. Convergence of EMU Equity Portfolios. Open Economic Review 23(2), pp. 381-419.
Giofre, M., 2017. Financial Education, Investor Protection and International Portfolio
Diversification. Journal of International Money and Finance 71, pp. 111-139.
Guerin, S.S., 2006. The Role of Geography in Financial and Economic Integration: A Comparative
Analysis of Foreign Direct Investment, Trade and Portfolio Investment Flows. World
Economy 29(2), pp. 189-209.
Hagen, J. Von and Zhang, H., 2014. Financial Development, International Capital Flows, and
Aggregate Output. Journal of Development Economics 106, pp. 66-77.
Haselmann, R. and Herwartz, H., 2010. The Introduction of the Euro and its Effects on Portfolio
Decisions. Journal of International Money and Finance 29(1), pp. 94-110.
Lane, P.R., 2000. International Investment Positions: A Cross-Sectional Analysis. Journal of
International Money and Finance 19(4), pp. 513-534.
38 Donny Tang
Lane, P.R. and Milesi-Ferretti, G.M., 2008. The Drivers of Financial Globalization. American
Economic Review 98(2), pp. 327-332.
Lloyd-Ellis, H. and Roberts, J., 2002. Twin Engines of Growth: Skills and Technology as Equal
Partners in Balanced Growth. Journal of Economic Growth 7(2), pp. 87-115.
Mandilaras, A. and Popper, H., 2009. Capital Flows, Capitalization, and Openness in Emerging
East Asian Economies. Review of International Economics 17(4), pp. 734-750.
McQuade, P. and Schmitz, M., 2017. The Great Moderation in International Capital Flows:
A Global Phenomenon? Journal of International Money and Finance 73, pp. 188-212.
Milesi-Ferretti, G.-M. and Tille, C., 2011. The Great Retrenchment: International Capital Flows
during the Global Financial Crisis. Economic Policy 26(66), pp. 285-342.
Miller, T., Kim, A. and Roberts, J., 2019. 2019 Index of Economic Freedom: 25th Anniversary
Edition. The Heritage Foundation, Washington DC.
Moral-Benito, E. and Roehn, O., 2016. The Impact of Financial Regulation on Current Account
Balances. European Economic Review 81, pp. 148-166.
Papaioannou, E., 2009. What Drives International Financial Flows? Politics, Institutions and Other
Determinants. Journal of Development Economics 88(2), pp. 269-281.
Pericoli, F.M., Pierucci, E. and Ventura, L., 2013. Cross-Border Equity Portfolio Choices and the
Diversification Motive: A Fractional Regression Approach. Economics Letters 121(2),
pp. 282-286.
Peter, A., 2012. Bilateral Trade, Openness, and Asset Holdings. Open Economies Review 23(4),
pp. 713-740.
Portes, R. and Rey, H., 2005. The Determinants of Cross-Border Equity Flows. Journal of
International Economics 65(2), pp. 269-296.
Reinhardt, D., Ricci, L.A. and Tressel, T., 2013. International Capital Flows and Development:
Financial Openness Matters. Journal of International Economics 91(2), pp. 235-251.
De Santis, R.A. and Gerard, B., 2009. International Portfolio Allocation: Diversification Benefits
and European Monetary Union. European Economic Review 53(8), pp. 1010-1027.
Sapir, A., 2011. Europe after the Crisis: Less or More Role for Nation States in Money and
Finance? Oxford Review of Economic Policy 27(4), pp. 608-619.
Schmitz, M., 2011. Financial Reforms and Capital Flows to Emerging Europe. Empirica 38(4),
pp. 579-605.
Schnabela, I. and Seckinger, C., 2019. Foreign Banks, Financial Crises and Economic Growth in
Europe. Journal of International Money and Finance 95, pp. 70-94.
Sheridan, B., 2014. Manufacturing Exports and Growth: When is a Developing Country Ready to
Transition from Primary Exports to Manufacturing Exports? Journal of Macroeconomics 42,
pp. 1-13.
Szabo, S., 2019. FDI in the Czech Republic: A Visegrád Comparison. European Economy
Economic Briefs 42, pp. 1-15.
Tinbergen, J., 1962. Shaping the World Economy: Suggestions for an International Economic
Policy. The Twentieth Century Fund, New York, NY.
Verdier, G., 2008. What Drives Long-Term Capital Flows? A Theoretical and Empirical
Investigation. Journal of International Economics 74(1), pp. 120-142.
Vermeulen, R., 2013. International Diversification during the Financial Crisis: A Blessing for
Equity Investors? Journal of International Money and Finance 35, pp. 104-123.
Von Hagen, J. and Zhang, H., 2014. Financial Development, International Capital Flows, and
Aggregate Output. Journal of Development Economics 106, pp. 66-77.
Table 1. 2SLS Estimation of the determinants of foreign portfolio investments in the CEE countries
Equity Equity Equity Debt Debt Debt
(1) (2) (3) (4) (5) (6)
2001-2004 2005-2009 2010-2017 2001-2004 2005-2009 2010-2017
BankCred -0.065 2.122*** 0.208 0.041 0.705 0.148
(-0.263) (2.435) (0.582) (0.218) (1.129) (0.701)
MktCap 0.557* 0.946*** 0.807*** 0.029 0.452*** -0.238*
(1.933) (4.125) (3.788) (0.129) (2.753) (-1.889)
Trade 2.253 -0.340 1.517 0.229 -0.361 -0.219
(0.778) (-0.183) (1.348) (0.102) (-0.273) (-0.330)
FDI -0.796 -0.439 0.474** 0.969 -0.328 -0.150
(-0.550) (-1.336) (2.010) (0.863) (-1.396) (-1.074)
Manuf -3.932** -1.324 -1.385* -2.756** -0.288 0.286
(-2.331) (-0.913) (-1.896) (-2.106) (-0.277) (0.662)
GDPPC -0.007 -0.012 0.132 0.620 -0.345 -0.064
(-0.011) (-0.013) (0.293) (1.403) (-0.528) (-0.236)
NFA -0.108 0.261 -0.332*** -0.476* -0.104 -0.084
(-0.318) (1.277) (-3.061) (-1.810) (-0.713) (-1.310)
GovDebt 0.091 -0.150 0.398** 0.970** 0.519* 1.512***
(0.172) (-0.360) (2.154) (2.377) (1.748) (13.801)
HiTech 0.530* 1.006*** 0.013 0.139 0.082 -0.271
(1.676) (2.245) (0.041) (0.569) (0.256) (-1.420)
PopSize 1.825*** 1.754*** 1.067*** 1.196** 0.910*** 0.664***
(2.625) (4.450) (5.255) (2.217) (3.229) (5.518)
FinFree 0.666 1.527 -0.628 0.758 1.575** 2.308***
(0.832) (1.295) (-0.868) (1.221) (1.869) (5.376)
Saving 3.480*** 0.500 1.672** 1.630** 1.235* 1.737***
(3.690) (0.492) (2.288) (2.229) (1.702) (4.008)
Adjusted R2 0.756 0.650 0.625 0.674 0.772 0.864
Observations 52 65 104 52 65 104
Notes: 2SLS refers to the two-stage least squares estimation.
All variables are in logarithm. T-statistics are reported in parentheses.
***, **, * indicate significance at 1%, 5%, and 10%.
40 Donny Tang
Table 2. 2SLS estimation of the determinants of foreign portfolio investments in the CEE countries
(1) (2) (3) (4) (5) (6)
2001-2004 2005-2009 2010-2017 2001-2004 2005-2009 2010-2017
DomCred -0.065 2.122*** 0.208 0.040 0.705 0.148
(-0.266) (2.429) (0.582) (0.210) (1.131) (0.701)
MktCap 0.557* 0.947*** 0.807*** 0.029 0.452*** -0.238*
(1.936) (4.120) (3.788) (0.132) (2.753) (-1.889)
Trade 2.245 -0.349 1.517 0.217 -0.363 -0.220
(0.776) (-0.183) (1.349) (0.097) (-0.274) (-0.330)
FDI -0.793 -0.440 0.473** 0.965 -0.328 -0.150
(-0.549) (-1.337) (2.010) (0.863) (-1.397) (-1.074)
Manuf -3.928** -1.324 -1.385* -2.754** -0.285 0.287
(-2.331) (-0.910) (-1.896) (-2.110) (-0.274) (0.661)
GDPPC -0.006 -0.013 0.133 0.620 -0.345 -0.063
(-0.011) (-0.014) (0.293) (1.405) (-0.529) (-0.237)
NFA -0.107 0.261 -0.332*** -0.474* -0.104 -0.084
(-0.318) (1.277) (-3.061) (-1.811) (-0.713) (-1.309)
GovDebt 0.092 -0.150 0.398** 0.971** 0.520* 1.513***
(0.175) (-0.360) (2.154) (2.385) (1.748) (13.800)
HiTech 0.530* 1.008*** 0.013 0.141 0.082 -0.271
(1.679) (2.247) (0.041) (0.575) (0.257) (-1.420)
PopSize 1.824*** 1.752*** 1.067*** 1.192** 0.910*** 0.664***
(2.627) (4.446) (5.255) (2.217) (3.229) (5.518)
FinFree 0.669 1.528 -0.628 0.762 1.576* 2.308***
(0.836) (1.293) (-0.868) (1.230) (1.868) (5.376)
Saving 3.480*** 0.500 1.672** 1.628** 1.234* 1.737***
(3.697) (0.492) (2.288) (2.229) (1.700) (4.008)
Adjusted R2 0.756 0.650 0.625 0.674 0.772 0.864
Observations 52 65 104 52 65 104
(1) (2) (3) (4) (5) (6)
2001-2004 2005-2009 2010-2017 2001-2004 2005-2009 2010-2017
BankCred -0.078 1.654* 0.407 0.038 0.602 0.324
(-0.303) (1.917) (0.840) (0.249) (1.088) (1.098)
MktCap 0.537* 0.729*** 1.034*** 0.095 0.331** -0.054
(1.919) (3.179) (3.878) (0.553) (2.250) (-0.327)
Trade 2.737 2.621 2.946*** -0.999 1.425 0.851
(1.118) (1.253) (3.229) (-0.668) (1.065) (1.513)
FDI -0.961 0.119 0.389 0.556 -0.020 -0.227
(-0.974) (0.353) (1.346) (0.920) (-0.095) (-1.274)
Manuf -4.225** -1.357 -1.167 -2.335** -0.657 0.436
(-2.348) (-0.837) (-1.536) (-2.111) (-0.633) (0.934)
GDPPC -0.110 -0.341 0.533 0.615 -0.004 0.281
(-0.177) (0.251) (0.778) (1.633) (-0.009) (0.667)
NFA -0.102 -0.150 -0.439*** -0.369*** -0.310*** -0.168***
(-0.413) (-0.806) (-3.821) (-2.440) (-2.595) (-2.376)
GovDebt -0.013 -0.197 0.271 1.170*** 0.662** 1.417***
(-0.023) (-0.567) (1.500) (3.714) (1.925) (12.735)
HiTech 0.538*** -0.025 -0.187 0.251** -0.399* -0.409**
(2.770) (-0.068) (-0.562) (2.107) (-1.733) (-1.999)
PopSize 1.915*** 2.005*** 1.272*** 0.871*** 1.165*** 0.820***
(4.287) (3.925) (6.745) (3.186) (3.566) (7.061)
FinFree 0.617 0.885 -1.235** 0.944** 0.870 1.845***
(0.803) (0.587) (-1.747) (2.006) (0.902) (4.241)
Saving 3.402*** 1.418 1.238 1.645*** 1.991*** 1.388***
(3.407) (1.001) (1.557) (2.690) (2.416) (2.835)
Adjusted R2 0.732 0.662 0.594 0.776 0.823 0.841
Observations 52 65 104 52 65 104
42 Donny Tang
(1) (2) (3) (4) (5) (6)
2001-2004 2005-2009 2010-2017 2001-2004 2005-2009 2010-2017
DomCred -0.078 1.644* 0.407 0.037 0.599 0.324
(-0.306) (1.914) (0.847) (0.233) (1.085) (1.097)
MktCap 0.537* 0.730*** 1.034*** 0.095 0.331** -0.054
(1.927) (3.179) (3.870) (0.553) (2.260) (-0.327)
Trade 2.736 2.603 2.946*** -1.004 1.479 0.850
(1.117) (1.253) (3.224) (-0.668) (1.061) (1.512)
FDI -0.957 0.117 0.389 0.559 -0.021 -0.227
(-0.970) (0.354) (1.347) (0.923) (-0.099) (-1.273)
Manuf -4.221*** -1.356 -1.266 -2.333** -0.655 0.437
(-2.340) (-0.837) (-1.536) (-2.106) (-0.631) (0.934)
GDPPC -0.109 -0.354 0.533 0.616 -0.008 0.281
(-0.177) (-0.251) (0.778) (1.632) (-0.009) (0.666)
NFA -0.102 -0.150 -0.439*** -0.368*** -0.310*** -0.168***
(-0.414) (-0.806) (-3.821) (-2.435) (-2.595) (-2.376)
GovDebt -0.012 -0.202 0.271 1.171*** 0.662** 1.418***
(-0.023) (-0.367) (1.500) (3.708) (1.923) (12.739)
HiTech 0.538*** -0.022 -0.187 0.250** -0.398* -0.409**
(2.771) (-0.068) (-0.562) (2.099) (-1.731) (-1.999)
PopSize 1.914*** 2.005*** 1.272*** 0.871*** 1.163*** 0.820***
(4.288) (3.925) (6.745) (3.176) (3.570) (7.062)
FinFree 0.619 0.893 -1.235* 0.947** 0.874 1.845***
(0.805) (0.587) (-1.747) (2.006) (0.907) (4.241)
Saving 3.403*** 1.412 1.238 1.644*** 1.988*** 1.388***
(3.414) (1.160) (1.557) (2.686) (2.414) (2.835)
Adjusted R2 0.733 0.662 0.594 0.775 0.824 0.841
Observations 52 65 104 52 65 104
Analysis of the correlation between taxation indicators

and economic growth at the macroeconomic level
in European Union and Romania
Cătălina MOTOFEI
The Bucharest University of Economic Studies, Romania
catalina.motofei@cig.ase.ro
Abstract. In this paper, the author attempts to verify, on the basis of a selected set of data, if the
influence of the causal variables (taxation indicators), namely Taxes on production and imports less
subsidies and Taxes less subsidies on products on the Gross Domestic Product falls under the
category of Granger causalities, and the research hypotheses are defined accordingly. Since the
variables were supposed to be, and actually found to be non-stationary, the chosen research method
is based on the Toda-Yamamoto approach, adapted to the structure of the data. The results allow
the validation of the hypotheses in a limited number of cases, the analysis aimed to test the
causalities between the nominal values and, on the other hand, their elasticities.
Keywords: causality, taxation, products, imports, GDP.
JEL Classification: H71, O40.

44 Cătălina Motofei
Introduction and literature review

The structure of the Gross Domestic Product includes the taxation component, and the main
objective of this research is to study the possible Granger causality between two taxation
indicators and the Gross Domestic Product (a measure of economic growth), in European
Union and Romania. This objective is determined upon the study of relevant literature, and
the chosen causal indicators cover products, production and imports.
International literature abounds in studies designed to analyze the connection between
taxation and the economic growth of a country (Stoilova and Patonov, 2013; Pessino and
Fenochietto, 2010; Tiwari and Mutascu, 2014).
Some studies considered that the level of economic development of a country influences
the level of taxes. Pessino and Fenochietto (2010) found, in their study developed on 96
countries, a significant connection between tax revenues-as percentage of the GDP
(as dependent variable) and the level of development of a country- as GDP per capita. They
consider that countries with a high level of per capita GDP may have high tax revenue as
percentage of the GDP. Tanzi and Zee (2000) consider that, in theory, there is a causality
connection that runs from economic development to the tax level.
Other studies support the hypothesis that the level of taxes determines the economic growth
of a country, due to taxation. Taxation is viewed as a fundamental element of the economy,
seeing that it contributes to the gathering of necessary funds for public expenditures and,
at the same time, supports the allocation of resources and income redistribution, and
contributes to the economic stability and the economic growth of a country (Stoilova and
Patonov, 2013; Stoilova, 2017).
Taxes may also have an impact on economic growth through investment in R&D and return
on capital accumulation (Tiwari and Mutascu, 2014), Păunică et al. (2019) have assessed
the impact of remittances on economic growth. In addition, taxation can impact economic
development through its influence on basic elements such as human and physical capital,
and productivity (Stoilova and Patonov, 2013).
Stoilova (2017) is among the authors that consider that tax revenues may have a positive
influence on economic growth. But, despite the benefits that taxation can generate, there is
no consensus in the international literature in regard to the correlation between taxes and
GDP: some studies report a positive connection, while others a negative one (Tiwari and
Mutascu, 2014).
This can also be seen in the study of Vasiliauskaitė and Stankevičius (2009), which
analyzes the interdependence connection between changes of tax burden and changes in
GDP per capita for EU countries, from 1995 to 2007 and find that in less rich countries, an
increase of taxes can negatively influence the GDP growth. In return, in developed
countries, a slow growth in tax variation has a positive influence on the increase of the
GDP, while a decrease of taxes determines a decrease of the GDP. The correlation between
changes in tax burden and GDP is considered to be stronger in countries with a stable tax
system and a historically high tax burden.
Analysis of the correlation between taxation indicators and economic growth 45
Taking in to consideration the importance of taxation and the mixed results generated by
previous analysis, many studies focus on the influence that tax structure has on the GDP,
or the economic growth (Stoilova and Patonov, 2013; Stoilova, 2017).
Tax structure can be seen as the key to understand the impact that the tax system of a
country has on its economic development.
By splitting tax revenues into direct taxes, indirect taxes and social contribution, Stoilova
and Patonov (2013) consider that a tax structure that relies on direct taxes contributes more
to the economic growth, based on their regression analysis of 27 EU countries for the period
1995-2010. The authors consider that indirect taxes are subject of inequity, so there are less
efficient.
By splitting tax revenues in to different categories, based on their source, we can observe
that international literature reports different results. For example, some studies report that
corporate and personal income taxes negatively influence economic growth, in oppose to
consumption and property taxes that are less harmful (Stoilova and Patonov, 2013).
Stoilova and Patonov (2013) considered that taxes on land, buildings and other types of
structure, income taxes and social contribution taxes have a positive influence on economy
growth. Stoilova (2017) conclude, from their study on 28 EU countries for the period 1996-
2013, that a tax structure that focuses on consumption, personal income and property taxes
will contribute more to the economic growth (are less harmful).
In this context, we wonder what about taxes on production and import? Are they important?
Are these types of taxes influencing the GDP? To what extent? Can tax policies that rely
on taxes on production and import be efficient for the economic development?
Taxes on production and imports are considered a part of tax revenues (Stoilova and
Patonov, 2013). Moreover, taxes less subsidies on production in considered a component
of the GDP in the income approach (Lacey, 2000). Therefore, we estimate that taxes on
production and imports will have an influence on the GDP. But taxes on production and
imports are also considered indirect taxes (Spoerer, 1998; Tiwari and Mutascu, 2014), and
indirect taxes are a component of the GDP (Alesina and Ardagna, 2010).
If indirect taxes are more harmful, then the taxes on production and imports will have a
positive or a negative impact on the economic development? Will they contribute to the
economic growth or diminish it?
The test performed by Tiwari and Mutascu (2014) on the particular case of the United
States of America, for the period 1947-2009, shows that taxes on production and imports
are causing, based on the Granger-causality test, the GDP.
Stoilova and Patonov (2013) found a quadratic connection between taxes on production
and imports and the growth rate of the GDP, concluding that these particular taxes are not
very effective for economic development. At the same time, Stoilova (2017) reports that
taxes on production and imports have a strong positive connection with economic growth.
Păunică et al. (2018) analyzed the influence of foreign trade on the Gross Domestic
Product, considering the foreign trade as a component of the globalization.
In regard to production, if we see taxation as taxes from income and taxes from production,
studies consider that taxes on production may be less evil in their impact on growth on the
long-run (Tanzi and Zee, 2000).
In regard to import, Pessino and Fenochietto (2010) consider that taxes depend, among
others, by the level of openness of an economy. The authors report different opinions from
the literature regarding this factor of influence. On one hand, an open economy can reduce
taxes on import and export, while increasing exports. These measures, in turn, can lead to
a decrease in taxes revenue. On the other hand, liberalization may increase tax revenues,
due to compensatory measures and improved customs procedures. But, while most
countries do not perceive taxes on export, imports are still restrained by taxes (Stoilova and
Patonov, 2013; Pessino and Fenochietto, 2010; Hines and Desai, 2005).
This aspect may encourage exports at the expense of internal consumption and imports
(Hines and Desai, 2005). Nevertheless, Hines and Desai, 2005 found a negative connection
between a country’s VAT reliance and international trade, regardless of the fact that export
are not, as imports are, subject to the VAT. That means that countries relying on VAT may
experience a decrease in exports and imports. But is a decrease in import detrimental to the
economic growth? How important are imports to the economic development of a country?
These are some of the main questions that we seek to answer in this present paper.
Research methodology and data

The character of non-stationary variable associated with the Gross Domestic Product
involves the application of a method designed to deal with such parameters. The taxation
indicators chosen are also components of the GDP, but the author wishes to outline the
existence of a Granger causality of those indicators on the GDP. All indicators are
macroeconomic, and the GDP was selected, in this article, the measure of the economic
growth. The analysis pursues the situation in Romania (author’s country) and the European
Union as a whole.
The following indicators have been selected for analysis:
a) Causal variables:
a. Taxes on production and imports less subsidies. This indicator is included in the
Eurostat dataset Annual national accounts (nama10). It is used in the computation
of the Gross Domestic Product through the income approach.
b. Taxes less subsidies on products. Extracted from the same dataset.
b) Target (dependent) variable:
a. Gross Domestic Product. Is the main indicator of the nama10 dataset, GDP and main
components (output, expenditure and income) [nama_10_gdp]
Based on the indicators, two research hypotheses have been defined:

H1. Taxes on production and imports less subsidies Granger causes the Gross Domestic
Product in the European Union and Romania.
H2. Taxes less subsidies on products Granger causes the Gross Domestic Product in the
European Union and Romania.
The data covers the interval between 1996 and 2019. All data were extracted from the
Eurostat database, and processed in EViews®(1), on the basis of guidelines of Giles (2011)
for applying Toda-Yamamoto methodology, adapted to our data panel:
a) Unit root tests. The purpose for their application is the determination of the maximum
order of integration for each pair of variables that allow the evaluation of a research
hypothesis. The author chose the Augmented Dickey-Fuller (ADF) and Phillips-Perron
tests, set with the following parameters:
a1. ADF: a maximum of five lags are included (as suggested by the software), the lag
selection is done automatically by the Schwarz Info Criterion. The evolutionary
characteristics of the variables led to the choice of the Trend and intercept option.
a2. PP: the Trend and intercept option is applied in this case too, together with the
default (Bartlett kernel) spectral estimation method, bandwidth is automatically
selected by the Newey-West method.
b) VAR estimation. For each pair of variables, an unrestricted Vector Autoregressive
(VAR) model is estimated and adjusted to the optimum number of lags, out of a
maximum of 5. This choice pursues the lag length indicated by the majority of the test
criteria and, if no valid conclusion is reached, the author resorted to the opinion of
Chirilă and Chirilă (2017), who plead for the Schwarz Info Criterion as being the most
appropriate for limited samples (in the present article, each indicator has a number of
24 observations).
c) Specification tests for each model. The models have been subjected to the stability (one)
and residual tests (three) in order to check for proper specification:
c1. AR roots test for stability.
c2. Autocorrelation LM test, for a number of lags equal to the one suggested by the
software.
c3. Normality test, with the orthogonalization method Cholesky of covariance
(Lutkepohl).
c4. White Heteroskedasticity test (no cross terms).
Considering both Giles (2011) and Hacker and Hatemi-J (2003), only models that pass all
tests are used in the next step. As Giles (2011), for the first stability and autocorrelation
test, the number of lags can be gradually extended by one unit (the author applies this
approach until the model passes the test or the number of observations makes impossible
to further increase the lag length). But, failure of normality or heteroskedasticity tests leads
to the interruption of the procedure for the model.
d) Modification of the VAR model (optimum lag length, eventually corrected against
stability and/or autocorrelation issues), by adding, as exogenous variables, the
additional number of lags corresponding to the maximum order of integration.
e) Application of the Wald test for the updated model.
f) Interpretation of the results – is there any Granger causality? Is it one-way or bidirectional?
If it exists, it can allow for the validation of the research hypothesis?
Results and discussions

1. H1 testing – European Union
1.1. Nominal values
The unit root tests have returned the following orders of integration for the two variables:
Table 1. Orders of integration, H1 hypothesis, European Union, nominal values
Variable ADF test PP test
Gross Domestic Product 1 1
Taxes on production and imports less subsidies 1 1
Maximum order of integration 1
Source: author’s representation of EViews results for ADF and PP tests.
As both variables are integrated of order one, the maximum order of integration is also
1. All tests reported the same value, so no reconciliation is necessary.
The VAR model was assessed as having the optimum form of VAR(5), a stable model but
with autocorrelation issues. The residuals for VAR(6) present also serial correlation,
VAR(7) is not stable, and this is the maximum number of lags permitted by the number of
observations. The autocorrelation tests were applied for a number of l+1 lags, where l is
the maximum limit of the model’s lag length.
According to the interpretation rules for the VAR specification tests, the model cannot be
used to test Granger causality.
1.2. Logarithm values
Subsequent to the unit root tests, the individual and maximum orders of integration have
been found to have the same values as the nominal values, as presented in Table 2.
Table 2. Orders of integration, H1 hypothesis, European Union, logarithm values
All lag order selection criteria have indicated the VAR(5) as the optimum designed model,
which responds well to the AR roots test. Serial correlation is found at the level of the
residuals (tests up to 12 lags), but VAR(6) passes the first two tests, the residuals are found
to be normal, but the heteroskedasticity test cannot be applied with the use of the software.
Given the fact that the model passed the initial three tests, and not failed the fourth one, the
procedure can be continued, but the results are to be treated with caution. The updated
model can be represented as such (without substituted coefficients):
LEUIT = C(1,1) × LEUIT(-1) + C(1,2) × LEUIT(-2) + C(1,3) × LEUIT(-3) + C(1,4) ×
LEUIT(-4) + C(1,5) × LEUIT(-5) + C(1,6) × LEUIT(-6) + C(1,7) × LEUG(-1) + C(1,8) ×
LEUG(-2) + C(1,9) × LEUG(-3) + C(1,10) × LEUG(-4) + C(1,11) × LEUG(-5) + C(1,12)
× LEUG(-6) + C(1,13) + C(1,14) × LEUIT(-7) + C(1,15) × LEUG(-7)
LEUG = C(2,1) × LEUIT(-1) + C(2,2) × LEUIT(-2) + C(2,3) × LEUIT(-3) + C(2,4) ×
LEUIT(-4) + C(2,5) × LEUIT(-5) + C(2,6) × LEUIT(-6) + C(2,7) × LEUG(-1) + C(2,8) ×
LEUG(-2) + C(2,9) × LEUG(-3) + C(2,10) × LEUG(-4) + C(2,11) × LEUG(-5) + C(2,12)
× LEUG(-6) + C(2,13) + C(2,14) × LEUIT(-7) + C(2,15) × LEUG(-7)
Coefficients in Table 3 display a bidirectional causality, but the validation of the first
hypothesis must be considered with caution.
Table 3. Block Exogeneity Wald Tests, H1 hypothesis, European Union, logarithm values
VAR Granger Causality/Block Exogeneity Wald Tests
Date: 10/25/20 Time: 16:46
Sample: 1996 2019
Included observations: 17
Dependent variable: LEUIT
Excluded Chi-sq df Prob.
LEUG 33.64697 6 0.0000
All 33.64697 6 0.0000
Dependent variable: LEUG
LEUIT 22.60259 6 0.0009
All 22.60259 6 0.0009
Source: author’s capture, of the VAR Granger Causality/Block Exogeneity Wald Tests in EViews.
2. H1 testing – Romania
2.1. Nominal values
The first step involves measuring the order of integration for each variable, by applying the
two instruments (ADF and PP tests) and establishing the maximum order. The tests were
designed as described in the research methodology section, and the results are presented in
Table 4.
Table 4. Orders of integration, H1 hypothesis, Romania, nominal values
The maximum order of integration is 2. After shaping a VAR model with default
parameters, as indicated by the software (this is not affecting the study, as the next step
involves the modification to the optimum lag), three tests out of five indicate a lag length
of 1,…, 5.
By evaluating the specification tests for VAR(5), the AR roots test has two values above 1.
If a correction to VAR(6) is made, the number of roots greater than 1 increases to three,
while for VAR(7) there are eight such roots that testify against the stability of the model.
As the model is not stable, the research methodology prevents the application of the
subsequent steps of the Granger causality test.

The two tests lead to the same order of integration as in the case of the nominal values, and
the same maximum order of integration applies for the future eventual modified VAR
model.
Table 5. Orders of integration, H1 hypothesis, Romania, logarithm values
Source: author’s representation of EViews results for ADF and PP tests
For Romanian logarithm variables, the initial VAR model is configured for an optimal lag
length of 1,…, 2, this model is stable and the residuals are not serially correlated. As
normality and heteroskedasticity tests display proper values, the model is suitable for
proceeding to the next step.
The updated model is characterized by the following equations:
LRUIT = C(1,1) × LRUIT(-1) + C(1,2) × LRUIT(-2) + C(1,3) × LROG(-1) + C(1,4) ×
LROG(-2) + C(1,5) + C(1,6) × LRUIT(-3) + C(1,7) × LROG(-3) + C(1,8) × LRUIT(-4) +
C(1,9) × LROG(-4)
LROG = C(2,1) × LRUIT(-1) + C(2,2) × LRUIT(-2) + C(2,3) × LROG(-1) + C(2,4) ×

LROG(-2) + C(2,5) + C(2,6) × LRUIT(-3) + C(2,7) × LROG(-3) + C(2,8) × LRUIT(-4) +
C(2,9) × LROG(-4)
No Granger causality can be asserted between the two variables that characterize the first
hypothesis, as the Chi-sq coefficients have values well below the limit corresponding to
the accepted significance level.
3. H2 testing – European Union

3.1. Nominal values
All variables are found to be stationary on their first differences, so the maximum order of
integration is 1 (see Table 6).
Table 6. Orders of integration, H2 hypothesis, European Union, nominal values
Taxes less subsidies on products 1 1
Source: author’s representation of EViews results for ADF and PP tests
For a maximum of 2 lags, the VAR presents autocorrelation issues. When modified for 3
lags, it complies with all tests and thus the model can be updated with another exogenous
lag.
The results of the Granger causality test are presented in Table 7:
Table 7. Block Exogeneity Wald Tests, H2 hypothesis, European Union, nominal values
VAR Granger Causality/Block Exogeneity Wald Tests
Date: 10/25/20 Time: 18:30
Sample: 1996 2019
Dependent variable: EUPT
EUG 5.362831 3 0.1471
All 5.362831 3 0.1471
Dependent variable: EUG
EUPT 8.988649 3 0.0294
All 8.988649 3 0.0294
Source: author’s capture, of the VAR Granger Causality/Block Exogeneity Wald Tests in EViews.
The interpretation of the coefficients and probabilities demonstrates the validation of the
H2 hypothesis for EU nominal values. EUPT Granger causes the Gross Domestic Product
in the case of the European Union.

After interpreting the values of the unit root tests, the situation matches the case of nominal
values. The initial VAR(2) has problems with serial correlation, so update to VAR(3) is made,
the result is a stable model, no autocorrelation, but it does not pass the normality test.
4. H2 testing – Romania
4.1. Nominal values
The orders of integration observed for the two variables are described in the Table 8, and
the maximum order is given by the measure of GDP, being equal to 2.
The optimum lag length is shown to be 1,…, 2. At this setting, the model is not stable, as
is VAR(3). VAR(4) is stable, but has issues of serial correlation, and VAR(5),VAR(6) and
VAR(7) are not stable. Thus, this correlation cannot be tested due to the improper results
of the specification tests.

For the logarithm values that describe the Romanian GDP and Taxes less subsidies on
products, the application of the two tests granted the results on the orders of integration:
The lag length criteria for the VAR model led to an optimum specification of 2 lags. When
tested, the model responds well to the Roots of Characteristic Polynomial test, there is no
serial correlation in the residuals, the null hypothesis of the VAR Residual Normality Tests
is accepted, thus the residuals are normal, and the heteroskedasticity test is also passed.
Therefore, this model respects the validity criteria in order to be used for the Granger
causality test. After the additional lags are added, the Wald test displays that no Granger
causality can be asserted, thus the research hypothesis H2 cannot be validated for Romania.
Conclusions
For the research hypotheses verified against nominal values, the only significant conclusion
reached is the fact that the Taxes less subsidies on products, in the European Union,
Granger causes the Gross Domestic Product. The other models were not suitable for the
actual testing for causality, because of the responses given to the specification tests. The
application of the research methodology pursued all steps in three of the four cases, two
tests indicated no causality, only one asserted bidirectional Granger causality between
Gross Domestic Product and Taxes on production and imports less subsidies but, due to
the error encountered during the application of the last specification test for that model, this
result cannot be accepted as a validation of the hypothesis. Thus, the results achieved by
this paper shows that the influence exerted by the two causal variables on the GDP cannot
be categorized, in all cases, as Granger causality.
Note
(1)
EViews® is a registered trademark of IHS Global Inc.
References
Alesina, A. and Ardagna, S., 2010. Large changes in fiscal policy: taxes versus spending. Tax policy
and the economy, 24(1), pp. 35-68.
Chirilă, V. and Chirilă, C., 2017. The Analysis of Romania’s External Migration and of the Causality
between Remittances and Romania’s Economic Growth, Amfiteatru Economic 19(46),
pp. 696-710.
Giles, D., 2011. Testing for Granger Causality, [online] Available at <https://davegiles.
blogspot.com/2011/04/testing-for-granger-causality.html> [accessed October 17th, 2020].
Hacker, R.S. and Hatemi-J, A., 2003. Tests for Causality Between Integrated Variables Based on
Asymptotic and Bootstrap Distributions, Working Paper [2003:02], Department of Statistics,
Lund University.
Hines Jr., J., and Desai, M.A., 2005. Value-added taxes and international trades: the evidence.
Unpublished manuscript.
Lacey, D., 2000. UK regional gross domestic product (GDP): methodological guide. Economic
Trends, [online] Available at: <https://escoe-website.s3.amazonaws.com/wp-content/
uploads/2020/01/01234148/ET-565-UK-regional-Gross-Domestic-Product-GDP-
methodological-guide-D-Lacey-Dec-2000.pdf> [Accessed 29 October 2020]
Păunică, M., Manole, A., Motofei, C., and Tănase, G.L., 2019. The impact of remittances
on GDP and household consumption. An European Union countries analysis, Economic
Computation & Economic Cybernetics Studies & Research, [online] Available at
<http://www.ecocyb.ase.ro/nr2019_4/6.%20%20Paunica%20Mihai,%20Al%20Manole.pdf>
[Accessed 29 October 2020].
Păunică, M., Manole, A., Motofei, C., and Tănase, G.L., 2018. The Globalization in the actual
Context of the European Union Economy. In: 12th International Conference on Business
Excellence: Innovation and Sustainability in a Turbulent Economic Environment. Bucharest,
Romania, 22-23 March 2018. De Gruyter.
Pessino, C., and Fenochietto, R., 2010. Determining countries' tax effort. Hacienda Pública
Española/Revista de Economía Pública, 195, pp. 65-87.
Spoerer, M., 1998. Taxes on Production and on Imports in Germany, 1901-1913. Jahrbuch fur
Wirtschaftsgeschichte, pp. 161-179.
Stoilova, D., 2017. Tax structure and economic growth: Evidence from the European Union.
Contaduría y Administración, 62(3), pp. 1041-1057.
Stoilova, D., and Patonov, N., 2013. An empirical evidence for the impact of taxation on economy
growth in the European Union. Tourism & Management Studies, pp. 1031-1039.
Tanzi, V., and Zee, H.H., 2000. Tax policy for emerging markets: Developing countries. National
Tax Journal, 53, pp. 299-322.
Tiwari, A. K., and Mutascu, M., 2014. A revisit on the tax burden distribution and GDP growth:
fresh evidence using a consistent nonparametric test for causality for the USA. Empirical
Economics, 46(3), pp. 961-972.
Vasiliauskaitė, A., and Stankevičius, E., 2009. Tax burden management and GDP growth: Case of
EU countries. Ekonomika ir vadyba, 14, pp. 202-209.
Measuring Adobe Company performance

from the perspective of a Complex Adaptive System
Nora CHIRIȚĂ
nora.chirita@csie.ase.ro
Andrei VIȘAN
visancristian2008@gmail.com
Mihaela POPESCU
mihaellap42@gmail.com
Abstract. Strategic management is a key component in any business. It represents the continuous
process of creating, implementing and evaluating decisions that help the organization to be more
productive, by guiding the allocation of resources to achieve goals. Depending on the strategy that
a company wants, which aims to create a competitive advantage, several techniques are used.
Making a correct diagnosis using the optimal ways to achieve this is a key element in any analysis.
Viewed as a cybernetic system, the company must be analyzed from the point of view of the efficiency
and effectiveness of internal processes, as well as from the perspective of analyzing the factors in
the environment in which it operates. In this research, the diagnosis was made by Adobe using
SWOT analysis, one of the most used analysis techniques, but correctly interpreted and performed
is still one of the most accurate techniques. The identification of key performance indicators and
their analysis is another topic that we address in this article.
Keywords: SWOT analysis, key performance indicators, BCG matrix.
JEL Classification: A10, G12, G17, L25, M2.

56 Nora Chiriță, Andrei Vișan, Mihaela Popescu
1. Introduction
Living in an age of digitization and the large volume of data and information, in which the
environment is constantly changing, and given the explosive development of society
associated with the limited nature of resources, it becomes increasingly difficult to adapt
and co-evolve units economic. The theory of complex adaptive systems, through the
properties they offer, is an important aspect in deepening and understanding the functioning
of economic units as complex adaptive systems. Over time, there have been several
researchers in the field who have contributed to the development of the literature, offering
numerous definitions and presenting the properties of CAS adapted to companies, banks,
financial markets, national economic systems or even the global economy.
The analysis and diagnosis of economic systems focuses and capitalizes on a series of
knowledge in various fields such as: operational research, decision theory, management,
finance, applied informatics, statistics, mathematics. In a broad sense, the analysis and
diagnosis of economic systems represents the totality of approaches to systems, exact
scientific methods and heuristic methods of investigation, diagnosis, modeling, design and
solution on this basis of real-world problems, focused on the systemic approach, in an
environment characterized by risk and uncertainty, in order to improve the quality of
decisions for the efficient management of economic systems.
When a diagnosis of an economic system is made, it must be analyzed from the point of
view of the component parts, but also from the cybernetic perspective in the sense of
analyzing the environment in which the economic system carries out its activities. Aristotle
gave the first definition of the notion of system, by stating “the whole is more than the sum
of its components”. A system is a set of elements that interact both with each other and
with the environment of which the system is part, with respect for rules, laws and principles,
in order to achieve a meaning, objective, purpose
In this research, SWOT analysis was presented as the usual technique for analyzing an
economic system by identifying strengths, capitalizing on opportunities and reducing or
eliminating weaknesses and threats. When it comes to launching a product, for example, it
is important that the business unit takes competitiveness into account and that strategic
decisions are made to promote a competitive advantage. For this, the importance of the
Boston Consulting Group matrix was presented. The last part of the research analyzes the
Adobe Company in terms of key performance indicators that correctly identified and
evaluated can contribute to the stability and increase of the company's performance. Before
any analysis it is necessary to know the organization in this way, for this purpose the SWOT
analysis was performed for the analyzed company.
2. Complex Adaptive System in enterprise

The enterprise is an obvious example of a complex adaptive system. This is due to the fact
that they have several properties, the most important being the following:
Connectivity: The ways in which the agents of an enterprise are interdependent and relate
to each other are of major importance for the survival and prosperity of the enterprise, as
Measuring Adobe Company performance from the perspective of a Complex Adaptive System 57
they lead to certain regulatory processes necessary for the existence of the enterprise.
Relationships between agents have a higher significance than the agents themselves.
Self-organization: There is no central command and control entity in a complex adaptive
system. However, there is an ongoing process of reorganization given the discovery of the
best model in relation to the environment.
Coevolution: All businesses exist and are an integral part of their environment. As a
consequence, with the change of the environment, companies also change so as to ensure
the best possible fit with it. However, when they change, so does the environment, so
change happens all the time.
Necessary variety: The greater the variety of an enterprise, the stronger it is. After all,
uncertainty and obscurity are high in companies, which use contradictions to co-evolve
with other systems in the environment.
Hierarchical systems: Most enterprises are structured and represent subsystems of larger
systems. For example, a bank branch is a system that has a management, employees and
customers. At a higher hierarchical level, it belongs to the network of bank branches in a
certain country, which is part of an international banking group.
Emergency: Planned and controlled interactions take place between the agents of an
enterprise, but rather randomly. As a result of these interactions, behaviors that are
characteristic of the company's agents emerge and shape the company's behavior as a
whole.
Simple rules: It can be said that companies do not have a high degree of complexity.
Emerging models can take many forms, but the spectrum of rules that govern the enterprise
is relatively simple. For example, how a product changes its market price is a result of
maintaining a balance between supply and demand.
Sensitivity to initial conditions: Small changes applied to the initial conditions in an
enterprise can have significant consequences after going through an emerging process such
as a feedback mechanism. Certain shocks in a financial network can lead to the bankruptcy
of a commercial bank and could have a domino effect on the international economy,
triggering a new economic crisis.
Suboptimality: Perfection is not necessary for an enterprise to thrive in its environment. It
just needs to be a little better than its competitors and any extra effort in the initiative to
become the best leads to a waste of energy.
Operation at the limit of chaos: Chaos has a special place in the theory of complexity
because any system with the capacity for self-organization operates far from balance, i.e.
at the limit of chaos. A system that is in equilibrium does not have its own dynamics that
give the system the ability to respond appropriately to its environment and will gradually
dissolve. A chaotic system actually loses its system characteristic. The state in which the
system manifests its highest productivity is on the verge of chaos, when creativity and
variety make their presence felt at the highest levels, leading the company to new
opportunities.
3. Analysis and diagnostic techniques applied in an enterprise

Strategic management allows a company to be more proactive in shaping its own future,
helping to generate competitive advantages by increasing the likelihood of choosing the
best strategy.
3.1. Analysis and diagnostic techniques applied in an enterprise
An important stage in the strategic management process is the external and internal analysis
called SWOT analysis. Through the external one, the company identifies the threats and
critical opportunities in its competitive environment. In addition, it examines how
competition in the same environment evolves and analyzes the future implications for the
threats and opportunities that economic unity may face. Internal analysis helps to identify
the strengths and weaknesses of the organization. This helps the company to understand
what its resources and characteristics can be an advantage over the competition and which
of them would rather be disadvantages. With the help of the SWOT matrix, companies can
resort to the appropriate strategies. Strategic choices can be associated with the mission,
objectives, vision, and internal and external analysis of the organization. It is important that
after the elections have been made, these strategies are implemented. According to Barney
and Hesterly (2006), the implementation of strategies occurs when an organization adopts
organizational practices and policies in accordance with its strategies. The completion of
the process described by the strategic management is rendered by gaining advantages over
competitors.
SWOT analysis is a tool used for strategic planning and strategic management in
organizations. Used efficiently, it is a rigorous technique for building organizational and
competitive strategy at the company level.
SWOT analysis is a process that involves four components (strengths, weaknesses,
opportunities and threats) and two dimensions (internal and external environment).
Weaknesses and strengths are attributed to internal factors, the organization, while
opportunities and threats are attributed to external factors attributed to the environment.
The SWOT analysis, as a rule, is drawn in a box with four dials that allows a summary that
is organized according to the four sections. In this case, the company manager must
contribute to balancing the strengths and weaknesses of the organization with the
opportunities and threats from the external environment, so as to arrive at the selection of
the successful strategy.
The strengths of the company: they are highlighted by the characteristics that add value to
the analyzed company compared to others. Thus, the strengths represent a positive,
favorable and creative component of a company. They play an active role in achieving
organizational goals because, through their analysis, the company knows and understands
its business potential. In addition, companies must respond to external environmental
threats by using their strengths.
Weaknesses of the company: in this case, the issues identified here mean that there is
something more disadvantageous compared to competition, i.e. the situation where the
current system and the company's capacity is weaker compared to others, the weakness of
the organization being a negative feature. Thompson and Strickland (1989) state that a
weakness is something that one organization lacks compared to others or a condition that
puts it at a disadvantage.
It is just as important for any organization to know its strengths as well as its weaknesses.
No strategy can be built on weaknesses, which have the potential to lead the organization
to inefficiency and inefficiency. Therefore, they need to be known and improved.
Environmental opportunities: means identifying a condition or situation that is considered
appropriate for a business activity. Opportunity is an advantage, a positive feature,
favorable to organizations. For organizational management, an opportunity is the
convenient time or situation that the environment presents to the organization to achieve
its goals.
Environmental threats: the threat is a situation or condition that endangers the development
of an activity, being a disadvantageous situation for the organization. For this reason, it is
a negative feature that should be avoided. Environmental factors that can impede
organizational efficiency and effectiveness are threats. The new world order formed as a
result of globalization involves both opportunities and threats. This system that improves
opportunities as well as threats guides organizational management to be more careful and
act more strategically in terms of developments inside and outside their environment.
External opportunities and external threats refer to economic, social, cultural, demographic,
environmental, political, legal, governmental, technological and competitive trends and
events that could bring significant benefits or harms to an organization in the future.
Strengths and internal weaknesses are the controllable activities of an organization.
SWOT analysis is a valuable technique for planning and decision making, being used in
the last fifty years in the field of strategic management. In this process, a number of analysis
techniques are used to achieve the long-term goals of an organization.
3.2. Boston Consulting Group Matrix
Today, sustainable business improvement faces various challenges for global economic
competition. Business models focus on providing products and services that are profitable
today, but also try to identify changes in offers that will keep the company profitable in the
future. A useful tool in strategic management, which has been used to analyze the
environment of an economic unit and to suggest more resource allocation strategies based
on the growth rate of the industry and the market force of the organization is the Boston
Consulting Group matrix. Also known as the growth quota matrix, the BCG matrix
provides a framework for analyzing products based on growth and market share. It is used
to help companies obtain information about products that would best help capitalize on
market share opportunities.
Boston Consulting Group helps companies increase their efficiency in terms of successful
business activities. To this end, the matrix plays an important role because it is a useful tool
for strategic product performance planning at industry and company level. The matrix was
introduced in 1970 by Bruce Doolin Henderson, president and CEO of Boston Consulting
Group, and has since been used by many corporations to improve their ability to run their
business efficiently and profitably.
To help companies analyze their assets in the medium and long term, the BCG matrix
segments products into four classes/quadrants:
Figure 1. Example of Boston Consulting Group Matrix
Source: The design is done by the authors.
Dilemmas or Question Marks: Indicates products with high market growth and low market
share.
The Stars quadrant shows that both the growth of the market and the market share reach
high values.
The Cash Cows dial is formed when market growth is low, but the market share remains
high.
The Dogs/Millstones quadrant shows that the products in this class have a small market
share and an almost insignificant market growth.
According to Temmerman (2011), the BCG matrix provides a simple two-dimensional
analysis of strategic business units such as industry growth rate and relative market share.
The vertical axis represents the growth rate of the industry, and the relative share on the
horizontal axis.
For a concrete explanation of the matrix, we will represent in the figure below five products
in the quadrants of the matrix, according to the market shares and the related growth rates.
To build the matrix it is necessary to know several notions such as market growth, relative
market share and market share.
In the figure above you can see that the products were divided into the following categories:
Dilemma (Product 4), Star (Product 3), Cash Cows (Product 2) and Dogs/Millstones
(Product 1 and Product 5).
The analyzes will be based on the following premises:
 Market share can be obtained through investments in the marketing sector.
 Market share gains will always create cash surpluses.
 Excess cash is generated when the product is in the maturity stage of the life cycle.
 The best opportunity to build a dominant position in the market is in the growth phase.
Dilemmas or question marks form the quadrant that collects business units that have a small
market share in a high-growth market. These units do not try to manage a lot of cash in
their industry and the products in this class, such as product 4, are called question marks or
dilemmas because the organization has to decide whether to invest in them using a rigorous
strategy such as market access, market development or product development, or eliminate
them because these products can either be turned into star products or move into the
millstone dial. Although the product 4, according to the matrix, has a high demand, the
realized profit is small due to the reduced market share. There is no clear strategy for this
business. In reality, most businesses start out in question because companies are trying to
enter a high-growth market with no market share. If dilemma products are ignored, they
become millstones but, on the other hand, have the potential to become stars and eventually
dairy cows when market growth slows.
The products in the star quadrant are indicated by obtaining a large market share in a fast-
growing market. According to the literature, these are considered the best opportunities for
the growth and benefits of the company. The stars are business leaders, but they still need
a lot of support. In this regard, they create large sums of cash to support a strong market
share. In our example, product 3 is a star product. With this product, the organization makes
real money. When the market share becomes very high, the industry matures and the market
growth rate decreases and the star product turns into a cash cow.
Cash cows are the quadrant of products that have a large market share in a mature period
of a slow-growing industry. This dial is called cash or milking cows because it produces
excess cash over the needs of these products. They require very little investment and create
significant cash to grow other businesses. According to the above matrix, product 2 is a
cash cow product. Its development is an attractive strategy to gain a competitive advantage.
Even capital reinvestment is necessary to maintain the current market share. Many products
in this dial have been stars in the past. Although cash cows are less attractive in terms of
growth, they are valuable in an enterprise.
The class of millstones or dogs are business procedures that have weak growing market
shares and cannot generate a large amount of cash due to weak business strategies. Both
product 1 and product 5 are in this dial. They face disadvantages caused by the costs of a
small market, their poor quality or inefficient marketing. According to the literature, these
products are often liquidated or reduced. To avoid this, these products need to start
distributing cash.
The BCG matrix needs a systematic classification rule, being an exploratory analysis tool
based on interactions, so as to obtain the consensus between different managers and the
analysis of the personalized classification scenario. The matrix provides recommendations
for the development of strategic resource allocation and strategy development in the typical
multi-business company. It is argued that relative competitive position and growth are the
two fundamental parameters that must be taken into account when determining the strategy
that an individual business must follow when viewed in the context of the company's
overall business portfolio.
4. Case study: Diagnosis and analysis of key performance indicators in Adobe

Adobe Inc. is an international software company that aims to provide the tools and solutions
needed to create, design and deliver the best digital experience for its customers. The
analyzed company represents one of the main players in this industry with a turnover of
over 11 billion USD (2019) and over 20,000 employees globally (2020).
Adobe Inc. (formerly Adobe Systems Inc.) was founded in 1982 by John Warnock and
Charles Geschke in Mountain View, California. The founders decided to start this company
to bring the PostScript language to market because their previous employer, Xerox, did not
want to do so. Steve Jobs wanted to buy Adobe, but the founders turned down the option.
However, Jobs managed to buy 19% of the company's shares at a price 5 times the actual
valuation and a 5-year, pre-paid PostScript license. Thus, Adobe became the first company
in the Silicon Valley area to make a profit in its first year of operation. PostScript was later
used in laser printing and dominated the market in the publishing industry. The company
was listed on the US NASDAQ in 1986 and launched the PDF document format in 1993.
In 1996, the company moved its headquarters to San Jose, California. Over the years,
Adobe has distributed creative software either developed in-house or taken over through
the acquisition of certain companies. Many of these software products are famous, such as
Acrobat Reader, Photoshop and Creative Cloud. The company currently has offices in more
than 25 countries, more than 20,000 employees globally and a turnover of more than $ 11
billion in 2019 alone.
4.1. Product history
The first products that Adobe hit were the PostScript language and the PDF format based
on that language. The company provided PostScript licenses to other printing companies,
and soon after became one of the main languages used for printing in the 1990s. The PDF
file format was created to store all the information needed to display any document on any
type of personal computer. Adobe later turned PDF into a free standard that was adopted
internationally in 2008 and is now the basic format for all shareable documents.
Adobe began launching creative software in the late 1980s with its main creative product,
Photoshop. This product was intended for use in graphics and image manipulation and soon
dominated the market after its launch in 1989. Other popular products were launched in the
following years: Premiere in 1991, Acrobat Reader in 1993, After Effects in 1994 and
Dreamweaver in 2005. It is noteworthy that some of Adobe's products have been taken
over from the acquisition of its competitors. In 2003, Adobe released Creative Suite 1
(CS1), a suite of creative activity applications bundled together to pay the one-time license.
Adobe continued to sell software suites until the CS6 version in April 2012. This business
model allowed Adobe to sell not only software licenses, but also software suites.
A radical transformation was made by Adobe by launching Adobe Creative Cloud in 2012
to move from a license-based business model to a monthly subscription. The paradigm shift
was abrupt as subscription-based models were relatively unknown to the market at the time.
In May 2013, Creative Cloud was announced as the main access platform for its creative
products and that individual products and suites will no longer be sold. Nearly 3 years after
its launch, Creative Cloud accounted for 75% of the company's revenue, valued at
approximately $ 3 billion, and in 2019 sales were $ 7.2 billion, or approximately 63% of
revenue for this year.
With the advent of new technologies such as artificial intelligence (AI), virtual reality (VR)
and the Internet of Things (IoT), Adobe has improved its existing products to better serve
its customers in these areas. Updates have recently been released for Premiere and After
Effects programs that use artificial intelligence to match colors and audio volume.
Improvements have also been added for creating virtual reality applications. Adobe has
also implemented a software program that allows companies to track user activity across
multiple types of devices, giving them a personalized experience. Until then, this aspect
was not known, and companies continued to show users the same ads for a long time.
Creative Cloud is Adobe's software creation and design suite that is based on a monthly
subscription. It is intended for those who have these activities as hobbies, but also for
industry professionals. The purpose of this product family is to be the best software for
photo, video, graphics and content creation.
Experience Cloud is the software suite used for marketing and advertising. It is aimed at
companies that want better tools for customer interaction by providing solutions for
advertising, analysis, content, data management and marketing campaigns.
Document Cloud is a set of tools focused on working with PDF documents. This suite is
aimed at businesses, but certain features are also available to the general public. Its purpose
is to better interact and manage PDF documents through features such as editing,
commenting and signing them.
4.2. Market position
Adobe operates in three sectors: Digital Media, Digital Experience and Publishing. The
first two of these are part of the company's long-term growth and represent the areas in
which it aims to attract investment and create solutions for the market.
The Digital Media sector is Adobe's most profitable and active area, providing solutions
and tools that enable individual consumers, but also small and medium-sized businesses to
create, publish, promote and monetize their content in digital form. It is the segment in
which the Creative Cloud and Document Cloud products are found, with revenues of 7.71
billion USD in 2019, representing 69% of this year's turnover.
The Digital Experience industry is Adobe's second line of business. Here is the Experience
Cloud product. The services and solutions in this segment aim at creating, managing,
executing, measuring and optimizing digital advertising and marketing activities. This
sector recorded revenues of 3.21 billion USD in 2019, i.e. 28.7% of turnover.
The publishing industry is Adobe's smallest market segment. It addresses the opportunities
that have arisen in connection with the authorization and publication needs related to the
technical and business aspects of companies. For 2019, revenues of USD 257 million were
recorded here, representing 2.3% of total revenues.
Figure 2. Share of revenues of each sector in turnover
Source: Refinitiv Eikon Thomson Reuters.
4.3. SWOT Analysis

Strength points:
 Global presence: Adobe has a diverse global presence in countries in Europe, Asia,
North America and South America. Its distribution channels are represented by retailers,
independent software vendors and original equipment manufacturers.
 Successful transition to the subscription-based business model: Adobe started with a
renewable license-based business model and then introduced the subscription-based
model. The company introduced the Adobe Creative Cloud platform, replacing Adobe
Creative Suite. This transition has led to an increase in recurring annual income, helping
to maintain the stability of its main source of income.
 Large portfolio: Adobe provides a wide range of products and services in the digital
media and digital marketing segments. They are available to customers such as
businesses, freelancers, marketers or individual consumers.
 Perfect in design: No other graphic design software on the market compares to Adobe
Photoshop. Adobe also provides specialized products for different niches, such as the
gaming, video or media industries, or web development.
 Trustworthy: Adobe products have proven to be reliable, and the brand has proven that
it can get the job done. Adobe has been on the market for decades and has slowly but
surely gained significant market share for design software products.
Weaknesses:
 Significant debts: At the end of fiscal year 2019, Adobe had debts of $ 4.1 billion. These
debts require a large part of the capital flow and limit the flexibility in terms of planning
and responding to changes in the market.
 General feedback: Over time, users have complained that Adobe does not take their
requirements into account when launching new software products.
 Relatively expensive subscriptions: Adobe creative and design software packages may be
considered too expensive by some customers, making them less attractive to them. It
applies in particular to emerging markets such as Eastern Europe, Asia and South America.
 Weak presence on the video animation market: Adobe has a very weak presence in the
3D video animation market, which is dominated by the products of its competitors 3ds
Max and Sony Vegas. In this area, Adobe operates on the market with Adobe Premiere
Pro, but is not as specialized in the development of three-dimensional structures such as
3ds Max or Sony Vegas.
Opportunities:
 Promising prospects for cloud computing: The demand for cloud computing services is
growing compared to previous years. This is influenced by capital and reduced
operating costs, resulting from a subscription-based model that contains only the
services requested by the customer.
 Growth in the digital marketing sector: Expenditure on digital marketing has increased
globally due to the widespread adoption of mobile phones and internet access. Adobe is
looking to increase its revenue from cloud marketing solutions.
 Partnership with Microsoft: Adobe and Microsoft have formed a strategic partnership
to help each other in their digital transformation efforts. Thus, Microsoft Azure became
Adobe's main cloud platform, and Adobe Marketing Cloud, Adobe Document Cloud
and Adobe Creative Cloud were chosen as the main creation and design tools by
Microsoft. Thus, Adobe will have a richer customer base by increasing the sales of its
top products.
Threats:
 Piracy: Adobe's main threat is piracy of its software products due to torrent sites and
illegal downloads of older versions available on the Internet. In emerging markets
(Eastern Europe, Asia, South America), piracy in particular is reaching astronomical
levels.
 Free software: There are several free design software products available on the market.
GIMP is a good alternative to Photoshop for amateur designers. There are also free
alternatives to Dreamweaver or other Adobe products. Free software products are not
generally used by professionals for commercial purposes, but they capture a large part
of the market share.
 Competition: Adobe products and services are in intense global competition. Innovative
technologies have allowed other participants to perform in the market, increasing
competition. This can lead to price competition, which has a negative impact on Adobe's
profit.
 Security threats: Adobe has confidential information about its customers, which has
always been a target for hackers. Computer attacks also constantly damage the system.
These potential security breaches will need to be taken seriously by Adobe to keep the
brand image on a positive note.
4.4. Analysis of key performance indicators of Adobe
To perform this analysis, we used Adobe share price data from January 4, 2010 to May 28,
2020, and based on these we determined the daily returns, average returns, risk and Sharpe
rate. These indicators will be detailed and interpreted below.
Figure 3. Probability density of Adobe stock returns
Source: Processing authors in Crystal Ball.
The daily returns of the Adobe stock are best represented as a Student t distribution. This
classification was based on the Anderson-Darling statistical index, in this case a value of
0.8331, i.e. the Student t distribution is true for specific data daily returns. Moreover, from
the analysis of the histogram, the asymmetry coefficient is 0.07, so we have an almost
perfectly symmetrical distribution, and the flattening coefficient is 16.02, i.e. the
distribution is leptokurtic, so that, in general, recorded small daily variations, with increases
and decreases in price around 1%.
Figure 4. The evolution of the daily profitability of Adobe's stock
Source: Processing authors in R Studio.
Thus, based on the correlation and the Augmented Dickey-Fuller stationarity test, found in
Figure A.1. (Appendix A), we can conclude that the profitability of the Adobe share is
presented in the form of a stationary time series. At the same time, the most significant
price movements were in the following data:
 September 22, 2010
An Adobe board conference call was held on September 21, highlighting the weak sales of
the Creative Suite 5 software package and the declining adoption of Adobe products in
Japanese education, a very important segment for Adobe. This event had the effect of
lowering the share price by 19% the next day.
 March 13, 2020
The share price of Adobe was forecast to rise to $ 375 next month by analysts at JP Morgan
Chase, which resulted in a positive outlook for investors, which led to a 17% increase in
the share price the next day.
 March 16, 2020
An analysis conducted by Wells Fargo & Co. estimated that it expects a drop from $ 335
to $ 315, but this had an even greater effect as investors decided to sell the shares in even
greater numbers, resulting in a price of $ 286 and a 15% decrease for the next day.
Figure 5. The evolution of the annual profitability of the Adobe share
Figure 6. The evolution of the annual risk of Adobe's action
Figure 7. Evolution of the annual Sharpe rate of Adobe stock
Source for Figures 5, 6 and 7: Processing authors in R Studio.

The previous graphs show the evolution of the average annual returns, the annual risk and
the annual Sharpe rate from 2010 to 2019 specific to Adobe. There is a close correlation
between the evolution of the average annual profitability and the annual Sharpe rate. We
also see some increases in average profitability in 2013 (due to the change in the
subscription-based business model, replacing the model based on annual licenses) and 2017
(experiencing one of the largest annual increases in revenue, 25% as following the
explosion of the digital media sector, on which the company relies most). Regarding the
company's risk (volatility), until 2019 it remained at a stable rate of approximately 1%. In
addition, the year 2020 started in full force, but with the start of March, the actions started
to decrease dramatically as a result of the crisis caused by the COVID-19 virus in March
and April. However, things started to return to normal in May, when policies were adopted
that involve employees being able to work remotely for an indefinite period, thus
facilitating the development and maintenance of Adobe solutions and further ensuring a
pleasant customer experience.
We also made some predictions regarding the price of the Adobe stock and the indicators
presented. The stock price forecast was made on the basis of data from 1 January 2010 to
29 May 2020 for 30 May 2020 to 30 May 2021, using the simple exponential smoothing
method, and the limits were set, based on a confidence interval of 95%. Thus, given the
following figure, the price of the Adobe stock would have an upward trend, reaching $ 487
at the end of the period, with a range of $ 401.88 - $ 572.14 (Figure 8).
Figure 8. Forecast of Adobe stock price evolution
The forecast for the average return on Adobe shares in 2021 is estimated at 0.0021, with a
probability of 95%, a stable value in relation to historical data, the confidence interval being
0.0075-0.0035 (Figure 9).
Figure 9. Forecast of the evolution of the average annual profitability of Adobe's stock
The risk for 2021 was estimated at 0.047, also with a probability of 95%, having as limits
the range 0.022-0.052. Thus, it can be said that the risk is generally expected to be lower
than in 2020, which may be an incentive to buy Adobe shares (Figure 10).
Figure 10. Forecast of the evolution of the annual risk of Adobe's action
Regarding the Sharpe rate, it was estimated at 9% for 2021, with a probability of 95%,
falling between the limits -2% - 20%. Considering the above about the average annual
return and the annual risk, the Sharpe rate is indeed an optimistic forecast for a potential
buyer (Figure 11).
Figure 11. Forecast of the evolution of the annual Sharpe rate of Adobe stock
5. Conclusions
Performance analysis is one of the main concerns of a company's management, because
only by analyzing the activity and identifying problems in the processes can maximize
profit. Thus, it is necessary to select the most appropriate methods according to the
company's objectives.
This research illustrated the purpose of performance management which has the role of
creating and developing an organizational structure capable of achieving it, stimulating
consistent decisions and ensuring the convergence of different objectives of the company
to established performance indicators.
When analyzing the performance of a company, you must take into account three basic
notions: economy (procuring the necessary resources at the lowest cost), efficiency
(maximizing the results obtained, starting from a given amount of resources, or minimizing
the amount of resources for a predetermined result) and effectiveness (the results obtained
to achieve the expected results).
Living in an age of digitalization in which the world is changing and adapting to the
complex systems of society, businesses must always be prepared with the tools necessary
for adaptation.
The need to identify useful tools that would help improve or increase performance is
relevant and of great interest to all organizations. Experiencing a world in which we are
surrounded by a multitude of data, the need to maintain the strategic direction set by a
company, to reduce the uncertainty of making the right decisions, exists.
KPIs should be understood and used there and when needed and help to complete the
performance management framework. In this research, several key performance indicators
were chosen to create a first view of Adobe's performance management.
The second part of the case study completes the first part of it, building a complex research
to make a diagnosis of Adobe. The analysis and diagnostic techniques highlighted the very
favorable ecosystem for the company, especially from a social and economic point of view,
which strengthens Adobe's position as an important player in the digital and cloud
technologies market. According to the analysis of key indicators, the company's profitability
is also on an upward trend and is expected to follow this trend in the next two years.
References
Aguinis, H., 2013. Performance Management. Third ed. Pearson.

Ajitabh, A. and Momaya, K., 2003. Competitiveness of Firms: Review of Theory, Frameworks and
Models. Singapore Management Review.
Barney, J.B. and Hesterly, W.S., 2006. Strategic Management and Competitive Advantage, USA:
Prentice Hall.
Bhatti M., Awan, H. and Razaq, Z., 2014. The key performance indicators (KPIs) and their impact
on overall organizational performance. Quality & Quantity: International Journal of
Methodology, Springer.
Bianchi, C., 2016. Dynamic Performance Management. Springer.
Brigham, E.F. and Ehrhardt, M.C., 2005. Financial Management Theory and Practice. Ohio:
Donnelley Willard.
Cojanu, V. and Bilbor, M.R., 2007. The SWOT Technique in Action: Strategic analysis of
Development in Romania. Review of Management and Economical Engineering, 6(5),
pp. 162-167.
Collins, D.J. and Montgomery, C.A., 2005. Corporate Strategy: A Resource-Based Approach. New
York: McGraw-Hill/Irwin.
Dimon, R., 2013. Enterprise Performance Management Done Right. John Wiley &Sons Inc.
Gurel, E., 2017. SWOT analysis: A theoretical review. Journal of International Social Research
10(51):994-1006, DOI: 10.17719/jisr.2017.1832.
Hoffman, K.D., Czinkota, M.R., Dickson, P.R., Dunne, P., Griffin, A., Hutt, M.D., Krishnan, B.C.,
Lusch, R.F., Ronkainen, L.A., Rosernbloom, B., Sheth, J.N., Shimp, T.A., Siguaw, J.A.,
Simpson, P.M., Speh, T.W. and Urbany, J.E., 2005. Marketing Principles and Best Practices.
Australia: Thomson, South-Western.
Joubert, J.C.N., Jooste, A. and Lotriet, R., 2011. The Cash Cows, Dogs, Stars and Problem Children
of the South African Agricultural Sector. The 18th International Farm Management Congress
Methven, Canterbury. New Zealand.
Nica, I. and Chiriță, N., 2019. Cibernetica firmei. Aplicații și studii de caz. Economica.
Paladino, B., 2013. Corporate Performance Management Best Practices. John Wiley &Sons, Inc.;
Parmenter, D., 2015. Key Performance Indicators. Developing, Implementing, and Using Winning
KPIs. Third ed. John Wiley & Sons, Inc.
Rashid, C., 2018. Efficiency of Financial Ratios Analysis for Evaluating Companies’ Liquidity.
International Journal of Social Sciences & Educational Studies.
Scarlat, E. and Chiriță, N., 2019. Cibernetica sistemelor economice, Third Edition, Economica.
Taticchi, P., 2010. Business Performance Measurement and Management. Springer.
Temmerman, R., 2011. Stars, Problem Children, Dogs and Cash Cows: Evocative Terminology.
In Multilingual Business Communication, SYNAPS. Journal of Professional Communication,
26(2), pp. 48-61.
Wheelen, T.L. and Hunger, D.J., 2002. Strategic Management and Business Policy, 8th Edition.
USA: Prentice Hall.
Appendix A
Figure A.1. Time series stationary analysis using correlation and Augmented Dickey-Fuller test
Date: 06/15/20 Time: 23:04
Sample: 1/04/2010 5/28/2020
Autocorrelation Partial Correlation AC PAC Q-Stat Prob Null Hypothesis: RENTABILITATE has a unit root
Exogenous: Constant
1 -0.144 -0.144 54.349 0.000
2 0.022 0.002 55.662 0.000
Lag Length: 0 (Automatic - based on SIC, maxlag=27)
3 -0.004 -0.000 55.695 0.000
4 -0.033 -0.034 58.515 0.000 t-Statistic Prob.*
5 -0.007 -0.017 58.642 0.000
6 -0.049 -0.053 65.075 0.000 Augmented Dickey-Fuller test statistic -59.11574 0.0001
7 0.043 0.029 69.883 0.000 Test critical values: 1% level -3.432656
8 -0.063 -0.055 80.405 0.000 5% level -2.862445
9 0.047 0.029 86.225 0.000
10% level -2.567296
10 -0.016 -0.008 86.914 0.000
11 -0.008 -0.012 87.098 0.000
12 0.032 0.025 89.829 0.000 *MacKinnon (1996) one-sided p-values.
13 -0.069 -0.059 102.45 0.000
14 0.012 -0.012 102.86 0.000
15 -0.039 -0.033 106.80 0.000 Augmented Dickey-Fuller Test Equation
16 0.064 0.051 117.63 0.000 Dependent Variable: D(RENTABILITATE)
17 -0.020 -0.004 118.70 0.000
Method: Least Squares
18 -0.004 -0.010 118.74 0.000
19 0.001 -0.011 118.74 0.000 Date: 06/15/20 Time: 23:01
20 -0.037 -0.030 122.31 0.000 Sample (adjusted): 1/05/2010 5/28/2020
21 0.032 0.014 125.06 0.000 Included observations: 2617 after adjustments
22 -0.051 -0.035 131.84 0.000
23 0.037 0.017 135.44 0.000 Variable Coefficient Std. Error t-Statistic Prob.
24 -0.001 0.011 135.44 0.000
25 0.007 0.006 135.57 0.000
RENTABILITATE(-1) -1.144020 0.019352 -59.11574 0.0000
26 -0.016 -0.023 136.27 0.000
27 0.003 0.002 136.29 0.000
C 0.001223 0.000367 3.329992 0.0009
28 0.021 0.009 137.42 0.000
29 -0.030 -0.010 139.75 0.000 R-squared 0.571990 Mean dependent var 1.53E-06
30 0.019 0.004 140.71 0.000 Adjusted R-squared 0.571826 S.D. dependent var 0.028659
31 -0.054 -0.046 148.50 0.000 S.E. of regression 0.018753 Akaike info criterion -5.114149
32 0.031 0.013 151.07 0.000 Sum squared resid 0.919642 Schwarz criterion -5.109663
33 0.006 0.009 151.17 0.000 Log likelihood 6693.864 Hannan-Quinn criter. -5.112524
34 -0.016 -0.011 151.86 0.000
F-statistic 3494.670 Durbin-Watson stat 1.999362
35 0.009 -0.006 152.05 0.000
36 -0.006 0.003 152.16 0.000 Prob(F-statistic) 0.000000
Source: Authors computation.

Welfare enhancing uncertainty
Sravaitri CHAUDHURI
Economics Department, Calcutta University, Kolkata, India
sravaitri@gmail.com
Abstract. In this paper it is shown that one of the intrinsic characteristics of the canonical Melitz
(2003) type framework is that inclusion of further uncertainties enhances the aggregate welfare of
the given economy, though reducing the total varieties available. Here the uncertainty assumed is,
payment risk associated with the exchange in international market. The key to such an occurrence
is the fact that uncertainty acts like an increased trade costs leading to added exit of low productive
firms. Since uncertainty seems to be desirable any efforts to decrease such can be deemed expensive.
However, this also proves for a fact that the developing countries with overall lower productivity
and higher uncertainty will lose out its import and export share to international competition in the
global market.
Keywords: firm heterogeneity, payment uncertainty, welfare.
JEL Classification: F10, F11, F12.

74 Sravaitri Chaudhuri
1. Introduction
International trade relations are by and large unpredictable because of constantly
fluctuating exchange rates, governmental policies, interests, prices of factor inputs, nature
etc. at the macro level, disturbing the decisions of the importers and the exporters at the
micro level. Literature that addresses this – both at empirical and theoretical levels – are
limited, but expanding. Export uncertainty is directly as well as indirectly implied in some
empirical research. For example, Besedes and Prusa (2006) using US import level data
observe that median duration of survival for an importing firm is very small, however if it
survives in the short run then its probability of continuing exporting for a long period to
US increases. Iacovone and Javorcik (2010) using Mexican export data after NAFTA,
found that not only a few varieties are able to survive the international market for a long
period of time, but the firms begin by selling only a very small amount to the importers.
The low survival rate of trade relationships does indicate the presence of international
market risk. Furthermore, using trade data of Dutch firms between 2002 and 2008 Cerusen
and Lejour (2011) show that not only, just 5% of all Dutch exporters have presently started
to export but also the same percentage cease to export with high productive firms more
inclined to export and less likely to leave so, among other findings. Bekes et al. (2017)
using monthly French customs data find that firms export less in markets with higher
demand volatility. Crowley et al. (2018) directly indicate that Chinese firms when
subjected to tariff policy uncertainty are less likely to explore new foreign market and also
exist from the established foreign markets also if there was no uncertainty Chinese firm
entry would have been 2% higher.
From a theoretical perspective, Segura-Cayuela and Vilarrubia (2008) introduced
informational externality to counter cost uncertainty using the Melitz (2003) model. In their
model, once a firm enters the foreign market, its success/failure reveals information to other
domestic firms who decide whether or not to enter the market based on the new
information. In this model of learning, it is shown that a sufficient degree of informational
externality can overcome the problems that arise from cost uncertainty. Another theoretical
model by Handely (2014) considers a ubiquitous but often ignored source of uncertainty,
i.e., trade policy uncertainty, in a dynamic heterogeneous firm model, and he observed that
risks reduced entry in new destinations and binding agreements did increase entry. Handely
(2014) also used Australian product level import data to prove the theoretical observations.
Also Carballo, Handley, Limao (2018) using a dynamic heterogeneous firm model show
that uncertainty about foreign income, trade protection and their interaction reduce export
investment and that it can be eased by the presence of trade agreement and by using US
firm level data from 2003-2011 show that uncertainty was the primary reason for trade
collapse in 2008 crisis. Now Rauch and Watson (2003) considered buyer’s side of the
ambiguity in a theoretical framework, where a developed countries buyer search for a new
Less Developed Country (LDC) seller, and its propensity to start small increases with the
cost of search but decreases as the probability of the LDC suppliers ability to supply large
increases with training.
For this study, the generic Melitz (2003) type framework is extended by incorporating
uncertainty in the form of payment defaults by importers. In doing so, it is seen that the
Welfare enhancing uncertainty 75
dynamics of this uncertainty percolates into the domestic market and affects the economy’s
welfare. Here the firms faced with uncertainty have to bear higher costs than the usual trade
costs, so only very efficient firms are able enter the new export market. These firms,
because they serve both the domestic and the international market and also charge much
lower price earn higher profits. These profits encourage more entrants, and along with the
increased cost of uncertainty more pressure builds up on the single factor market, and
pushes the real wages higher than in a situation without uncertainty. This causes more firms
drawing low productivity exit, as they are unable to survive the increased cost competition
which in turn increases the aggregate welfare of the economy, so any policy to reduce the
cost by for e.g., reducing credit constraints suggested in Muuls (2008), and Manova (2012)
can actually be detrimental. The study may not be fundamentally new, but it has quite a
strong implication, it is a well observed fact that south has lower levels of productivity, and
higher levels of uncertainty, then not only will its exports be compromised but also will its
imports which implies that it will completely loose out to north in the global market and
any kind of effort to reduce or buffer the cost may be deemed inefficient. The policy should
therefore aim to increase the productivity and transparency levels of the south at par with
the north.
In section 2, the model specification is briefly described. In section 3, the pro-competitive
effect of incorporating uncertainty is shown, while the intuitive explanation of the findings
is provided in section 4. Finally, section 5 consists of the conclusion and the direction for
further research.
2. The model
Without loss of generality it is assumed that the world is composed of symmetric
economies, home and rest of the world. Where the preference of a representative consumer,
is given by a usual C.E.S utility function used in Krugman (1980), Eaton and Kortum
(2002), Anderson and Wincoop (2003), Chaney (2008) and primarily Melitz (2003)
however there are several theories, for e.g., Feenstra (2003), Kokovin et al. (2010), Behrens
and Murata (2012), Arkolakis et al. (2019), Demidova (2017) which move beyond the
restrictions of C.E.S utility function but maintain monopolistic competition of the market:
𝑈 ∈Ω
𝑞 𝜔 𝑑𝜔 1
the definitions of all the variables in the above equation and the rest if not explicitly
expressed are equivalent to Melitz (2003), also the mass of varieties is considered to be an
aggregate good, 𝑄 ≡ 𝑈 Dixit and Stiglitz (1977) which has an aggregate price, associated
with it:
𝑃 ∈Ω
𝑝 𝜔 𝑑𝜔 2
Now through the usual maximisation mechanism, the demand for a single variety 𝜔 is
obtained:
𝑞 𝜔 𝑄 3
The economies have a single industry, with continuum of firms, each choosing to produce
a single distinct variety. The quantities of these varieties (q) are produced using a single
homogeneous input; labour (l), and the technology applied is:
𝑙 𝑓 4
again, if w is regarded as the wage rate then total cost of the firm is:
(1)
𝑤𝑙 𝑤 𝑓 5
now 𝑓 0 is the fixed cost, identical for all the firms; however, the variable cost differs
on the basis of labour productivity, 𝜃 0, randomly drawn by the firms and it is further
assumed that even if the quantity demanded is higher total cost of high productivity firm is
always lower. Labour is inelastically supplied at its aggregate level L which also indicates
the economy’s size.
The market is characterized by monopolistic competition so each firm faces downward
sloping a residual demand curve with the constant elasticity 𝜎.
3. The pro-competitive effect of export market uncertainty

3.1. Firm behaviour
Since the market structure is that of monopolistic competition, the price charged by the
firm in the domestic market is:
𝑝 𝜃 6
subsequently, the revenue earned is:

(2)
𝑟 𝜃 𝑅 7
and therefore, the profit from the domestic market is:
𝜋 𝜃 𝑓 𝑓 8
Now, when the economy opens up, the firm, if decides to export, has to bear further costs.
One is the usual iceberg type cost, 𝜏 1. The other cost is a one-time investment cost of
entering the international market, 𝑓 0 which is paid every period out of the export
profits, i.e., 𝑓 𝛿𝑓 , where 𝛿 is the exogenously given probability of death. But apart
from the above usual costs, the firms also face(3), the cost of uncertainty(4), which is intrinsic
to the export market, as much of the business is done on credit basis and information on
credibility of international buyers are either unobservable or extremely expensive to
observe, therefore making it unobservable. Now the price that the firm charges from the
export market representing the increased variable cost is:
𝑝 𝜃 𝜏𝑝 𝜃 9
Since there is an ambiguity regarding the importer’s payment, so the cost of uncertainty
enters the model as the expected revenue of a firm (Neumann and Morgenstern, 1944),
which is:
𝐸 𝑟 𝜃 1 𝛼 𝜏 𝑟 𝜃 10
where 0 𝛼 1 is the exogenously known probability of default by the importers. Now,
the firms expected profit from the export market is:
𝐸 𝜋 𝜃 1 𝛼 𝛼 𝑓 11
The combined revenue is:

𝑟 𝜃 𝑟 𝜃 if the firm serves only the domestic market
𝑟 𝜃 𝐸 𝑟 𝜃 if the firm exports too 12
Again, the combined profit is:
𝜋 𝜃 𝜋 𝜃 𝑚𝑎𝑥 0, 𝐸 𝜋 𝜃 13
Now, there is an abundant pool of identical entrants in the industry. But to enter the market,
these prospective entrants must make an initial investment, which is modelled by a sunk
entry cost 𝑓 0. Only after paying this sunk cost can the firm draw a productivity level,
𝜃 out of a common ex-ante distribution ℎ 𝜃 , and which remains unaffected overtime.
ℎ 𝜃 is a general distribution defined over 𝑅 and has a continuous cumulative distribution
𝐻 𝜃 . After drawing the productivity, the firms with productivity such that, 𝜋 𝜃 0
(excluding 𝑓 ) immediately exit the market, while the rest produce. Furthermore, the firms
that draw productivity such that 𝐸 𝜋 𝜃 0 too apart from 𝜋 (θ) 0 serve both the
domestic and the international market, and they continue to do so until faced with the death
shock. Thus, present value function of the firms’ expected profit can be given by
𝑣 𝜃 max 0, . So, from the above analysis, it is evident that there are two threshold
productivity levels; 𝜃 ∗ 𝑖𝑛𝑓 𝜃: 𝑣 𝜃 0 is the cut-off productivity above which the
firms serve the domestic market, i.e., the firms successfully enters the market, and
𝜃∗ 𝑖𝑛𝑓 𝜃: 𝜃 𝜃 ∗ 𝑎𝑛𝑑 𝐸 𝜋 𝜃 0 𝜃 ∗ is the cut-off of entering the international
(5)
market and is greater than cut off without uncertainty. The ex-ante distribution of the
productivity levels is no longer the ex-post distribution of productivity, which is:
𝜇 𝜃 ∗ if 𝜃 𝜃∗,
0 otherwise.
where 𝜇 𝜃 is determined by the ex-ante distribution, conditional on 𝑝 ≡1 𝐻 𝜃 ∗ , the
∗
probability of successful entry in the domestic market. Furthermore, 𝑝 ∗
represents the ex-ante as well the ex-post probability that one of the successful firms will
enter the export market.
Again, let N is the equilibrium total number of incumbent firms in both the countries,
and 𝑝 𝑁 is the number of firms exporting, so the total number of varieties available to the
consumer of any given country is 𝑁 𝑁 𝑝 𝑁.
Now the aggregate price in (2) can be written as:
𝑃 𝑁 𝑝 𝜃 ,
where 𝜃 𝑁𝜃 𝑁 𝜏 𝜃 is the weighted average of productivity

of all of the firms competing in a single country, that reflects the combined market share
of all firms as well as the output leakage due to exports, as 𝜃 𝜃 𝜃 ∗ and
𝜃 𝜃 𝜃 ∗ alone, which is average the productivity of all firms and only exporting
firms, respectively, does not reflect. Since, this study considers the steady state equilibria;
the aggregate variables remain unchanged over time. Therefore the other aggregate
variables can be expressed as; 𝑄 𝑁 𝑞 𝜃 ,𝑅 𝑁 𝑟 𝜃 ,𝛱 𝑁𝜋 𝜃

3.2. Open economy equilibrium with payment uncertainty
The overall average of the combined revenue and profit is given by:
𝑟̅ 𝑟 𝜃 𝑝 𝐸 𝑟 𝜃
𝜋 𝜋 𝜃
𝑝 𝐸 𝜋 𝜃 14
Now, the zero-cut-off profit condition implies a relationship between the average profits
and the cut-off productivity levels:
𝜋 𝜃∗ 0 𝜋 𝜃 𝑓𝑘 𝜃 ∗ and,
𝐸 𝜋 𝜃∗ 0 𝐸 𝜋 𝜃 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗ 15
∗ ∗
where 𝑘 𝜃 ∗ ∗ 1 ,𝑘 𝜃 ∗ ∗ 1,
∗
θ θ∗ q θ∗ q θ θ∗
𝑗 𝜃 .
θ∗ θ∗ θ θ∗
Again, 𝜃 ∗ can be represented as the function of 𝜃 ∗ :
𝜃∗ 𝜃∗ 16
∗
So, now 𝜋 can be written as a function of the cut-off of successful entry in the market, i.e.,
𝜃 ∗ using (14), (15), and (16):
𝜋 𝑓𝑘 𝜃 ∗ 𝑝 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗ 17
The above equation represents the zero-cut-off profit condition of an open economy with
payment uncertainty.
All of the firms, apart from the firms that have just successfully entered the market, earn
positive profits, so the average profit 𝜋 must be positive. Let 𝑣̅ denote the present value of
the average profit flows, then 𝑣̅ . Furthermore, the net value of entry can be defined as:
∗
𝑣 𝑝 𝑣̅ 𝑤𝑓 𝜋 𝑓 18 .
Since the average profit flow of the firm that is able to successfully enter the market is
positive, the firms want to invest in the sunk cost of entering the market. Since the market
is characterized by monopolistic competition, this implies that there is a free entry and exit.
Hence, 𝑣 0 in equilibrium. If, 𝑣 0, then the firms will observe a positive value of
entry; i.e., the firms will see that the present value of future profits more than covers the
sunk cost. Thus, there will be more entries. If 𝑣 0, then the firms will exit. Therefore,
the free entry condition is:
𝜋 ∗ 𝑓 19
The zero-cut-off profit condition and the free entry condition establish a relationship
between the combined average productivity and the entry cut-off, and the unique
equilibrium values of the two are obtained:
𝜋 𝑓𝑘 𝜃 ∗ 𝑝 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗ (ZCP)
𝜋 ∗ 𝑓 (FE)
Proposition 1: The equilibrium domestic entry cut-off is greater for an export market with
uncertainty than without uncertainty.
Once the equilibrium value of the cut-off productivity for successful entry is obtained, the
value of other endogenous variables, such as the cut-off for entering the export market and
the average productivity levels, are also obtained. Thus, the mass of incumbent firms can
also be determined:
𝑁 20
Proposition 2: The mass of firms serving the domestic market (i.e., the total varieties
available to the consumers) decreases when there is uncertainty in the export market.(6)
As N decreases, the number of firms exporting decreases as 𝑁 𝑝 𝑁 (i.e., number of
imported varieties) decreases. Thus, the total varieties consumed by individuals decrease
for trade under uncertain conditions, so the welfare must decrease because consumers
prefer variety. The welfare function is:
𝑊 𝜌𝜃 ∗ 21
In the above equation, all the variables are equivalent to Melitz, but 𝜃 ∗ 𝜃 ∗ where 𝜃 ∗
represents equilibrium domestic entry cut off without uncertainty is less than it is under
uncertain conditions, which is already stated in proposition 1. This suggests:
Proposition 3: The welfare in trade under uncertain conditions is greater than the welfare
in trade under certain conditions.
4. A graphical analysis
From the above model, it is clear that uncertainty is welfare enhancing. This phenomenon
would further be clarified if the equilibrium conditions in case of closed economy, open
economy with certainty, and open economy with uncertainty are compared. The ZCP
(Zero-cut-off Profit) condition and the FE (Free Entry) condition for a closed economy are:
𝜋 𝑘 𝜃 ∗ 𝑓 (ZCP)
𝜋 ∗ (FE)
Both of the above equations provide a relationship between 𝜋 and 𝜃 ∗ , and so the
equilibrium values of these variables can be obtained by solving them. The equations can
also be represented diagrammatically:


f e
eq m * a 
The intersection of the two curves provides the equilibrium values of the two variables (the
downward sloping schedule represents the ZCP condition while the upward sloping
schedule the FE condition). Now considering the equilibrium conditions of an open
economy with certainty:
𝜋 𝑓𝑘 𝜃 ∗ 𝑝 𝑓 𝑘 𝜃 ∗ (ZCP)
𝜋 ∗ (FE)
Diagrammatically,
ZCPT
 FE
f e
ZCPa
eq m * a  T
* 
From the above diagram, it is observed that the ZCP shifts upwards raising the equilibrium
values of both of the variables when the economy moves from autarky to free trade. This
occurs because trading involves trade costs, which in turn implies that only the efficient
firms enter the export market that charge lower prices and still earn higher profits. Now as
explained by Melitz (2003) that there are two potential effects of trade leading to the entry
cut-off to increase. One is the fact that trade causes competition to increase in the domestic
country due to new foreign firms entering who have high productivities which forces the
prices down. But due to CES assumption price elasticity is constant so demand is not
affected. But the higher profits of the firms attract more entrants, this increases the demand
for labour in the domestic factor market, thereby increasing the real wage. This increased
real wage causes the firms with lower productivities not being able to break even and
therefore exiting the market. Therefore, increasing the equilibrium market entry cut-off,
and since only efficient firms remain in the market, the average profit and so does the level
of welfare. Now, the equilibrium conditions of an open economy with uncertainty are:
𝜋 𝑓𝑘 𝜃 ∗ 𝑝 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗ (ZCP)
𝜋 ∗ 𝑓 (FE)
Diagrammatically,
ZCPT
 FE
f e ZCPTU
ZCPa
eq m * a T*  TU
*

In the above diagram, the ZCP curve shifts even higher to ZCPTU, which increase the
equilibrium values of both the domestic entry cut-off and the average productivity than
when the shift was to ZCPT. The explanation of this shift is similar to that explained
previously. However due to the uncertainty faced by the firms that plan to export, they have
to bear further costs apart from the trade costs. This implies that only very highly efficient
firms are able to enter the export market. The profits of these firms are higher which attracts
higher number of entrants along with the higher costs pushes the real wages higher than
that in case of trade under certain conditions. So, more firms in the lower tail exit, which
not only pushes up the equilibrium domestic entry cut off but also increase the average
profit and the aggregate welfare even though the mass of variety available declines.
5. Conclusion and future research

The literature concerning international trade, heterogeneous firms and export market
uncertainty is fast expanding. As the repercussions, of uncertainty in international market
cannot be ignored. Furthermore, the literature that exists which considers export market
volatility almost always considers it to be the main peril towards the economy and explains
ways to eliminate it or at least reduce it because international trade has always been
considered to be the engine of growth. In this paper, payment uncertainty in the export
market has been incorporated into the standard Melitz (2003) model; however, once
uncertainty is introduced, it can be observed that rather than turning out to be welfare
diminishing, it enhances it by making the competitive effect of trade even more profound.
This occurs because as the market is uncertain, the firms who plan to export products expect
higher costs, which is why only highly efficient firms enter the international market. After
entering the international market, the firms incur higher profits this results in other firms
observing the expanded profits and therefore entering the market. Both increase in costs
and entrants leads to an increase in factor demand which drives up the real wages. The
higher real wage leads to some more of the less productive firms leaving the export market
compared to that in the certain. Since, only the highly productive firms survive, the average
profit increases, and so does the aggregate welfare even though the variety available
decreases.
While there is extensive theoretical and empirical literature concerning heterogeneous
firms and trade, there are still many possible scopes of extensions. It is apparent from these
analyses that bigger firms are more likely to enter a risky market, while smaller firms will
exploit the safer options. Countries with highly productive firms will be able to export more
because they will be able to take more risks, while countries with smaller, less productive
firms will be able to export less because they will try to avoid risks and they will lose out
to the global competition. This could be verified empirically as well as modelled
theoretically by doing away with the symmetric country assumption. Furthermore, in this
paper, uncertainty increases efficiency; however, normally it is deemed inefficient, and
safeguards are provided to diminish its consequences. Therefore, the theoretical model
could be further extended in that direction. Also, the uncertainty considered can be made
heterogeneous i.e. the importers probability of default might differ as different firms have
different responses to changes in the macroeconomic variables.
Notes
(1)
The wage rate is henceforth normalized to unity.
(2)
𝑅 𝑃𝑄 𝑝 𝜃 𝑞 𝜃 𝑑𝜃 𝑟 𝜃 is the aggregate revenue as well as aggregate expenditure.
(3)
For simplicity it is assumed that there is no financial sector, firms bear their own costs.
(4)
The model is static hence the level of uncertainty remains unaltered.
(5)
The threshold for entering the export market is greater with payment ambiguity compared to than
without because if 0 𝛼 1 then 1 𝛼 𝛼 𝑓 𝑓.
(6)
For the derivation of equation (20), refer to appendix B.
References
Anderson, J.E. and Wincoop, E.V., 2003. Gravity with Gravitas: A Solution to the Border Puzzle,
American Economic Review, Vol. 93 (1), pp. 170-192.
Arkolakis, C., Costinot, A., Donaldson, D. and Rodriguez-Clare, A., 2019. The Elusive Pro-
Competitive Effects of Trade, The Review of Economic Studies, Vol. 86 (1), pp. 46-80.
Behrens, K., Muruta, Y., 2012. Trade, Competition and Efficiency. Journal of International
Economics, Vol. 87(1), pp. 1-17.
Bekes, G., Fontagne, L., Murakozy, B. and Vicard, V., 2017. Shipment Frequency of Exporter and
Demand Uncertainty, Reveiw of World Economics, 153, pp. 779-807.
Besedes, T. and Prusa, T.J., 2006. Ins, Outs, and The Duration of Trade, Canadian Journal of
Economics, Vol. 39 (1), pp. 266-295.
Carballo, J., Handley, K. and Limao N., 2018. Economic and Policy Uncertainty: Export Dynamics
and the Value of Agreements, NBER Working Paper, No. 24368.
Chaney, T., 2008. Distorted Gravity: the Intensive and Extensive Margin of International Trade.
American Economic Review. 98 (4), pp. 1707-1721.
Creusen, H. and Lejour, A., 2011. Uncertainty and Export Decision of Dutch Firms, FIW Working
Paper, No. 69.
Crowley, M., Meng, N. and Song, H., 2018. Tariff Scares: Trade Policy Uncertainty and Foreign
Market Entry by Chinese Firms, Journal of International Economics, Vol. (114), pp. 96-115.
Demidova, S., 2017. Trade Policies, Firm Heterogeneity, and Variable Markups, Journal of
International Economics, Vol. 108, pp. 260-273.
Dixit, A. and Stiglitz, J.E., 1977. Monopolistic Competition and Optimum Product Diversity
American Economic Review, Vol. 67(3), pp. 297-308.
Eaton, J. and Kortum, S., 2002. Technology, Geography and Trade, Econometrica, Vol. 70 (5), pp.
1741-1779.
Feenstra, R.C., 2003. A Homothetic Utility Function for Monopolistic Competition Models without
Constant Price Elasticity. Economic Letters.78, pp. 79-86.
Handely, K., 2014. Exporting Under Trade Policy Uncertainty: Theory and Evidence, Journal of
International Economics, Vol. 94 (1), pp. 50-66.
Iacovone, L. and Javorcik, B.S., 2010. Multi-Product Exporters: Product Churning, Uncertainty and
Export Discoveries. The Economic Journal, 120, pp. 481-499.
Kokovin, S., Thisse, J.F. and Zehlobodoko, E., 2010. Monopolistic Competition: beyond the CES.
CEPR Discussion Paper. 7947.
Krugman, P.R., 1980. Scale Economies, Product Differentiation and the Pattern of Trade. American
Economic Review. 70, pp. 950-59.
Manova, K., 2012. Credit Constraints, Heterogeneous Firms and International Trade, The Review of
Economic Studies. Vol. 80(2), pp. 711-744.
Melitz, M.J., 2003. The Impact on Intra- Industry Reallocations and Aggregate Industry
Productivity. Econometrica. 71, pp. 1695-725.
Muuls, M., 2008. Exporters and Credit Constraints: A Firm Level Approach, NBB working paper,
No. 139
Von Neumann, J. and Morgenstern, O., 1944. Theory of Games and Economic Behaviour, Second
edition, 1947; third edition, 1953. Princeton, New Jersey: Princeton University Press.
Rauch, J.E., and Watson, J., 2003. Starting Small in Unfamiliar Environment, International Journal
of International Organisation, 21, pp. 1021-1042.
Segura-Cayuela, R., and Vilarubia, J.M., 2008. Uncertainty and Entry into Export Markets. Banco
de Espa𝑛a Working Paper, No. 0811.
Appendix A
Zero profit cut off condition:
Equation (17) represents the zero cut off profit condition,
𝜋 𝑓𝑘 𝜃 ∗ 𝑝 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗
The average revenue of all the firms in the domestic market can be written as,
𝑟 𝑟 𝜃
Now,
∗ ∗ (from equation (7) in the paper)
∗ ∗
𝑟 𝜃 ∗ 𝑟 𝜃∗ 𝐴1
Again, at 𝜃 ∗ (cut off of successful entry in the market)

𝜋 𝜃∗ 0
∗
𝑓 0
𝑟 𝜃∗ 𝜎𝑓 𝐴2
The average profit of the all the firms in the source country is,
𝜋 𝜃 𝑓
∗
𝜋 𝜃 ∗ 𝑓 from (A1)
𝜋 𝜃 ∗ 𝑓 from (A2)
𝜋 𝜃 𝑓 ∗ 1
Now, let 𝑘 𝜃 ∗ ∗ 1 ,𝜃 𝜃 𝜃 ∗ (for the derivation refer to Melitz (2003)
𝜋 𝜃 𝑓 𝑘 𝜃∗ 𝐴3
Again,
∗ ∗
∗ ∗
𝑟 𝜃 ∗ 𝑟 𝜃∗ 𝐴4
Now at 𝜃 ∗ (the cut off of entering the international market)

𝐸 𝜋 𝜃∗ 0
∗ ∗
1 𝛼 𝛼 𝑓
∗
𝑟 𝜃∗ 𝛼 ∗ 𝑓 𝐴5
The average profit of only the exporting firms,
𝐸 𝜋 𝜃 1 𝛼 𝛼 𝑓
𝐸 𝜋 𝜃 ∗ 𝑟 𝜃∗ 𝛼 𝑓 from (A4)
∗
𝐸 𝜋 𝜃 ∗ 𝛼 ∗ 𝑓 𝛼 𝑓 from (A5)
∗
𝐸 𝜋 𝜃 𝛼 ∗ ∗ ∗ 1 𝑓
∗
∗ 0 and ∗ as average productivity > cut-off productivity
level i.e. 𝜃 𝜃 ∗ so the variable labour cost of firms with productivity 𝜃 is less than
of the firms with productivity 𝜃 ∗ .
∗
Now, let 𝑘 𝜃∗ ∗ 1 and 𝑗 𝜃∗ ∗ ∗
∗
where, 𝜃 𝜃 𝜃 so,
𝐸 𝜋 𝜃 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗ 𝐴6
Since only very efficient firms enter the export market as compared to under certain export
market the expected average profit is higher,
𝐸 𝜋 𝜃 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗ 𝜋 𝜃 𝑓 𝑘 𝜃∗
The combined average profit across all domestic firms is (earned from domestic and
international sales),
𝜋 𝜋 𝜃 𝑝 𝐸 𝜋 𝜃
Using equation (A3) and (A6) combined average profit can be written as,
𝜋 𝑓𝑘 𝜃 ∗ 𝑝 𝑓 𝑘 𝜃∗ 𝛼𝑗 𝜃 ∗ , hence the ZCP.
International market entry cut off as a function of domestic entry cut off:
It is known that,
∗
𝑟 𝜃∗ 𝑅 𝐴7
and,
∗
𝐸 𝑟 𝜃∗ 1 𝛼 𝑅 𝐴8
Now, using (A7) and (A8),

∗ ∗
∗ 1 𝛼 𝜏 ∗ 𝐴9
Again at 𝜃 ∗ ,
𝜋 𝜃∗ 0
𝑟 𝜃∗ 𝜎𝑓 𝐴10
At 𝜃 ∗ ,
𝐸 𝜋 𝜃∗ 0
∗
1 𝛼 𝑟 𝜃∗ 𝜎 𝛼 ∗ 𝑓 𝐴11
So, using (A10) and (A11)

∗
∗ ∗
∗ 𝐴12
Equating (A9) and (A12),

∗
∗ ∗
1 𝛼 𝜏 ∗
∗
∗ ∗
∗
𝜃∗ 𝜃∗
∗
Appendix B
Equilibrium number of domestic firms:
The combined average revenue of all the firms is,
𝑟̅ 𝑟 𝜃 𝑝 𝐸 𝑟 𝜃
also,
𝑟̅ 𝑁 ̅
Now, L=R (refer Melitz (2003)).

Again,
𝜋 𝜃 𝑓
𝑟 𝜃 𝜎 𝜋 𝜃 𝑓 𝐵1
and
𝐸 𝜋 𝜃 1 𝛼 𝛼 𝑓
1 𝛼 𝑟 𝜃 𝜎 𝐸 𝜋 𝜃 𝛼 𝑓 𝐵2
So now, mass of firms, 𝑁 using (B1) and (B2) and putting it in
𝑁 ̅
(it is assumed that 𝑝 𝑝 i.e. the fraction of firms exporting out of total firms
exporting remains unaltered in equilibrium in uncertainty). Comparing N in case of export
without uncertainty and N with, it is clear that this N is lesser.
Business process modeling.

Using Unified Modeling Language to streamline
the design of the TO-BE system within a company
Ionuț NICA
ionut.nica@csie.ase.ro
Ștefan IONESCU
stefion09@gmail.com
Abstract. The UML modeling language is a set of methods that describe the dynamic and structural
properties and is used to evaluate the structure of a systems architecture. This modeling language
presents several types of predefined diagrams that have the role of synthesizing in visual form the
most eloquent information necessary to characterize a system or to understand a process. In this
research, we will use UML to propose the improvement of an existing application within Adobe, the
purpose of which is to send promotional offers through multiple channels to members of the Adobe
community. We will also approach at a theoretical level the description of the basic notions and
analyzes used in modeling business processes and necessary for the diagnosis of business systems.
Keywords: business analysis, UML, PESTELE analysis.
JEL Classification: C63, L21, M21, O22, O3.

90 Ionuț Nica, Ștefan Ionescu
1. Introduction
Business analysis is a set of activities that coordinate change in an organizational context
by defining needs and proposing solutions that bring value to the business. This value is
determined by achieving benefits, reducing costs and identifying new opportunities within
the organization.
We chose to approach this field in this research because we consider that business analysis
has a significant impact on the evolution of companies. Thus, we want to actively
participate in the change for the better of the current state (AS-IS), based on the
uninterrupted action of feedback processes.
In this research, we will highlight the importance of business analysis within an IT
company. This will be achieved both through process-oriented methods, widely used by
business analysts, and through a systemic, cybernetics type approach that highlights the
appearance of the enterprise's complex adaptive system.
A business process is a set of structured and interconnected activities that can be performed
by people or equipment and serve a specific business purpose, when performed in a certain
order. It can often be viewed as a flowchart of a sequence of activities with interleaved
decision nodes or as a matrix consisting of series of activities with relevance rules based
on the data involved in the flow. The benefits of using business processes include improved
customer satisfaction and increased agility in responding to sudden market changes.
Organizations that adopt process-oriented approaches break down the barriers of classical
structuring by departments and try to avoid the emergence of subsystems isolated from the
rest of the functional ensemble.
The premise of a business process is the achievement of objectives set for the realization
of a project or the understanding of a system, and the conclusion is the satisfaction of the
objective of having a result that brings value to the client. The person responsible for the
execution of the process throughout the life of the process without interrupting it is the
process owner. He is represented by a person who has in-depth knowledge in this field and
the skills required for this position. Its attributions include defining the mission, vision and
objectives of the process, respectively satisfying the key performance indicators. They
evaluate the success of an organization, a particular activity (such as projects, programs,
products, and other initiatives) or a particular individual.
In general, business processes can be classified into three types, in von Rosling's view:
 Operational: They are the core of the business and create the main value stream (for
example, taking orders from customers, opening an account or producing components).
 Management processes: They have the role of supervising the operational ones, such as
corporate governance, budgetary supervision or employees.
 Support processes: They support the core of operational processes (i.e. accounting,
recruitment, customer relations, job security).
It is possible to break down a complex process into several sub-processes, each with its
own specific list of attributes, but which, at the same time, contributes to the overall purpose
of the business. Their analysis includes the mapping or modeling of them at a general level,
but also of the subprocesses at the level of individual activity.
Business process modeling 91
2. Software support systems for business processes

Lucidchart(1) is a web platform that facilitates user access to contribute, review and share
various charts and diagrams.
Process modeling is the main goal of the Lucidchart application. To this end, it offers the
possibility to optimize the workspace with the help of custom settings and templates,
creating interactive layers of threads, as well as making recommendations within them.
Also, the documentation can be easily made and stored in applications such as Confluence
and Jira, which facilitate project management and collaboration between team members.
The forms of visualization are extremely varied and can be presented to stakeholders in a
professional manner and easily adapted to the approach.
ProcessMaker BPM is a web application that allows users to create and edit process
diagrams using the BPMN 2.0 standard. Dynamic shapes (Dynaforms) in ProcessMaker
are based on the intuitive drag-and-drop interface. Users have a wide variety of options,
including creating custom templates, integrating with variables and external files. Although
no programming experience is required, developers can work directly in JavaScript if they
wish. Processmaker BPM also provides multiple ways to view dashboards. Dashboards are
tools that give decision-makers full visibility into processes and employees. ProcessMaker
has the following key performance indicators (KPIs):
 Process Efficiency Index (PEI): This index “learns” from the behavior of processes over
time and establishes optimal levels of performance based on several factors, such as
standard deviations, resource costs and other statistics.
 Employee Efficiency Index (EEI): Similar to PEI, this index measures the efficiency of
each employee over time and makes rankings based on efficiency and cost reductions
for the organization.
Another modeling and design tool used in business processes, which offers much more
complex techniques, is Sparx Systems Enterprise Architect(2). It uses visual design
techniques that are based on the UML Object Management Group. Enterprise Architect is
a software that is used by many enterprises in order to design and model the architecture of
business systems but also to implement certain models that will be processed throughout
the development life cycle.
Microsoft Visio(3) is a desktop application used to create a wide variety of diagrams. These
include flow charts, flow charts, process charts, 3D maps and more. Microsoft Visio also
provides comprehensive sets of templates and shapes for making diagrams.
For IT companies, Visio provides the ability to make data flow diagrams (DFDs). These
can be a software development-focused approach, which involves thorough research before
moving on to the actual programming part. Business analysts use DFDs to evaluate existing
systems. The representation of the process may reveal steps that might otherwise be omitted
or misunderstood. Regarding the business part, the modeling of the processes according to
the BPMN standard is performed for the other participants, respectively stakeholders in a
business process in order to fully understand it in a visual and easily deducible way of the
stages. From a slightly deeper perspective, it is intended for the people who will handle the
implementation of the process, providing an appropriate number of details for a successful
execution.
3. Important notions in the analysis of business processes

Business analysis: According to BABOK® (A Guide to the Business Analysis Body of
Knowledge – Babok Guide, 3rd edition), is the practice of enabling change in an
enterprise/organization by defining needs and recommending solutions that provide
value to stakeholders.
Business analysis is a relatively new economic discipline that promises to provide great
benefits to organizations, ensuring that business requirements are aligned with the business
change solutions implemented. Many of these solutions will include new or improved IT
systems, and others may have a broader purpose, incorporating changes in areas such as
business processes and employee roles.
Issues that arise: Organizations have introduced Business Analysis to ensure that business
requirements are met when new information technology (IT) systems are introduced.
However, recognizing their importance is in principle easier if one considers how they can
be achieved. Some business analysts are experienced as systems analysts and are less
accustomed to considering business requirements and types of potential solutions when
meeting such requirements. Many business analysts come from the business environment
and have a limited understanding of IT and how computing systems have developed. While
business knowledge is important to these business analysts, problems can arise when IT is
part of the solution and the analyst has an insufficient understanding of IT. This leads to
difficulties in communicating with developers and can lead to mistakes in formulating an
integrated solution from both a business and IT perspective. Some business analysts, even
if they have experience and knowledge, have failed to provide beneficial advice to their
organizations and, as a result, have a misunderstanding of their role in organizations, which
has led organizations to refuse to apply the solutions they propose. The activity of business
analysts today is well defined on the basis of technical standards that have been established
for many years. In fact, some of these techniques were used even before the role of the
business analyst became clear in the existence of organizations.
Design: Viable representation of a solution. It focuses on how the value will be delivered
through the solution, if realized. The nature of the representation can be a document (or a
set of documents) and can vary significantly depending on the circumstances.
Plan: Proposal to do or accomplish something. The plans describe a set of events, the
relationships between the events, the program, the expected results, the necessary resources
and the stakeholders involved.
Requirement: Viable representation of a need. The requirements focus on understanding
the value that can be achieved if a requirement is met. The nature of the representation can
be a document (or a set of documents) and can vary significantly depending on the
circumstances.
Risk: The effect of uncertainty about the value of a change or solution. Business analysts
work with other stakeholders to identify, assess and prioritize risks, applying risk
management methods: mitigating consequences, eliminating the source of risk, avoiding
starting a risky business, sharing risk, accepting or even assuming an even greater threat, if
a suitable opportunity arises.
Classification of requirements: According to the BABOK® Guide, requirements are

classified under the following categories:
 Business requirements: Expresses objectives and results that describe why a change was
initiated. They can be valid for the whole enterprise, a certain department or a certain
project.
 Usage requirements: Describe the needs of stakeholders that must be met in order for
business requirements to be met. They can serve as a bridge between business
requirements and solution-specific requirements.
 Solution requirements: Describe the characteristics of a solution that meets the
requirements of stakeholders. They provide an appropriate level of detail so that the
solution can be developed and implemented. These can be divided into two
subcategories:
– functional requirements – describe the characteristics that a solution should have in
terms of its behavior (mode of operation);
– non-functional requirements – are not directly related to the operation of the solution,
but refer to the situations in which the solution must remain effective, respectively
to the qualities that the solution will have to have.
– Transition/implementation requirements: Presents the specifications that a solution
must have, but also its conditions of efficient operation to facilitate the transition
from the current state (AS-IS) to a future state (TO-BE), but which they are no
longer needed as soon as the change has taken place. They differ from other types of
requirements in that they are temporary in nature and often refer to areas such as data
conversion, employee training or the implementation of a different management
approach.
Stakeholders: Each task includes a list of stakeholders who will participate in or be
affected by it. A stakeholder is an individual with whom a business analyst will interact
directly or indirectly during the ongoing task. Any stakeholder can be a source of
requirements, assumptions or constraints.
Some of the stakeholders involved in a business initiative can be found in the following
list, which is not exhaustive and can vary depending on requirements.
 Business analyst: Inherently a stakeholder in all activities specific to business analysis.
According to BABOK, he is responsible and accountable for these activities.
 Customer: Uses products or services provided by the company and may have
contractual or moral obligations that the company undertakes to fulfill.
 Expert in the field (business): Has detailed knowledge about a specific topic relevant to
the business need or scope of the solution. Here you can find people such as managers
or consultants.
 End user: Interacts directly with the solution. Here you can find all those who participate
in a business process or use the proposed solution.
 Expert in the field (implementation): Has extensive knowledge regarding the
implementation of one or more components of the solution. Some of the people who
can fulfill this role are: developer, solution architect, database administrator, change
manager.
 Operational support: Responsible for the regular maintenance of a system or solution.

It can be provided by: operations analysts, technicians, launch managers.
 Project manager: It manages the activities necessary to deliver a solution that meets a
specific business need and the achievement of project objectives, taking into account
factors such as scope, budget, program, resources, quality and risk.
 Supervisory authority: Responsible for defining and imposing standards. Standards can
be imposed on the solution through legislation, corporate governance methodologies or
auditing. These include: the government, public regulators and auditors.
 Sponsor: Responsible for initiating the effort in defining a business need and developing
a solution that meets that need. They authorize the efforts to be made and control the
budget and the scope of the initiative.
 Supplier: Stakeholder who is outside the boundaries of a particular organization or
business area. It provides the company with products or services in accordance with
contractual or moral obligations to be fulfilled. These include: consultants, vendors,
providers.
 Tester: Responsible both for verifying how the solution meets the requirements defined
by the business analyst, and for directing the verification process. They also ensure that
the solution meets quality standards and that the risk of gaps is well determined and
minimized. An alternative role is that of Quality Assurance analyst.
Collaboration between stakeholders involves frequent and two-way communication.
Collaborative relationships help maintain the free flow of information when obstacles arise
and promote a common effort to solve problems and achieve the desired results.
3.1. UML Notations
A Use Case represents a collection of possible scenarios, regarding the communication
between the system and the external actors, characterized by certain purposes. These
scenarios are different in the sequence of steps to which lower-level use cases may
correspond. Use cases show WHAT THE SYSTEM SHOULD DO and NOT HOW.
Actor – The notation of actor is used as a
behavioral classifier that has a role that plays
in the interaction with the initiative proposed
in the analysis. The actor can represent an
internal or external entity that interacts with
the subject to analysis, the use case. It can be
a human user of the carefully designed
system, it can be another system that
interacts with the attentive system or even a
hardware device that uses the services of the
analyzed system.
The identification is made by answering the following questions: Who is interested in the
information in the system? Who changes data? Who interacts with the system? Use cases
are represented in the form of an ellipse inside which the name of the Use case is written.
The noun usually begins with a verb. The association is used to indicate the connection
between an actor and a use case, in the sense that that actor participates in some way in that
use case. There may be a generalizing relationship between the actors. If an actor inherits
another actor, then he can communicate with the same use cases of the system as the parent.
When a use case includes the behavior of another use case, the include relationship is
formed. “Included” use cases cannot be used independently, but only as part of the use
cases that include them. The inclusion relationship is used to avoid repeatedly describing
the same stream of events.
When a use case is inserted in another, but only under certain conditions, the extended
relationship is formed. The extend relation (and implicitly the use extension cases) are
used when modeling an optional or exceptional behavior, which does not condition the
finality of the basic use case. When a UML element depends on other elements for
implementation, the dependency relationship is created, which is a directed relationship.
The activity diagram models sequences of actions executed by elements of the system
and shows the business and the flow of activities.
Action represents the elementary unit of
behavior (calling behavior/operation).
Node is a branch point of a stream. The
initial state represents the point of entry
into the activity in attention. This node is
unique and from it always starts a single
transition. The final state represents the
point of exit from the activity. This can
be the end point when the goal in focus
is achieved or a node that concludes a
flow of attention activity. The decision
node is used when modeling a point in
the flow of attention where a choice must
be made based on a particular branch of
the flow. The notation of the decision
node is made in the geometric shape of a
rhombus. The same notation is used if
flows are reunited after a previous
decision, in which case no more
conditions are required. This node is
called the merge node. When two other actions are performed simultaneously from one
action, the fork node is used. When we have the situation in which two simultaneous
actions are performed on which the next action depends, the join node is used.
4. PESTELE Analysis
The management of the environment in which the actions of a company are carried out is
a main factor of efficient development that includes planning, protection, monitoring,
evaluation, research, education, conservation and sustainable use of resources.
PESTLE analysis is the most common approach for considering the external business
environment. It is a political, economic, social, technological, legal and environmental
analysis and describes a framework of macroeconomic environmental factors used in
strategic management. If in the case of SWOT analysis the assessment of the four
components is made both to the internal environment of the organization and to the external
environment, the basis of the PESTLE analysis starts from the fact that the company must
react to changes in the external environment and adapt.
Another very important factor that we must consider in the existing society is the ethical
factor. Thus, an optimized method of PESTLE analysis is PESTELE analysis which also
includes analysis from an ethical perspective.
Political changes could include general changes in the internal political climate, the effects
of European integration and the subsequent effects of the collapse of the Soviet Union, the
change of government, changes in world power, as well as those specific to legislation and
regulations. All these are examples related to the political analysis within PESTELE.
Examples of economic changes can be the effects of business cycles, world trade patterns,
changes in the exchange rate, commodity prices, changes in capital markets, and economic
effects on suppliers and particular groups of customers.
Social change includes the effects of categories, demographic habits and environmental
concerns and sustainable development. Behavior patterns of individuals and certain groups
of agents reflect their attitudes, beliefs, and values. The social environment includes the
attitudes and values of society, as well as the behavior, which is motivated by these values.
Technological changes cover the effects that take place on products, processes and
distribution channels. The technological factor includes the innovations with which the
technology is developed and impacts in a favorable or unfavorable way all the processes of
the organization or even the market operations.
The legal component of the PESTELE analysis refers to the fact that all legal processes
such as taxes, quotas, labor, imports, etc. will be taken into account. Depending on the
country in which it operates, an organization may be influenced by certain policies,
restrictions, laws, for example, consumer law or labor rules.
Environmental factors represent the last component of the PESTELE analysis and include
all the parameters that impact the business environment.
In the literature, ethics (a term derived from the Greek “ethos” – usually, character) is “the
philosophical discipline that studies the theoretical and practical problems of morality”; in
everyday speech, the term ethics is often used in the sense of morality. Thus, on the one hand,
“ethics is the science that studies moral principles, their origin, nature, essence, development
and content”, and, on the other hand, “ethics represents the set of moral rules, values and
norms that regulate human behavior in the company and/or determines their obligations, in
general, or in a certain field of activity, in particular”. Practically, all the factors mentioned
above from the PESTLE analysis must be analyzed from an ethical perspective.
PESTELE analysis is performed in general, which makes it difficult to define clear rules
on how best to apply it in different circumstances. Globally dispersed businesses will need
to perform separate PESETLE analyzes for different regions, as trends occur at different
frequencies in different locations. The value of PESTELE is directly related to the quality
of the effort made and to the correctness and documentation of possible influences of
external changes on the analyzed enterprise and industry.
PESTELE analysis is a useful tool for understanding the growth or decline of the market
and the position, potential and direction of a business. Such an analysis is a useful tool for
measuring business.
5. Study case: Business process modeling in Adobe Company

Adobe Inc. is an international software company that aims to provide the tools and solutions
needed to create, design and deliver the best digital experience for its customers. With a
turnover of over $ 11 billion (2019) and over 20,000 employees globally (2020), Adobe is
one of the leading players in the software industry. As users become more aware and
technical-oriented, and technologies such as virtual reality and artificial intelligence
become more common, Adobe has begun to face challenges in maintaining market
dominance, a need to renew business strategies. In order to produce reliable strategic
recommendations, it was necessary to analyze Adobe's history and current market position.
Any business analyst, before performing the analysis for an initiative, must know the
analyzed organization and understand the processes that impact the analyzed initiative.
Adobe Inc. (formerly Adobe Systems Inc.) was founded in 1982 by John Warnock and
Charles Geschke in Mountain View, California. The founders decided to start this company
to bring the PostScript language to market because their previous employer, Xerox, did not
want to do so. Steve Jobs wanted to buy Adobe, but the founders turned down the option.
However, Jobs managed to buy 19% of the company's shares at a price 5 times the actual
valuation and a 5-year, pre-paid PostScript license. Thus, Adobe became the first company
in the Silicon Valley area to make a profit in its first year of operation. PostScript was later
used in laser printing and dominated the market in the publishing industry. The company
was listed on the US NASDAQ in 1986 and launched the PDF document format in 1993.
In 1996, the company moved its headquarters to San Jose, California. Over the years,
Adobe has distributed creative software either developed in-house or taken over through
the acquisition of certain companies. Many of these software products are famous, such as
Acrobat Reader, Photoshop and Creative Cloud. The company currently has offices in more
than 25 countries, more than 20,000 employees globally and a turnover of more than $ 11
billion in 2019 alone.
Adobe operates in three sectors: Digital Media, Digital Experience and Publishing. The
first two of these are part of the company's long-term growth and represent the areas in
which it aims to attract investment and create solutions for the market.
5.1. Study case: PESTELE Analysis in Adobe Company
It is a political, economic, social, technological, legal and environmental analysis and
describes a framework of macroeconomic environmental factors used in strategic
management. The PESTELE analysis describes how the company should react to the
changes in each of the mediated sectors. All the factors described below were analyzed
taking into account the ethical factor.
Political factors
 The impact of the Trump administration's policies: Because of the new US government,
there are more regulations regarding employment rules in this country, so organizations
like Adobe could be forced to update their employment policies and employ fewer
foreigners. This could affect the overall quality of the products and services provided by
Adobe. Political ambiguity has also forced more immigrants to drop out of the United
States and choose other countries to pursue higher education. Since a large proportion of
Adobe employees are immigrants, these policies have adverse effects on the company.
 Promoting digitalization in emerging economies by governments: Emerging economies
such as India, China, Brazil and other countries in Asia and South America are focusing
on the digitalization process, in which they are investing heavily. The annual growth
rates of the software industry in Asia and Africa are 11% and 10%, respectively,
according to the CIO, which results in a significant increase in revenues in these regions.
A new initiative has been taken by the Government of India, namely the digitization of
public services. Thus, several start-up companies have appeared on the market that offer
digitization services in this regard. Most of these platforms are designed based on Adobe
products and enjoy a consistent demand. The government itself supports digitalization
by setting up development centers for day-to-day operations in public institutions. Such
an initiative in developing countries is good for Adobe. Brazil has publicly discussed
its digitization strategy and is actively pursuing its implementation, and Mexico has
launched its National Digital Strategy, which is aggressively pursuing various avenues
of digitization.
Economic factors
 Growth in emerging markets: Emerging markets are responsible for more than half of
global GDP growth and about 40% of total global GDP, according to Barakaat
Consulting. By 2030, Asia will have the largest share of global GDP at 40%. Investments
in cloud infrastructure and associated technologies and applications will contribute to the
growth of emerging economies at a significantly faster rate than developed economies.
Sectors such as e-commerce, cloud health services, payment and banking ecosystems will
be the main growth channels for emerging economies, such as India, part of Africa and
the Middle East. Hiring more people in the software fields and encouraging the adoption
of online technologies by consumers will significantly contribute to the growth of the
company. Thus, emerging markets represent real opportunities for Adobe.
 Foreign currency fluctuations: Users of Adobe products and services are present all over the
world, and currency fluctuations pose a serious threat to Adobe's profitability. Revenues
outside the United States were also affected by fluctuations in foreign currencies. At the end
of fiscal 2018, Adobe lost approximately $ 8.3 million due to currency fluctuations. The
increase in crude oil prices together with the increase in revenues in the APAC region (Asia-
Pacific) can also have a considerable impact on Adobe's profitability.
Social factors
 The explosion of the mobile application market: Awareness of the importance of using
a mobile platform for the day-to-day sales activities has grown exponentially since the
advent of e-commerce platforms. Multinational companies have realized the importance

of providing mobile services to customers, and this has set the standard for a suite of
cloud service providers such as Adobe. Currently, according to a Digital Future in Focus
report, this point has been far exceeded in countries such as India, Mexico and
Indonesia, with a 4: 1 ratio of the number of smartphone users versus desktops.
According to PwC, 51% of online advertising budgets went to smartphones. This aspect
is also supported by the company Smart Insights, which certifies that 80% of Internet
users will own a smartphone in the near future. This growth is expected to continue for
at least another ten years and is clearly beneficial to Adobe.
 Concerns about privacy and data security policies: Poor data protection is a major
concern in the software industry, and Adobe is no exception. In the past, there have been
security breaches within Adobe, and these incidents have had the effect of misinforming
some of its customers. In October 2013, a group of hackers managed to steal the data of
38 million Adobe users, and later found out that they were very poorly encrypted.
Norton recently announced that it has not found any vulnerabilities in the widely
distributed Adobe Flash Player program.
Technological factors
 Evolution of SaaS (Software as a Service) and IaaS (Infrastructure as a Service)
platforms: The evolution of SaaS and IaaS platforms has reduced the exorbitant
installation and maintenance costs in terms of online presence. As a result, more and
more companies want to establish an online presence to reach customers more easily.
This evolution laid the foundations of Adobe SaaS platforms.
 The evolution of cloud technology: Cloud technology is considered one of the main
vectors of digital transformation, and Adobe is at the center of them through solutions
such as Creative Cloud, Marketing Cloud and Document cloud, which serve customers
through the most advanced technologies available on the market. The innovation
brought by Adobe attracts many new customers in the use of cloud solutions. Also, the
evolution of cloud technology has resulted in the implementation of the subscription-
based business model, generating massive profits for Adobe.
 Fast mobile networks in developing countries: The speed of mobile internet from 2G,
3G, 4G to 5G has gradually increased in developing countries, and people in these areas
have access to cutting-edge technologies from anywhere in the world, and this is
beneficial for Adobe by increasing the overall number of customers.
Legal factors
 Compliance with legal regulations for the use of electronic signatures.
The main market for digital signatures is North America, governed by associated laws and
rules, giving rise to a significant number of legal companies in countries such as the USA
and Canada. Directive 1999/93/EC on electronic signatures was adopted by the European
Parliament in 1999. The adoption of the digital signature in the Asia-Pacific region has
enjoyed a significant boost due to the growth of the banking, financial services and
insurance sector and the initiatives taken by the governments of countries such as China
and India to promote the use of digital technologies. The global digital signature market is
facing challenges from alternative technologies such as vein biometrics, voice and
keystrokes. Digital signatures are the most advanced and secure type of electronic
signature, used for credit applications, healthcare, the emergence of new drugs and other
business processes that are strictly regulated due to the specificity of high risk and high
value. In the European Union, digital signatures are the only form of electronic signature
with a legal status equivalent to a handwritten signature. To ensure compliance with
government and industry-specific regulations, organizations use digital identities from
trusted providers found on certain official lists. Some of these are the European Union
Trusted Lists (EUTL) and the Adobe Approved Trust List (AATL).
 Criminal prosecution of damages caused by security breaches: Security breaches are the
ghosts that haunt the software industry. Adobe is no exception to this, and as a result of
these recent breaches, the US government is preparing to implement a legal framework
through which customers can sue the service provider in the event of a security breach.
As Adobe has suffered greatly from these shortcomings in the past, it will need to act
appropriately to minimize these incidents in the future.
 Data protection requirements: With the misuse of customer data during the scandal
involving Facebook and Cambridge Analytica, keeping this data safe has become the
main goal of every business. The European Union's General Data Protection Regulation
(GDPR) entered into force on 25 May 2018. With the help of these rules, businesses
will need to focus on building brand loyalty and prioritizing customer security. The
GDPR is the latest European Union law addressing data protection requirements. These
rules have a wide applicability and have an impact on any business that collects personal
data of EU citizens. Adobe hopes to make a drastic contribution to the GDPR's transition
to many businesses. Adobe Experience Cloud will support brands (“data controllers”),
technology providers (“data processors”), and areas where processors could partner with
controllers through a variety of tools, processes, and documentation. Alisa Bergman,
director of security at Adobe, says that “Adobe is at the forefront of brand support to
become an experienced business, and GDPR brings the perfect opportunity for
companies to position their core customer base, gain trust through transparency and
improving the customer experience with security in mind”.
Environmental factors
 Services for electronic documents: Electronic document services have drastically reduced
paper consumption. Adobe Acrobat is the main vector of this transformation, and books
in electronic format have become the new rule, those in physical format becoming
obsolete. Thus, a major market was born, and Adobe has a monopoly within it.
 Adopting sustainability in the way we work: At Adobe, sustainability is embedded in
the work of every employee. Each Adobe project is monitored for this purpose and
approved only when it is considered to be environmentally viable. In addition, Adobe
invokes strict waste and water management policies.
 Corporate social responsibility initiatives: As part of its corporate social responsibility
policy, Adobe has several initiatives to make its offices more environmentally friendly.
For example, the Adobe office in Bangalore is powered by solar energy. Adobe's plan
is to supply 100% of its energy by 2035. Adobe has received the Dow Jones Gold
Standard for Corporate Social Responsibility for two years in a row, a testament to
Adobe's contributions to the environment.
5.2. UML modeling within the implementation of a system for sending promotional offers on
several channels
In this section, we will suggest improving an existing application within Adobe, which
aims to send promotional offers through multiple channels to members of the Adobe
community 4 .
Description of the actors:
 Marketing representative: Is the person who logs into Adobe Experience Cloud to
access the members database and uses the Adobe Campaign platform to send them
promotional offers to those who have provided their phone number and email address
and who, depending on their response, will send a reminder to the member either by
email (if the email was not read) or by SMS (if the email was read but the promotional
link was not accessed) or will update the members database, reviewing audience groups
in case success (email has been opened and the promotional link has been accessed).
 Member: This is the person who receives the promotional email and first decides
whether to read the email and secondly, if he will read it, will have to decide whether to
access the promotional link or not. If you do not read this email, after three days you
will receive another informative email about the same offer. If, instead, he reads the
email but does not access the offer, he will receive an SMS with the same purpose.
 Adobe Campaign: The two actors described above are joined by the actor “System”,
the Adobe Campaign platform being its first instance. Through this platform, the
marketing representative will log in to his email account and access it in order to
perform flow-specific activities.
 Adobe Experience Cloud: This is the second component of the “System” actor. Within
this cloud is the database of members in which the marketing representative will initially
authenticate. After accessing the email using the Adobe Campaign platform, the
representative will use the Adobe Experience Cloud to enter the members database.
AS-IS Situation:
Currently, there is a management flow of the activity of sending promotional offers. In the
first phase, this involves authentication by the marketing representative in the database of
members of the Adobe community, found in Adobe Experience Cloud. After that, the
representative logs in to the email account and accesses it in order to be able to add the
recipients in the email according to the members found in the database. Only members who
have provided both their phone number and email address become recipients of the email.
Then, the marketing representative will compose the email message, along with which he
will attach the link to the promotional offer and will send the email to the eligible members.
Recipients will be given 3 days to open the email with the promotional offer. After that,
the following actions will be performed depending on the effects produced by the
previously sent email:
 Another email will be sent as a reminder of the offer to the people who did not open
the email, they will receive it, and the flow is over;
 An SMS with the same purpose will be sent to the people who opened the email, but
did not access the link to the offer, they will receive the SMS and the flow is over;
 The database will be updated in Adobe Experience Cloud, in the case of people who
opened the email and accessed the promotional offer.
Figure 1. AS-IS activity diagram of the bidding process based on email and SMS channels

Figure 2. TO-BE use case diagram of the bidding process based on email and SMS channels
In the new form, the application involves a new actor, represented by the system
administrator. It provides extra control and security over the business by automatically
blocking the Adobe Experience Cloud and Adobe Campaign accounts of the marketing
representative. These accounts will be unlocked after successfully verifying the user's
identity. Otherwise, the flow ends if the marketer fails to identify himself.
From the point of view of carrying out the activity, the verification will no longer be
conditional on the availability of the email address and telephone number of the members.
These will take place in two separate and parallel sub-streams, based on checking the email
address and phone number, respectively. The waiting time of 3 days is kept between the
email, respectively the initial SMS and the one for the purpose of reminder, but, unlike the
AS-IS variant, these days will coincide and the waiting time is shortened, because they will
no longer produces successively (SMS by email).
The use-case diagram of the desired state, presented above, suggestively expresses the
interaction of each of the actors presented (marketing representative, member, Adobe
Experience Cloud, Adobe Campaign and system administrator) in the process of sending
promotional offers.
As a result of the transformation of the process, Adobe will gain added value in its multi-
channel marketing. The added value is increased security from the involvement of a system
administrator if database and email authentications are unsuccessful. In addition, the fact
that in the new state the flows of sending offers by email, respectively SMS take place in
parallel (simultaneously) and not successively, the execution time of the whole process
decreases from 6 days to 3 days, so there is a halving of the actual time, working time that
can be used to more effectively target potential customers, which would result in an
increased conversion rate.
6. Conclusions
Business analysis is vital for the company as it capitalizes on the internal potential, but also
the components of the external environment, in order to gain a favorable position in the
market. Following the presentation of the concepts and methods in the previous sections,
the complexity of the field of business analysis and the fact that this area is in perpetual
development, as the companies themselves have a continuous evolution.
In the first part we presented the general concepts related to the analysis of business
processes and we highlighted the importance of modeling them. We also presented some
software support systems used for efficient process management.
In the second part we highlighted the methodologies used for business processes (UML
modeling and BPMN modeling). Moreover, the PESTLE and PESTELE analysis,
respectively, proved to be useful to understand the factors from the external environment
that impact the organization in attention. Organizational modeling must be done from a
cybernetic perspective and this means, compared to the traditional approach, to understand
the processes in the external environment that influence the organization, giving a complete
view of the company and the enterprise ecosystem. Cybernetics is the science of
communication and control and, from this point of view, an understanding of systems as
cybernetic systems facilitates modeling and this proposes optimal solutions to remedy some
identified deficiencies. The case study was approached in the last part of the research,
constituting a complex research in order to model the business processes within the Adobe
Company. Regarding the activity of the marketing department of Adobe, for the process of
sending offers based on SMS and email channels, a topic that comes to the attention of the
research analysis, improvements have been proposed compared to its current state.
Therefore, if the flows of sending offers by email and SMS will take place simultaneously,
the duration of the process will be halved (from 6 to 3 days), and the time saved can be
used to streamline the segmentation of potential customers. The modeling of business
processes comes to the attention of today's specialists and the techniques in this field offer
a precision and a clear understanding of the system so as to propose the optimal solutions
for achieving the value that the proposed change in the enterprise must bring. For the next
research we will approach the Agile framework and its importance in optimizing business
initiatives.
Notes
1 The Luchidchart modeling program can be accessed from the following available address:
https://www.lucidchart.com
2 The Enterprise Architect modeling program can be accessed from the following available address:
https://sparxsystems.com/
3 The activity diagram and the use case diagram were represented using the Microsoft Visio tool.
4 For recommendations, suggestions and observations, you can contact the authors Ionuț Nica
(ionut.nica@csie.ase.ro) or Ștefan Ionescu (stefion09@gmail.com).
References
Adobe official site, 2020. Available online: <https://www.adobe.com/ro/> [accessed on

06.11.2020].
Al-Jubran, H.H. and Bu Hassan, S.A., 2009. Safety and operations measures taken by one company
to conduct onshore rigless critical well intervention operations in water, oil and gas wells.
DOI: <https://doi.org/10.2118/121667-MS> Society of Petroleum Engineers.
Charroux, B., Osmani, A. and Thierry-Mieg, Y., 2009. UML 2. Pratique de la modélisation. 2e
édition. Pearson Education.
Chiriță, N. and Nica, I., 2019. Cibernetica firmei. Aplicații și studii de caz. Economica.
Glykas, M., 2013. Business Process Management Theory and Applications, Springer.
González, C.R. (editor) and Calpe, S.A., 2000. Diccionario Enciclopédico Nuevo Espasa Ilustrado.
Editorial Espasa.
Hay, D.C., 2011. Enterprise Model Patterns – The UML version. Technics Publications, LLC.
Iamandi, I.E., 2012. Etică şi responsabilitate socială corporativă în afacerile internaţionale.
International Business Master Course, ASE.
International Institute of Business Analysis, 2015. BABOK. A Guide to the Business Analysis Body
of Knowledge®. v3 ed.
Krol-Mach, M., Olzsak, C.M. and Petech-Pilichowschi, T., 2015. Advances in ICT for Business,
Industry and Public Sector. Studies in Computational Intelligence, Springer.
Miller, F.P., Vandome, A.F. and McBrewster, J., 2011. Pest Analysis. VDM Publishing.
Nica, I. and Chiriță, N., 2020. Analiza și Diagnoza Sistemelor Economice. Seminar/course support,
ASE.
Nica, I., 2020. Bazele Analizei Afacerii. Seminar support, Business Analysis and Enterprise
Performance Control Master, ASE.
Pooley, R. and Wilcox, P., 2003. Applying UML. Advanced Application, 1st edition. Butterworth-
Heinemann.
Von Rosing, M., Kemp, N., Hove, M., Hove, M. and Ross, J.W., 2014. Process Tagging – A Process
Classification and Categorization Concept. Elsevier, Morgan Kaufmann.
Scarlat, E. and Chiriță, N., 2019. Cibernetica sistemelor economice, Third Edition, Economica
Slamanig, M., 2012. PEST Analysis Hungary: Country Evaluation and Selection of Hungary. 1st
ed. GRIN Verlag.
UML Diagrams, 2020. Available online: <https://www.uml-diagrams.org/> [accessed on
06.11.2020].
Unified Modeling Language, 2017. OMG® Unified Modeling Language® (OMG UML®), version
2.5.1. OMG® Unified Modeling Language® Publication.
Wikipedia, The Free Encyclopedia, 2020. Microsoft Visio. Available online:
<https://en.wikipedia.org/wiki/Microsoft_Visio> [accessed on 06.11.2020].
Static and dynamic analysis

of intra-industry trade of BRICS countries
Kuldeep Kumar LOHANI

Visva-Bharati (A Central University), Santiniketan, West-Bengal, India
Indian Council of Social Science Research, New Delhi, India
lohanikuldeep@gmail.com
Abstract. In this article, we tried to estimate Intra-Industry Trade among the BRICS countries. IIT
calculated by employing Grubel and Lloyd IIT Index for static analysis and Thom and McDowell
(1999) MIIT index for dynamic analysis. Additionally, the decomposition of IIT carried out to
distinguish between Horizontal and Vertical IIT. The unit of analysis selected at one- digit and
two-digit SITC Industry level for GL IIT index, Further, to conduct MIIT analysis, industry is
defined at two-digit-SITC level data by aggregating four-digit SITC sub-industry level data of IIT
of BRICS countries. Further, study analysed the Pre and Post-BRICS trade pattern of IIT.
Therefore, this study emphasises that do emerging economies IIT among themselves? On the basis
of estimated results of this study revealed that IIT occurs at higher level of aggregation. This
signifies that developing countries are trading in the same Industry for love for variety and cost
effectiveness. Hence, the empirical result contradicts conventional Krugman (1979, 1985)
hypothesis of IIT trade takes place in developed nations (industrialist nations). This implies that
BRICS countries should focus on opportunities of trade complementary of intermediates products.
This will enhance cost effectiveness of product development or production. Further, this will
promote innovation in the BRICS region. To achieve this, countries needs to conduct constructive
trade dialog among the BRICS countries.
Keywords: intra industry trade, marginal intra industry trade, Grubel and Lloyd index, Thom and
McDowell index, BRICS.
JEL Classification: F1, F11, F15.

108 Kuldeep Kumar Lohani
Introduction
Conventionally, trade takes place between the nations on the basis of comparative
advantage of a nation until 1960 e.g. Adam (1776), Ricardo (1817), Heckscher (1919)
and Ohlin (1933). From 1960s, pattern of trade has changed. This has been nicely
observed by Grubel and Lloyd (1971). In that paper, it is argued that trade takes place
between nations in similar industry. The shadow has been cleared on this finding by
consecutive papers by Krugman (1979, 1980), Helpman (1981) and Krugman and
Helpman (1985) where clearly argued that trade takes place on the basis of product
differentiation, love for variety, economies of scale, and increasing returns. The market
structure is imperfect competition. The structure of trade is intra industry or two way
trade in same industry between the nations. Further, this phenomenon was observed in
developed nation, where factor endowment, income level (i.e. per capita income) and
geographical conditions are similar. Nonetheless, this was not true for the developing
countries and the least developed countries.
In general, IIT performance measured by Grubel and Lloyd (1971) index for a year.
Where, it shows changes in trade flow in each period. Nonetheless, marginal intra-
industry trade (MIIT) index reveals changes in trade flow of IIT over/between the
period(s). In addition, MIIT is related to adjustment cost, which is indirectly related to
trade liberalization and reallocation of resources. The reallocation of resources happen
between industries (IT: inter industry) as well as within industry (IIT). It also covers
structural adjustment cost. In literature, number of studies covered adjustment cost and
employment relationship discussion. However, these issues will not discuss in details in
the present study. Here, the article will focus on IIT and MIIT in reference to BRICS
countries. Additionally, adjustment cost also reveals disequilibrium phenomena; it is
related to structure change in trade flow (exports and imports).
Further, IIT can be distinct into two parts such as horizontal (HIIT) and vertical IIT
(VIIT). HIIT defined as economies of scale where products are differentiated and
consumers choose products for product variety. Conversely, VIIT defined at the industry
level, where production process of goods is vertically integrated, whereas firms are
situated in trading partner’s countries and trade takes place on the basis of specialization
(comparative advantage) as illustrated in inter-industry trade (IT). For instance, a
component is produced in one nation and processed in another and the finished good
could be exported again, either as final good or intermediate goods. Moreover, a
computer processor produced in one nation and exported to nation two, where computer
is assembled and produced final good as computer that can be exported to nation one.
BRICS Acronyms made from Brazil, Russia, India, China and South Africa. Initially,
BRIC term coined by O’Neill (2001), where, a comparison has been made between G7
countries and BRIC countries. This was all about future leading economies in 2050. The
finding of the study suggested that the largest economy in the world would be China
followed by United States of America, India, Japan and Brazil. Consequently, it gave the
voice to forming the BRIC as a group. The term BRIC becomes BRICS after the
inclusion of South Africa in 2010, when 2nd BRICS summit held in Brazil. Overall,
BRICS accounts for 24% of the global GDP in nominal terms and in absolute figure
Static and dynamic analysis of intra-industry trade of BRICS countries 109
stands for $20.3 trillion GDP. The population size is 3.6 billion by the World Bank
database. Further, the IIT among BRICS countries stands for $33.8 billion in 2001. In 2010,
IIT was $414 billion, which increased to $675.3 billion in 2018 as per two-digit-SITC level
data, UNCOMTRADE database, 2020.
Furthermore, in this article, we tried to shed light on intra industry trade (IIT) among
developing nations such as BRICS countries, where, two-way trade analysis has been
carried out for the BRICS nations. The purpose of the study is to trace the evidence on
IIT among the BRICS countries. Further, to reach the purpose of the study, we setup
followings question: whether BRICS economies trade among themselves in the same
industry or not? If so, what level of aggregation of data (industry)? To search the answer
of these basic questions, we applied Grubel and Lloyd (1971) IIT index and marginal
intra-industry trade index as proposed by Thom and McDowell (1999). In addition, we
also distinguished IIT between horizontal and vertical IIT.
The few study tried to analyse IIT with respect to BRICS countries. For example, Proença
and Faustino (2015), Filipowicz (2016), Şahbudak and Şahin (2016), Maxir and Masullo
(2017), Mutambara (2017) and Cattaneoand Snowball (2019), Aditya and Gupta (2019)
tried to analyse IIT by GHM method with gravity model setup. On the hand, Dwesar and
Kesharwani (2019) and Varshini and Manonmani (2019) tried to apply GLI and MIIT
index. Further, Lohani (2020) analysed trade flow of BRICS countries using gravity
model and found trade creation such implies that nearly 20% of exports are above the
normal level in the bloc since inception of BRICS bloc. In addition, Kubendran (2020)
also applied gravity model of trade for BRICS countries. However, did not focus on IIT.
Thus, present study tried to fill the gap and present study prompts.
This organisation of the paper is in followings sequence i.e. Introduction, Review of
Literature, Methodology, Results; Discussion and Conclusion and Policy Implications.
Review of literature
Traditionally, trade takes place on basis of comparative advantage. This was argued by
classical economist i.e. Adam (1776), Ricardo (1817), etc. Thereafter, neoclassical
economist said trade takes place on the basis of factor abundance and specialisation i.e.
Heckscher (1919) and Ohlin (1933). Finally, new trade theorist said trade takes place on
account of product differentiation, love for variety, consumer preference, economies of
scale and increasing returns to scale in the same industry i.e. Krugman (1979, 1980), Eaton
and Kierzkowski (1984), Kierzkowski (1989). To verify this argument, next sub-sections
will discuss recent empirical studies related to IIT. Further, Balassa (1979), Falvey and
Kierzkowski (1987) argued that economies with less trade barriers have larger levels of IIT.
Static intra-industry trade

The first attempt to analyse the IIT has been made by Grubel and Lloyd (1971) analysed
intra-industry trade at three digit SITC level of aggregation of data of nine OECD
member countries. The empirical evidence revealed that IIT taking place below the three
digits SITC level of aggregation between Australia and other OECD countries. IIT
mainly taking place in primary goods or processed manufacture goods. Further, Casson
and Pearce (1988) discussed intra-firm trade and its four factors i.e. propensity to:
circulate, trade, internalise, and the composition of final output that affect intra-firm trade
and related to horizontal product differentiation as well. In addition, multinational
enterprise may exploit opportunities by horizontal differentiation of products. For
instance, Japanese production would have benefited and developing nations as well. This
leads to intermediate products inflowing into intra-firm trade and developing countries
surely going to gain from it. Furthermore, Islam (2018) analysed Bangladesh and India
trade flow by using RCA and GLI for inter-industry trade and IIT respectively at HS two
digit level of disaggregation. The results showed that India has comparative advantage in
more than 10 products whereas Bangladesh has comparative advantage in 8 products
exports. IIT is in mostly in oil/mineral fuels and cotton industry and the magnitude was
very low i.e. 16% and 4% respectively. Boyrie and Kreinin (2011) estimated GLI for IIT
and analysis was concentrated on inter-industry trade of manufacturing products from
five to eight digits SITC level data, and then aggregated in the four digit classification.
Further, the analysis carried out at nation’s total trade and bilateral trade as well. The
estimated results illustrate that IIT is huge among OECD nations, whereas IIT largely
taking place from North to South and South to South region.
Dynamic intra-industry trade

The pioneer attempt has been made by Brulhart (1994), tried to analyse dynamic
approach of GLI, which is considered as static method to analyse IIT. This study
proposed MIIT (marginal IIT) index to examine trade flow including adjustment costs.
Another limitation of GLI was that the dynamic adjustment factors including the sectoral
and geographical distribution benefit could not able to capture. Further, the Irish
chemicals sector was analysed between 1985 and 1990 period. The study concluded that
single indicator is not enough to state all crucial information; hence, a multi-stage
evaluation may the only practicable way out. Further, Thom and McDowell (1999)
estimated marginal intra-industry trade (MIIT) in light of trade liberalisation and
structural adjustment costs. Where, point out that the Briilhart (1994) MIIT index was
able to measure only horizontal IIT, which lead to overestimation of adjustment cost,
thus, study introduced new measure of MIIT such as dynamic MIIT by combining of
Brulhart's index and the aggregate index as suggested in that study. Consequently, VIIT is
the difference of two measures i.e. residual of two indices. This was estimated for the
trade flows between the EU and three Central and Eastern Europe countries. Further, the
study suggested that if we avoid the distinction between vertical IIT and inter industry
trade leads to risk of underestimating the level of IIT and, ceteris paribus. Another study
by Al-Mawali (2006) examined country-specific factors of VIIT and HIIT in South
Africa employing the panel gravity model of trade. Three types of econometric model
have been applied. Conventional assumptions of the gravity model and IIT have been
found true in case of South Africa. However, geographical distance is inversely related to
IIT. Further, Jing (2009) analysed data on HS two digit up to 24 chapters related to
agriculture trade of China. The indices were computed such as GLI, Brülhart marginal IIT
index, and Thom and McDowell marginal IIT index during 1996-2005. Results were
found that agricultural IIT is considerably low, and Brülhart index shown that IIT
improving over the decades, Thom and McDowell index revealed that gain from trade
was mostly coming from growth of VIIT.
Studies on BRICS bloc

Proença and Faustino (2015) examined IIT including HIIT and VIIT with respect to
Portugal and the BRIC countries, the European Union, and the five Portuguese-speaking
African countries. Further, the semi parametric gravity model methods have been applied
to control for unobserved heterogeneity. The sample size consists 38 countries and for the
period of 1995-2006. The outcome of empirical model showed that misspecification
assumption of the parametric model has been corrected by the present approach. IIT has
been adversely affected by distance and the effect of economic integration on IIT found
inconclusive. BRIC economies trade in similar quality of goods. Additionally, IIT is
inversely affected by GDP per capita difference between the trading partners. Further,
Filipowicz (2016) analysed the Russian Federation RTAs with reference to global value
chains (GVCs). This study covered India, China, South Africa, Mexico and Russia
Federation and concluded that favourable RTA must be decided by structure of GVCs of
respective country. Furthermore, Şahbudak and Şahin (2016) analysed IIT in agriculture
products between China and Brazil during 2000-2014. In order to analyse IIT, Grubel-
Lloyd Index (GLI), HIIT and VIIT have been computed at SITC three digit level of
aggregation. The study revealed that IIT in agriculture product is very low and low
quality of VIIT taking place between China and Brazil. Moreover, Maxir and Masullo
(2017) examined Brazil’s trade in forest products for the period of 2000-2014. The RCA
and GL indices have been measured to determine IIT and IT. The results found that
Brazil has comparative advantage (CA) in fuel wood, wood panels, wood floors, wood
articles and wood pulp and imports very low volume of wood and paper. Another study
by Mutambara (2017) examined South Africa’s trade relations with conventional trading
partners and with China as well. The methods were used such as the factor intensities of
goods traded, trade intensity with each country, ease of market access of its products, and
IIT opportunities during 2001-2015. Results infer that China is top trading partner of
South Africa. South Africa’s trade with China is purely complimentary in comparison to
conventional trading partners. China is a market mostly for non-fuel primary
commodities for South Africa, whereas the conventional trade partners gives markets for
both low value and high value added products and further scope for IIT to South Africa.
In addition, South Africa conventional comparative advantages are in low value added
products primary commodities such as non fuel primary commodities and minerals and
mineral fuels. In addition, Cattaneo and Snowball (2019) analysed trade in cultural and
creative industries of South Africa with BRICS countries and results found that South
Africa has trade deficit in cultural good with China and India during 2008-2016. On the
other hand, South Africa trade in cultural goods with Russia shows surplus. Moreover,
crafts and audio-visual sectors are reducing significant trade deficit. Further, study
suggested that trade preferences favoured SADC, the EU and EFTA above BRICS
partners. Further, Dwesar and Kesharwani (2019) examined the magnitude of IIT between
India and China using Brülhart weighted marginal IIT A-Index during 1999-2018. Further,
the study covered data on HS two digit i.e. 99 commodities and compared decade-wise
IIT performance. The evidence was found that India exports to China raw-materials and
some semi-finished goods. While China’s exports to India primary products and
secondary manufactured goods. Percentage share of IIT in total trade are ranging from
10% to 20% over the years. Lastly, Varshini and Manonmani (2019) analysed trade
advantage of Indian pharmaceutical industry 2000-2014. The methods were used i.e.
Marginal Intra-Industry Trade, Horizontal and Vertical Intra-industry, Balassa Revealed
Comparative Advantage (RCA), Revealed Symmetric Comparative Advantage,
Normalized RCA, and Vollrath’s Trade Advantage indices were estimated. The empirical
results revealed that intra-industry trade taking place at the higher level. However,
comparative disadvantage has been observed for this industry. In addition, India exports
high quality of pharmaceutical products to the world.
Other methods
Caporale et al. (2015) analysed Chinese trade flows and trade specialisation using gravity
model of trade from 1992-2012. The analysis consists major trading partners from Asia,
Europe and North America and fixed effect vector decomposition model has been used to
estimate gravity model. The empirical results revealed that there is a change in structure
of foreign trade of China specifically to paradigm shift from resource – and labour-
intensive to capital – and technology-intensive exports has been observed. Further, In
addition, Konno (2016) estimated inter industry trade and IIT including HIIT and VIIT
for Russian trade flow with major trading partners. Thereafter, determinants of IIT have
been analysed by using log- linear gravity model of trade for the period of 2000-2013.
The feasible generalised lest square, fixed effect and random effect econometric model
have been estimated. The results were found that conventional assumptions of gravity
model are valid for Russia. In addition, FTA, FDI and Commonwealth Independent
States/Customs Union membership positively influencing trade of Russia. Nonetheless,
distance is adversely affecting trade flow. Furthermore, Marzábal et al. (2016) analysed
performance of IIT between EU and Latin America with reference to Brazil and Mexico
and studied the factors affecting trade and foreign direct investment. Study also focused
on input-output framework. In addition, Mhaka and Jeke (2018) analysed trade
performance between South Africa and China during 1995-2014. The OLS method has
been used to analyse bilateral trade flow, economic size, population or market size, and
exchange rate of respective countries. The result showed that economic size positively
affecting trade whereas exchange rate is inversely related to trade flow. Aditya and Gupta
(2019) examined India’s IIT specifically to decomposition of IIT into horizontal and
vertical trade. It analysed by using Greenaway-Hine-Milner and support vector machines
method. The results of both the method have been compared and contrasted during
1978-2013 at SITC 5 digit level of data aggregation. The results showed that India’s trade
is horizontal IIT. However, vertical IIT are increasing in comparison to horizontal IIT
over the period. In addition, the industry wise analysis namely bakery industry,
phosphorus industry, aluminium industry results revealed that is horizontal IIT. Chin et
al. (2019) computed IIT decomposing into HIIT and VIIT and analysed VIIT and
economic size for the Malaysian economy for the period of 1988-2016. The study
covered Malaysia’s trade with its top trading partners and panel vector autoregression
method used for data analysis. The results revealed that positive bidirectional causality
relationship exist between VIIT and economic size and trade-led growth hypothesis is
valid in case of Malaysia. Lastly, Brodzicki et al. (2020) examined factors affecting VIIT
and HIIT of Spain and Poland employing gravity model of trade estimated by fixed effect
and PPML models from 2005-2014. The factors were included such as convention factors
of standard gravity model of trade including various indicators of trade cost, regional
factors and FDI. The study formulated hypothesis based on economic theory and most of
them found true and some them were false.
The purpose of this study is to know the BRICS bloc formation effect on IIT of BRICS
countries. However, after reviewing the literature, we did not find ample of evidence on
IIT and MIIT analysis specific to IIT of BRICS countries. Very few study tried to analyse
IIT with respect to BRICS countries. For example, Proença and Faustino (2015),
Filipowicz (2016), Şahbudak and Şahin (2016), Maxir and Masullo (2017), Mutambara
(2017) and Cattaneoand Snowball (2019) tried to analyse IIT with gravity model setup
and GHM method. On the other hand, Aditya and Gupta (2019), Dwesar and Kesharwani
(2019) and Varshini and Manonmani (2019) tried to apply GLI and MIIT index. Thus,
present study tried to fill the gap and present study prompts. The next section will
elaborate on methodology of this study.
Methodology
The Intra Industry Trade is measured by following methods: The pioneer attempt made
by Grubel and Llyod (1971) to measure IIT. Thus, it usually known as Grubel and Llyod
Index (hereinafter referred to as GLI). The GLI is calculated by:
| |
GLIx = 1- (1)
Where, E stands for export of x country to y country in the similar industry and M stands
for import of x country from y country in the similar industry. Subscript x and y is country
name. The index values remain within zero and one or 0 GLI 1. The index explains that
if GLI =1, it implies that country’s IIT is very high with respect to trading partner.
Conversely, If GLI = 0, this implies that country’s trade in the inter industry only. The other
classification of values of GLI is following: if GLI .25, it infers that IIT is low, if .25
GLI .5 it implies that IIT is in lower level; further, when, GLI values lies between .75
and 1 or .75 GLI 1, this reveals that IIT is high in the trade flow.
Marginal Intra Industry Trade index (MIITI): This is measured by Brulhart (1994). The
MIITI uses to analyse the dynamic aspects of IIT. Whereas, GLI was unable to captured
and considered as static analysis. Nonetheless, Brulhart (1994) was able to calculate only
HIIT or IIT in similar products in the same industry. Therefore, this study uses Thomand
and McDowell (1999) approach because it is more comprehensive. Additionally, this
approach uses aggregate of all sub-IIT and present IIT at industry level. Further, also,
able to differentiate IIT between HIIT and VIIT at industry level. The definition of VIIT
and HIIT defers from GHM approach. In this case, trade in different sub sector in the
same industry or sector called as VIIT e.g. trade in computer and computer parts or
assembly and trade in similar product of the same sector or industry defined as HIIT.
Furthermore, inter industry is calculated as residual form total trade. The equation (3)
illustrate Briilhart (1994) unweighted MIITI for ‘s’ industry. To measure MIIT at sub
sector, equation (4) formed as suggested by Briilhart (1994), where Aw denotes weighted
MIITI for ‘s’ industry; W denotes a weight that is calculated by
|∆ ∆ |
As = 1- |∆ | |∆ |
(3)
As value lies between zero and one, where, As = 0 implies that pure inter-industry trade
(IT) and As =1 refers to high degree of IIT. If the values of As remain between 0 and .5
or 0 MIITI .5, this implies that MIIT at high level.
Aw = ∑ 𝑊𝐴 (4)
The equation (4) is weighted MIITI and calculated at sub-industry level and aggregate
sub industry’s calculated value by multiplying weighted value to get equation (4) give
aggregate MIITI at industry level.
|∆ | |∆ |
Ws = ∑ |∆ | |∆ |
(5)
To estimate Thomand and McDowell (1999) aggregate approach, the equation (6) would
be estimated. The equation (6) also written as equation (7). Furthermore, to decompose
total trade flow, the differentiation of IIT have been done by computing Aw, it reveal IIT
then computing (As - Aw) denotes VIIT and finally, to get IT (1- As).
|∆ ∆ |
As = 1- ∑ |∆ | ∑ |∆ |
(6)
|∆ ∆ |
Ai = 1- |∆ | |∆ |
(7)
Eq. 7 is weight index suggested by Briilhart (1994) for As. MIITI measures the trade
direction and structure of changes between distinct time periods. In other words, it
measures the gap between change in total export and import and change in net trade. This
also captures the adjustment process of IIT. This was not possible to trace by GLI
method. Further, When, As reveals positive value and closer to one, implies that sectoral
trade is converging trend and shows predominance of MIIT in the adjustment process.
Nonetheless, if As shows negative value, this implies that divergence trend in sectoral
trade flows, while other thing remaining constant, leads to greater transitional adjustment
costs. The next section will describe results of the study.
Empirical results
Static IIT analysis
The GLI results of BRICS countries are presented in Tables 1-6 on one digit-SITC level
data and Tables 7-9, reports GLI results for the year 2001, 2010 and 2018 at two-digit
SITC level data. The year 2001 depicts pre-BRICS scenario, year 2010 shows at the time
of BRICS formation level of IIT prevailing among the BRICS countries and year 2018
reveals post-BRICS IIT performance. Further, Brazil’s IIT with BRICS countries are in
followings sectors: IIT with China is at low level only in [6] manufacture goods over the
period; IIT with India is at low level in [6] manufacture goods and [7] Machinery and
transport equipment; IIT with Russia is low in [2] Crude materials, inedible, except fuels,
and [8] miscellaneous manufactured articles; and IIT with South Africa is high in [1]
Beverages and tobacco, [2] Crude materials, inedible, except fuels, [3] Mineral fuels,
lubricants and related materials, [5] Chemicals and related products, n.e.s., and [6]
Manufactured goods. Overall, Brazil’s IIT takes place with BRICS countries mostly in
[6] manufacture goods and [3] Crude materials, inedible, except fuels (Table 1).
Table 1. Brazil’s IIT with BRICS Countries at SITC-One Digit Data Level
BRAZIL CHINA INDIA RUSSIA SOUTH AFRICA
PRODUCT 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018
ALL 0.82 0.91 0.70 0.69 0.90 0.97 0.59 0.63 0.66 0.81 0.73 0.65
0 0.50 0.64 0.21 0.23 0.03 0.12 0.03 0.00 0.04 0.07 0.02 0.02
1 0.00 0.00 0.04 NA 0.09 0.20 0.01 0.01 0.03 0.24 0.25 0.52
2 0.02 0.01 0.01 0.14 0.16 0.18 0.15 0.74 0.47 0.60 0.64 0.93
3 0.48 0.10 0.04 0.00 0.80 0.17 0.00 NA 0.00 1.00 0.12 0.94
4 0.03 0.00 0.02 0.00 0.29 0.08 NA NA NA 0.00 0.01 0.06
5 0.25 0.25 0.20 0.21 0.34 0.29 0.04 0.04 0.03 0.75 0.69 0.60
6 0.85 0.43 0.51 0.72 0.54 0.34 0.14 0.12 0.23 0.48 0.63 0.85
7 0.74 0.10 0.04 0.86 0.55 0.55 0.78 0.45 0.12 0.13 0.32 0.09
8 0.06 0.02 0.03 0.83 0.26 0.26 0.09 0.95 0.25 0.18 0.27 0.32
9 NA 0.00 0.00 NA NA NA NA NA 0.00 NA 0.75 NA
Source: Author own calculation based on UNCTAD database accessed on 02/02/2020.
Table 2. China’s IIT with BRICS Countries at SITC-One Digit Data Level
CHINA BRAZIL INDIA RUSSIA SOUTH AFRICA
PRODUCT 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018
ALL 0.73 0.78 0.58 0.95 0.68 0.42 0.51 0.93 0.86 0.94 0.84 0.71
0 0.55 0.55 0.29 0.42 0.70 0.92 0.52 1.00 0.97 0.81 0.59 0.87
1 0.00 0.01 0.03 0.17 0.12 0.50 0.01 0.17 0.63 0.06 0.92 0.43
2 0.01 0.01 0.01 0.51 0.07 0.27 0.08 0.05 0.05 0.10 0.03 0.02
3 0.00 0.14 0.05 0.25 0.87 0.53 0.12 0.03 0.01 0.53 0.27 0.90
4 0.12 0.00 0.02 0.22 0.03 0.03 0.57 NA 0.04 0.46 0.30 0.33
5 0.56 0.37 0.28 0.80 0.30 0.36 0.14 0.71 0.82 0.99 0.54 0.39
6 0.86 0.48 0.74 0.85 0.54 0.80 0.26 0.73 0.80 0.86 0.71 0.61
7 0.63 0.11 0.22 0.28 0.08 0.09 0.22 0.06 0.09 0.38 0.05 0.02
8 0.11 0.02 0.02 0.51 0.18 0.14 0.17 0.01 0.03 0.02 0.02 0.03
9 NA 0.01 0.00 0.16 0.46 NA 0.00 0.65 NA 0.01 0.00 NA
Table 3. India’s IIT with BRICS Countries at SITC-One Digit Data Level
INDIA BRAZIL CHINA RUSSIA SOUTH AFRICA
PRODUCT 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018
ALL 0.80 0.98 0.89 0.57 0.58 0.35 0.98 0.46 0.47 0.46 0.82 0.82
0 0.42 0.03 0.10 0.51 0.65 0.44 0.01 0.09 0.25 0.22 0.14 0.45
1 0.02 0.17 0.15 0.12 0.29 0.82 0.00 0.00 0.04 0.30 0.03 0.43
2 0.13 0.14 0.15 0.95 0.11 0.50 0.32 0.23 0.43 0.21 0.12 0.05
3 0.05 0.86 0.19 0.01 0.79 0.56 0.01 0.01 0.01 0.00 0.78 0.34
4 0.02 0.21 0.06 0.17 0.05 0.09 0.00 0.35 0.81 0.51 0.37 0.02
5 0.33 0.45 0.36 0.63 0.30 0.44 0.98 0.49 0.95 0.42 0.66 0.33
6 0.89 0.61 0.42 0.56 0.79 0.42 0.48 0.20 0.22 0.88 0.70 0.99
7 0.62 0.63 0.50 0.12 0.05 0.07 0.28 0.89 0.78 0.90 0.09 0.20
8 0.92 0.28 0.26 0.38 0.13 0.15 0.36 0.28 0.81 0.11 0.06 0.17
9 0.60 0.10 0.00 0.91 0.02 0.01 0.07 0.12 0.00 0.02 0.01 0.00
China’s IIT with BRICS countries are in sectors such as IIT with Brazil is at high level
only in [6] manufacture goods and in the low level in [0] Food and live animals and [5]
Chemicals and related products, n.e.s. over the period; IIT with India is in high level in
[6] manufacture goods and at low level in [0] Food and live animals[5] Chemicals
and related products, n.e.s.; IIT with Russia is high level in [0] Food and live animals,
[6] manufacture goods and [5] Chemicals and related products, n.e.s. and at low level in
[1] Beverages and tobacco over the period; and IIT with South Africa is high in [0] Food
and live animals and [3] Mineral fuels, lubricants and related materials, and at low level in
[1] Beverages and tobacco, [4] Animal and vegetable oils, fats and waxes, [5] Chemicals
and related products, n.e.s., and [6] Manufactured goods. Overall, China IIT does with
BRICS countries mostly in [6] manufacture goods and [5] Chemicals and related
products, n.e.s. (Table 2).
Further, India’s IIT with BRICS countries are in sectors like: IIT with Brazil is at low level
only in [7] Machinery and transport equipment and at the lowest level in [5] Chemicals and
related products, n.e.s., [6] manufacture goods and [8] Miscellaneous manufactured articles
over the period; IIT with China is at high level in [1] Beverages and tobacco and at low
level [2] Crude materials, inedible, except fuels, [3] Mineral fuels, lubricants and related
materials, [5] Chemicals and related products, n.e.s., and [6] Manufactured goods; IIT with
Russia is at high level in [4] Animal and vegetable oils, fats and waxes, [5] Chemicals and
related products, n.e.s., [7] Machinery and transport equipment, and [8] Miscellaneous
manufactured articles, and at the lowest level in [0] Food and live animals, and [2] Crude
materials, inedible, except fuels, over the years and IIT with South Africa is high in [6]
Manufactured goods, and at lowest level in [0] Food and live animals, [1] Beverages and
tobacco, [3] Mineral fuels, lubricants and related materials, and [5] Chemicals and related
products, n.e.s.. Overall, India doses IIT with BRICS countries mostly in [6] manufacture
goods and [5] Chemicals and related products, n.e.s., [7] Machinery and transport
equipment and [3] Mineral fuels, lubricants and related materials (Table 3).
Table 4. Russia’s IIT with BRICS Countries at SITC-One Digit Data Level
RUSSIA BRAZIL CHINA INDIA SOUTH AFRICA
PRODUCT 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018
ALL 0.29 0.58 0.88 0.62 0.74 0.96 0.83 0.48 0.52 0.24 0.23 0.67
0 NA 0.00 0.02 0.69 0.80 0.93 0.01 0.03 0.16 0.00 NA 0.79
1 0.00 0.01 0.01 0.01 0.10 0.95 0.00 0.00 0.01 NA NA 0.00
2 0.63 0.27 0.35 0.10 0.07 0.08 0.94 0.41 0.57 0.56 0.11 0.21
RUSSIA BRAZIL CHINA INDIA SOUTH AFRICA

PRODUCT 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018
3 0.00 0.00 0.00 0.22 0.04 0.01 0.02 0.02 0.01 NA 0.25 0.04
4 NA NA NA 0.28 0.01 0.03 0.00 0.47 0.78 NA 0.01 NA
5 0.10 0.04 0.04 0.22 0.90 0.52 0.88 0.58 0.94 0.88 0.94 0.67
6 0.40 0.26 0.72 0.51 0.24 0.36 0.50 0.29 0.32 0.65 0.37 0.93
7 0.97 0.89 0.15 0.69 0.14 0.14 0.17 0.67 0.61 0.39 0.15 0.46
8 0.10 0.52 0.27 0.23 0.02 0.03 0.64 0.30 0.78 0.24 0.21 0.84
9 NA 0.63 0.60 0.00 0.16 0.00 0.03 0.07 0.00 0.95 0.00 0.00
Russia’s IIT with BRICS countries are in followings sectors: IIT with Brazil is high only
in [6] manufacture goods and the lowest in [2] Crude materials, inedible, except fuels,
and [8] Miscellaneous manufactured articles over the period; IIT with China is at high
level in [0] Food and live animals, and [1] Beverages and tobacco and low level in [5]
Chemicals and related products, n.e.s., and [6] Manufactured goods; IIT with India is
high level in [4] Animal and vegetable oils, fats and waxes, [5] Chemicals and related
products, n.e.s., and [8] Miscellaneous manufactured articles, and the low level in [7].
Machinery and transport equipment, [2] Crude materials, inedible, except fuels, and [6]
Manufactured goods, over the years and IIT with South Africa is very high in [0] Food
and live animals, [6] Manufactured goods, and [8] Miscellaneous manufactured articles
and the low level in [5] Chemicals and related products, n.e.s. and [7] Machinery and
transport equipment. Overall, Russia does IIT with BRICS countries mostly in [6]
manufacture goods and [5] Chemicals and related products, n.e.s. [8], Miscellaneous
manufactured articles (Table 4).
Table 5. South Africa’s IIT with BRICS Countries at SITC-One Digit Data Level
SOUTH AFRICA BRAZIL CHINA INDIA RUSSIA
PRODUCT 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018
ALL 0.56 0.68 0.48 0.60 0.83 0.67 0.80 0.97 0.93 0.85 0.55 0.88
0 0.05 0.03 0.02 0.48 0.50 0.74 0.36 0.16 0.36 0.12 0.00 0.98
1 0.49 0.30 0.39 0.14 0.91 0.09 0.23 0.05 0.75 0.05 0.01 0.01
2 0.91 0.62 0.68 0.17 0.04 0.04 0.25 0.10 0.06 0.74 0.16 0.31
3 0.01 0.20 0.81 0.27 0.37 0.15 0.22 0.63 0.30 0.13 0.22 0.01
4 NA NA 0.02 0.00 0.35 0.23 0.03 0.00 0.02 NA 0.01 NA
5 0.79 0.71 0.84 0.54 0.51 0.17 0.49 0.81 0.36 0.04 0.14 0.16
6 0.90 0.79 0.79 0.88 0.86 0.63 0.98 0.78 0.77 0.42 0.72 0.38
7 0.30 0.54 0.17 0.31 0.04 0.01 0.89 0.21 0.24 0.47 0.47 0.99
8 0.48 0.43 0.32 0.01 0.01 0.07 0.13 0.11 0.12 0.36 0.97 0.92
9 0.00 0.00 0.00 0.04 0.00 0.00 0.57 0.01 0.00 0.88 NA 0.02
South Africa does IIT with BRICS countries in the sectors like: IIT with Brazil is high in
[3] Mineral fuels, lubricants and related materials, [5] Chemicals and related products,
n.e.s., and [6] Manufactured goods, and the lowest in [1] Beverages and tobacco, [2] Crude
materials, inedible, except fuels, and [8] Miscellaneous manufactured articles over the
period; IIT with China is high in [0] Food and live animals, and [6] Manufactured goods;
IIT with India is high level in [6] Manufactured goods, and [1] Beverages and tobacco
and the low level in [0] Food and live animals, [3] Mineral fuels, lubricants and related
materials, and [5] Chemicals and related products, n.e.s., over the years and IIT with
Russia is very high in [0] Food and live animals, [7] Machinery and transport equipment,
and [8] Miscellaneous manufactured articles and the low level in [6] Manufactured goods
and [2] Crude materials, inedible, except fuels,. Overall, South Africa’s IIT does with BRICS
countries largely in [0] Food and live animals, [6] manufacture goods and [5] Chemicals
and related products, n.e.s., [2] Crude materials, inedible, except fuels, [3] Mineral fuels,
lubricants and related materials, [8] Miscellaneous manufactured articles (Table 5).
Table 6. Overall BRICS Countries’ IIT with BRICS Countries at SITC-One Digit Data Level
BRICS BRAZIL CHINA INDIA RUSSIA SOUTH AFRICA
PRODUCT 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018 2001 2010 2018
ALL 0.84 0.91 0.82 0.48 0.68 0.94 0.67 0.63 0.43 0.79 0.80 0.92 0.71 0.86 0.75
0 0.06 0.14 0.17 0.53 0.84 0.69 0.24 0.82 0.82 0.13 0.29 0.76 0.57 0.57 0.76
1 0.05 0.17 0.10 0.38 0.18 0.19 0.04 0.13 0.29 0.00 0.02 0.33 0.96 0.45 0.63
2 0.04 0.03 0.03 0.15 0.03 0.03 0.84 0.31 0.97 0.16 0.20 0.31 0.24 0.08 0.08
3 0.29 0.64 0.13 0.73 0.11 0.07 0.02 0.90 0.68 0.22 0.04 0.01 0.57 0.57 0.47
4 0.00 0.03 0.07 0.20 0.01 0.03 0.21 0.57 0.87 0.00 0.47 0.10 0.00 0.22 0.05
5 0.21 0.25 0.20 0.60 0.77 0.42 0.78 0.46 0.62 0.38 0.70 0.94 0.94 0.71 0.31
6 0.73 0.43 0.45 0.33 0.93 0.91 0.68 0.79 0.49 0.51 0.52 0.60 0.96 0.86 0.65
7 0.91 0.18 0.11 0.31 0.71 0.11 0.28 0.25 0.18 0.54 0.26 0.24 0.35 0.10 0.05
8 0.18 0.06 0.05 0.47 0.80 0.06 0.67 0.29 0.31 0.31 0.12 0.09 0.04 0.02 0.08
9 NA 0.02 0.00 0.01 0.03 0.00 0.05 0.03 0.00 0.01 0.12 0.00 0.09 0.00 0.00
BRICS Countries’ IIT with BRICS countries are in followings sectors: IIT with Brazil is
low in [6] Manufactured goods, over the period; IIT with China is high in [6] Manufactured
goods and the low level in [0] Food and live animals, and [5] Chemicals and related
products, n.e.s.; IIT with India is high level in [0] Food and live animals, [2] Crude
materials, inedible, except fuels and [4] Animal and vegetable oils, fats and waxes, and
the low level in [5] Chemicals and related products, n.e.s., [6] Manufactured goods, and
[3] Mineral fuels, lubricants and related materials, In addition, the lowest level IIT is in
[1] Beverages and tobacco and [8] Miscellaneous manufactured articles, over the years;
IIT with Russia is high in [0] Food and live animals, and [5] Chemicals and related
products, n.e.s., the low level IIT is in [6] Manufactured goods and the lowest level of IIT
is in [1] Beverages and tobacco and[2] Crude materials, inedible, except fuels; and IIT with
South Africa is very high in [0] Food and live animals, the low level IIT is in [1] Beverages
and tobacco, [6] manufacture goods and the lowest level of IIT is in [5] Chemicals and
related products, n.e.s., and [3] Mineral fuels, lubricants and related materials. Overall, BRICS
IIT with BRICS countries take place mostly in [0] Food and live animals, [6] manufacture
goods and [5] Chemicals and related products, n.e.s., [2] Crude materials, inedible, except
fuels, [3] Mineral fuels, lubricants and related materials (Table 6).
Table 7. Overall BRICS Countries’ IIT with BRICS Countries at SITC-Two Digit Data Level in 2001
Country Name Partner Country Name 0-0.25 0.25-0.5 0.5-0.75 0.75-1 Total No. of Industry
China 28 13 5 3 49
Brazil India 13 6 6 9 34
Russian Federation 8 3 2 13
South Africa 14 9 7 6 36
Brazil 28 10 6 3 47
China India 21 8 15 5 49
Russian Federation 22 9 3 7 41
Brazil 11 13 4 6 34
India China 26 7 7 11 51
Brazil 9 2 3 2 16
Russian Federation China 24 11 5 10 50
India 16 9 4 1 30
Brazil 15 10 6 8 39
South Africa China 33 2 5 8 48
India 23 10 7 7 47
Source: Author own calculation based on UNCOMTRADE database using WITS accessed on 02/04/2020.
The results on two digit-SITC data have been reported in Tables 7-9. It shows distribution
of indices’ value of IIT of BRICS countries for the years such as 2001, 2010 and 2018.
Further, the three years represents pre-BRICS 2001, at the BRICS formation-2010 and
post-BRICS formation-2018. Furthermore, the analysis revealed that IIT at two industry
level expanded over the period. This is clearly shown by three different year analysis (see
Tables 7-9). However, the high level of IIT performance has observed in the nearly in
single digit over the years of the analysis. Further, the results and break down high IIT
indices have been reported and discussed in next paragraph. For instance, in 2001,
Brazil’s IIT with BRICS countries at two-digit SITC chapters are high in followings
industries: IIT with China in [9] Misc food products (0.81), [66] Non-metal mineral
manuf. (0.95), and [72] Industry special machine (0.86); IIT with India in [59] Chem
material/prods n.e.s. (0.95), [66] Non-metal mineral manuf. (0.85), [67] Iron and steel
(0.93), [72] Industry special machine (1.00), [73] Metalworking machinery (0.75), [74]
Industrial equipment n.e.s. (0.90), [81] Building fixtures etc. (0.80), [82] Furniture/
furnishings (0.98), and [89] Misc manufactures n.e.s. (0.85); and IIT with South Africa in
[29] Crude anim/veg mater n.e.s. (0.97),[54] Pharmaceutical products (0.77), [64] Paper/
paperboard/article (0.95), [69] Metal manufactures n.e.s. (0.79), [73] Metalworking
machinery (0.90), and [76] Telecomms etc. equipment (0.85).
Table 8. Overall BRICS Countries’ IIT with BRICS Countries at SITC- Two Digit Data Level in 2010
China 38 7 5 3 53
Brazil 35 8 5 3 51
Brazil 24 7 6 5 42
Brazil 10 3 7 6 26
India 22 8 5 4 39
Brazil 27 7 6 3 43
India 31 7 9 4 51
In addition, in 2001, China’s IIT with BRICS countries at two-digit SITC chapters are
high in followings industries: IIT with Brazil in [5] Vegetables and fruit (0.99), [26] Textile
fibres (0.92), and [68] Non-ferrous metals (0.88); IIT with India in [9] Misc food
products (0.76), [27] Crude fertilizer/mineral (0.85), [29] Crude anim/veg mater n.e.s.
(0.94), [33] Petroleum and products (0.90), and [65] Textile yarn/fabric/art. (0.88); IIT
with Russian Federation in [26] Textile fibres (0.83), [52] Inorganic chemicals (0.83),
[61] Leather manufactures (0.77), [63] Cork/wood manufactures (0.82), [77] Electrical
equipment (0.87), [78] Road vehicles (0.89), and [89] Misc manufactures n.e.s. (0.99); and
IIT with South Africa in [3] Fish/shellfish/etc. (0.85), [11] Beverages (0.91), [33] Petroleum
and products (0.96), [53] Dyeing/tanning/color mat (0.95), [63] Cork/wood manufactures
(0.79), [71] Power generating equipment (0.98), and [74] Industrial equipment n.e.s.
(0.90).
Table 9.Overall BRICS Countries’ IIT with BRICS Countries at SITC- Two Digit Data Level in 2018
China 37 9 3 7 56
Brazil 38 11 3 5 57
Brazil 19 12 13 7 51
Brazil 18 9 4 5 36
India 31 6 10 4 51
Brazil 24 12 7 8 51
India 30 9 11 5 55
Further, in 2001, India’s IIT with BRICS countries at two-digit SITC chapters are high in
industries like: IIT with Brazil in [5] Vegetables and fruit (0.92), [67] Iron and steel
(0.76), [74] Industrial equipment n.e.s. (0.81), [75] Office/dat proc machines (0.82),
[82] Furniture/furnishings (0.81), and [85] Footwear (0.94); IIT with China in [4] Cereals/
cereal preparatn (0.77), [8] Animal feed ex unmlcer. (0.99), [9] Misc food products
(0.78), [27] Crude fertilizer/mineral (0.82), [29] Crude anim/veg mater n.e.s. (0.79), [55]
Perfume/cosmetic/cleansr (0.96), [59] Chem material/prods n.e.s. (0.98), [62] Rubber
manufactures n.e.s. (0.80), [65] Textile yarn/fabric/art. (0.95), [67] Iron and steel (0.77),
and [93] UN Special Code (0.85); IIT with Russian Federation is in [51] Organic
chemicals (0.98), [59] Chem material/prods n.e.s. (0.86), [66] Non-metal mineral manuf.
(0.80), [76] Telecommsetc equipment (0.84), [87] Scientific/etc instrument (0.92), [89]
Misc manufactures n.e.s. (0.81), and [93] UN Special Code (0.76); and IIT with South
Africa is in [57] Plastics in primary form (0.89), [63] Cork/wood manufactures (0.87),
[67] Iron and steel (0.99), [71] Power generating equipment (0.98), and [74] Industrial
equipment n.e.s. (0.92). Furthermore, in 2001, Russia Federation’s IIT with BRICS
countries at two-digit SITC chapters are high in followings industries: IIT with Brazil in
[64] Paper/paperboard/article (0.76), and [65] Textile yarn/fabric/art. (0.77); with China
in [0] Live animals except fish (0.88), [26] Textile fibres (0.86), [27] Crude fertilizer/
mineral (0.84), [52] Inorganic chemicals (0.95), [59] Chem material/prods n.e.s. (0.96),
[64] Paper/paperboard/article (0.87), [69] Metal manufactures n.e.s. (0.90), [72] Industry
special machine (0.94), [76] Telecomms etc. equipment (0.77), and [77] Electrical
equipment (0.80); IIT with India only in [62] Rubber manufactures n.e.s. (0.80); and with
South Africa only in [51] Organic chemicals (0.87). Lastly, in 2001, South Africa’s IIT
with BRICS countries at two-digit SITC chapters are high such as IIT with Brazil in [27]
Crude fertilizer/mineral (0.88), [51] Organic chemicals (0.87), [67] Iron and steel (0.98),
[69] Metal manufactures n.e.s. (0.83), [73] Metalworking machinery (0.88), [84]
Apparel/clothing/access (0.86), [87] Scientific/etc. instrument (0.90), and [88]
Photographic equ/clocks (0.85); with China in [11] Beverages (0.93), [27] Crude
fertilizer/mineral (0.86), [51] Organic chemicals (0.91), [53] Dyeing/tanning/color mat
(0.91), [64] Paper/paperboard/article (0.75), [71] Power generating equipment (0.99), [73]
Metalworking machinery (0.99), and [74] Industrial equipment n.e.s. (0.97); IIT with
India in [33] Petroleum and products (0.76), [51] Organic chemicals (0.91), [59] Chem
material/prods n.e.s. (0.94), [71] Power generating equipment (0.77), [74] Industrial
equipment n.e.s. (0.88), [76] Telecomms etc equipment (0.85), and [77] Electrical
equipment (0.90); and IIT with Russian Federation is in [75] Office/dat proc machines
(0.83), and [77] Electrical equipment (0.77).
Further, in 2010, Brazil’s IIT with BRICS countries at two-digit SITC chapters are high in
followings industries: IIT with China is in [6] Sugar/sugar prep/honey (0.81), [7] Coffee/
tea/cocoa/spices(0.78), and [66] Non-metal mineral manuf.(0.87); IIT with India in [23]
Crude/synthet/rec rubber (0.98), [68] Non-ferrous metals (0.99), [71] Power generating
equipment (0.90), [88] Photographic equ/clocks (0.90), and [89] Misc manufactures n.e.s.
(0.96); IIT with Russian Federation is in [25] Pulp and waste paper (0.97), [68] Non-ferrous
metals (0.75), [74] Industrial equipment n.e.s. (0.79), [75] Office/dat proc machines (0.79),
[78] Road vehicles (0.80), and [89] Misc manufactures n.e.s. (0.94); and IIT with South
Africa is in [53] Dyeing/tanning/color mat (0.88), [59] Chem material/prods n.e.s. (0.94),
and [62] Rubber manufactures n.e.s. (0.94). In addition, in 2010, China’s IIT with BRICS
countries at two-digit SITC chapters are high in industry like: IIT with Brazil is in [54]
Pharmaceutical products (0.85), [57] Plastics in primary form (0.98) and [61] Leather
manufactures (0.99); IIT with India is in [7] Coffee/tea/cocoa/spices (0.95), [22] Oil
seeds/oil fruits (0.92), [23] Crude/synthet/rec rubber (0.91), [41] Animal oil/fat (0.82), [51]
Organic chemicals (0.81), [54] Pharmaceutical products (0.96), [57] Plastics in primary
form (0.80), and [67] Iron and steel (0.94); IIT with Russian Federation is in [8] Animal
feed ex unmlcer. (0.78), [57] Plastics in primary form (0.88), [67] Iron and steel (0.98), and
[72] Industry special machine (1.00); and IIT with South Africa is in [3] Fish/shellfish/etc.
(0.91), [11] Beverages (0.92), [27] Crude fertilizer/mineral (0.92), [51] Organic chemicals
(0.96), [52] Inorganic chemicals (0.99), [57] Plastics in primary form (0.89), [68] Non-
ferrous metals (0.98), [78] Road vehicles (0.94), and [93] UN Special Cod (1.00). Further,
in 2010, India’s IIT with BRICS countries at two-digit SITC chapters are high in industries
like: IIT with Brazil in [7] Coffee/tea/cocoa/spices (0.89), [57] Plastics in primary form
(0.82), [61] Leather manufactures (0.76), and [71] Power generating equipment (0.85); IIT
with China in [3] Fish/shellfish/etc. (0.84), [6] Sugar/sugar prep/honey (0.96), [28] Metal
ores/metal scrap (0.81), [32] Coal/coke/briquettes (0.78), [59] Chem. material/prods n.e.s.
(0.76), [61] Leather manufactures (0.84), and [88] Photographic equ/clocks (0.82); IIT with
Russian Federation is in [26] Textile fibres (0.97), [28] Metal ores/metal scrap (0.80), [57]
Plastics in primary form (0.99), and [77] Electrical equipment (0.79) and IIT with South
Africa is in [29] Crude anim/veg mater n.e.s. (0.96), [66] Non-metal mineral manuf. (0.88),
[68] Non-ferrous metals (0.90), and [87] Scientific/etc instrument (0.97). Furthermore, in
2010, Russia’s IIT with BRICS countries at two-digit SITC chapters are high in followings
industries: IIT with Brazil in [58] Plastics non-primry form (0.98), and [65] Textile yarn/
fabric/art. (0.91); IIT with China in [33] Petroleum and products (0.94), and [51] Organic
chemicals (0.79); IIT with India in [9] Misc food products (0.95), [23] Crude/synthet/rec
rubber (0.97), [64] Paper/paperboard/article (0.76), [71] Power generating equipment
(0.92), and [75] Office/dat proc machines (0.90); and IIT with South Africa is in [76]
Telecomms etc equipment (0.90), [78] Road vehicles (0.98), and [87] Scientific/etc
instrument (0.95). Finally, in 2010, South Africa’s IIT with BRICS countries at two-digit
SITC chapters are high industries such as: IIt with Brazil in [51] Organic chemicals (0.95),
[52] Inorganic chemicals (0.98), and [65] Textile yarn/fabric/art. (0.85); IIT with China in
[9] Misc food products (0.91), [42] Fixed veg oils/fats (0.81), [65] Textile yarn/fabric/art.
(0.99), [73] Metalworking machinery (0.84), and [88] Photographic equ/clocks (0.85); IIT
with India in [9] Misc food products (0.89), [26] Textile fibres (0.95), [52] Inorganic
chemicals (0.86), [55] Perfume/cosmetic/cleansr (0.95), [71] Power generating equipment
(0.86), and [75] Office/dat proc machines (0.93); IIT with Russian Federation in [52]
Inorganic chemicals (0.89), [64] Paper/paperboard/article (0.94), [65] Textile yarn/fabric/
art. (0.94), [78] Road vehicles (0.84), and [89] Misc manufactures n.e.s. (0.87).
Furthermore, in 2018, Brazil’s IIT with BRICS countries at two-digit SITC chapters are
high in followings industries: IIT with China in [8] Animal feed ex unmlcer. (0.89), [23]
Crude/synthet/rec rubber (0.81), [55] Perfume/cosmetic/cleansr (0.99), [57] Plastics in
primary form (0.91), [63] Cork/wood manufactures (0.86), [67] Iron and steel (0.87), and
[68] Non-ferrous metals (0.77); IIT with India is in [26] Textile fibres (0.83), [55]
Perfume/cosmetic/cleansr (0.97), [67] Iron and steel (0.98), [72] Industry special machine
(0.97), [73] Metalworking machinery (0.75), and [76] Telecomms etc equipment (0.82);
IIT with Russian Federation is in [4] Cereals/cereal preparatn (0.81),[51] Organic
chemicals (0.96), and [75] Office/dat proc machines (0.80); and IIT with South Africa is
in [26] Textile fibres (0.78), [28] Metal ores/metal scrap (0.78), [53] Dyeing/
tanning/color mat (0.91), [65] Textile yarn/fabric/art. (0.92), [67] Iron and steel (0.96),
and [84] Apparel/clothing/access (0.86). In addition, in 2018, China’s IIT with BRICS
countries at two-digit SITC chapters are high in industries like: IIT with Brazil in [5]
Vegetables and fruit (0.93), [26] Textile fibres (0.79), [57] Plastics in primary form
(0.79), [67] Iron and steel (0.85), and [68] Non-ferrous metals (0.93); IIT with India is in
[5] Vegetables and fruit (0.79), [8] Animal feed ex unmlcer. (0.78), [23] Crude/
synthet/rec rubber (0.87), [29] Crude anim/veg mater n.e.s. (0.97), [33] Petroleum and
products (0.95), [41] Animal oil/fa (0.87), and [66] Non-metal mineral manuf. (0.79); IIT
with Russian Federation takes place in [11] Beverages (0.78), [12] Tobacco/manufactures
(0.79), [52] Inorganic chemicals (0.79), [61] Leather manufactures (0.80), [63]
Cork/wood manufactures (0.84); and IIT with South Africa is in [6] Sugar/sugar
prep/honey (0.96),[27] Crude fertilizer/mineral (0.88), and [54] Pharmaceutical products
(0.88). On the other hand, in 2018, India’s IIT with BRICS countries at two-digit SITC
chapters are high in followings: IIT with Brazil is in [12] Tobacco/manufactures (0.96),
[26] Textile fibres (0.87), [52] Inorganic chemicals (0.80), [66] Non-metal mineral
manuf. (0.80), [67] Iron and steel (0.78), [73] Metalworking machinery (0.87), and [85]
Footwear (0.98); IIt with China takes place in [6] Sugar/sugar prep/honey (0.88), [12]
Tobacco/manufactures (0.96), [22] Oil seeds/oil fruits (0.85), [29] Crude anim/veg mater
n.e.s. (0.79), and [57] Plastics in primary form (0.97); IIT with Russian Federation largely
happens in [51] Organic chemicals (0.86), [61] Leather manufactures (0.91), [72]
Industry special machine (0.77), [75] Office/dat proc machines (0.86), [79]
Railway/tramway equipment (0.86), and [87] Scientific/etc instrument (0.89); and IIT
with South Africa is in [11] Beverages (0.84), [23] Crude/synthet/rec rubber (0.84), and
[57] Plastics in primary form (0.79). Additionally, in 2018, Russian Federation’s IIT with
BRICS countries at two-digit SITC chapters are high in industries such as: IIT with Brazil
is in [25] Pulp and waste paper (0.75), [57] Plastics in primary form (0.77), [62] Rubber
manufactures n.e.s. (0.87), [63] Cork/wood manufactures (0.77), and [82] Furniture/
furnishings (0.98); IIT with China in [7] Coffee/tea/cocoa/spices (0.97), [8] Animal feed
ex unmlcer. (0.98), [64] Paper/paperboard/article (0.86), [71] Power generating
equipment (0.86), and [79] Railway/tramway equipment (0.96); IIT with India in [61]
Leather manufactures (0.82), [69] Metal manufactures n.e.s. (0.90), [72] Industry special
machine (0.87), and [77] Electrical equipment (0.83); and IIT with South Africa takes
place in [59] Chem material/prods n.e.s. (0.99), [84] Apparel/clothing/access (0.92), [88]
Photographic equ/clocks (0.99), and [89] Misc manufactures n.e.s. (0.82). Lastly, in 2018,
South Africa’s IIT with BRICS countries at two-digit SITC chapters are high in
followings industries: IIT with Brazil is in [3] Fish/shellfish/etc. (0.78), [5] Vegetables
and fruit (0.91), [27] Crude fertilizer/mineral (0.91), [28] Metal ores/metal scrap (0.75),
[51] Organic chemicals (0.99), [68] Non-ferrous metals (1.00), [69] Metal manufactures
n.e.s. (0.81), and [89] Misc manufactures n.e.s. (0.86); IIT with China takes place in [6]
Sugar/sugar prep/honey (0.82), [8] Animal feed ex unmlcer. (0.75), [12]
Tobacco/manufactures (0.90), [27] Crude fertilizer/mineral (0.82), and [42] Fixed veg
oils/fats (0.79); with India in [5] Vegetables and fruit (0.77), [26] Textile fibres (0.90),
[66] Non-metal mineral manuf. (0.88), [75] Office/dat proc machines (0.97), and [79]
Railway/tramway equipment (0.96); and IIT with Russian Federation is in [0] Live
animals except fish (0.90), [55] Perfume/cosmetic/cleansr (0.87), [59] Chem material/
prods n.e.s. (0.95), [71] Power generating equipment (0.81), and [79] Railway/tramway
equipment (0.96).
Dynamic IIT analysis
The analysis has been carried out at four digit-SITC data for the three sub period such as
2000-2005, 2009-2014 and 2015-2018. The followings process has been adopted to carry
out analysis of MIIT: Firstly, data of four digits SITC has been deflated to get the real
change during the selected period by consumer price index of United States of America
collected from World Development Indicator, World Bank database and its base year is
2010. This method used because of non-availability of export/import unit price index for
all BRICS countries. Secondly, regular trade flow of selected sub-period has been
considered in the analysis, e.g. Ferto (2008). Thirdly, sub-sectors are defined in this
analysis at four digits and aggregated to calculate industry at two-digit SITC level.
Fourthly, both indices have been added over all sectors by trade weight. Finally, the
results have been reported in Tables 10-12.
The analysis sample size has been selected as per above said criteria such as in 2000-2005,
554 sub-industries and 55 industries were selected, 707 sub-industries and 60 industries
were chosen during 2009-2014 and in 2015-2018, 760 sub-industries and 60 industries
were selected for the analysis of MIIT of BRICS countries. Further, in 2000-2005, Brazil
trade with BRICS countries is more or less IT except with India; it was VIIT nearly 50%.
China trade with BRICS countries is almost IT. India trade with BRICS countries is
dominated by IT except with Russia, it was HIIT. Russia trade with BRICS countries is
approximately IT. South Africa trade with BRICS countries is largely IT. In addition,
South Africa trade with Brazil show 30% VIIT. Nonetheless, South Africa has not been
observed an active trade partner with the BRICS countries during 2000-2005 (see Table 10).
Table 10. Decomposition Trade Flow of BRICS Countries during 2000-2005
AJ AW AJ-AW 1-AJ
Country Name Partner Country Name IIT HIIT VIIT IT
China 0.22 0.13 0.09 0.78
Brazil India 0.67 0.17 0.50 0.33
Russian Federation 0.14 0.14 0.00 0.86
Brazil 0.30 0.17 0.13 0.70
China India 0.32 0.18 0.13 0.68
Brazil 0.28 0.10 0.18 0.72
India China 0.25 0.14 0.11 0.75
Brazil 0.00 0.00 0.00 1.00
Russian Federation China 0.25 0.12 0.13 0.75
India 0.31 0.09 0.22 0.69
Brazil 0.46 0.16 0.30 0.54
South Africa China 0.10 0.09 0.02 0.90
India 0.29 0.15 0.14 0.71
Additionally, in 2009-2014, Brazil trade with BRICS countries is more or less IT except
with India; it was both HIIT and VIIT nearly 16% and 15% respectively. China trade with
BRICS countries almost concentrated to Brazil and India, where, trade with Brazil is
almost IT. On the other hand, trade with India is vertically integrated within the same
Industry. Further, India trade with BRICS countries is dominated by IT except with Brazil
and China, where, it was both HIIT (nearly 26% and 17%) and VIIT (around 21% and
22%), respectively. Russia trade with BRICS countries is approximately IT. Nonetheless,
Russia trade with Brazil is HIIT nearly 20% and trade with India is close to 38%
vertically integrated in the same industry. And South Africa trade with BRICS countries
is largely IT. In addition, South Africa trade with Brazil reveals both nearly 22% HIIT
and 21% VIIT has been observed. Nonetheless, South Africa trade with India and China
accounts for 18% and 15% of VIIT respectively, during 2009-2014 (see Table 11).
Table 11.Decomposition Trade Flow of BRICS Countries during 2009-2014
AJ AW AJ-AW 1-AJ
Country Name Partner Country Name IIT HIIT VIIT IT
China 0.11 0.04 0.07 0.89
Brazil India 0.31 0.16 0.15 0.69
South Africa 0.06 0.06 0.00 0.94
Brazil 0.15 0.05 0.10 0.85
China India 0.39 0.09 0.29 0.61
Brazil 0.35 0.13 0.22 0.65
India China 0.37 0.12 0.25 0.63
South Africa 0.00 0.00 0.00 1.00
Brazil 0.29 0.20 0.09 0.71
India 0.47 0.09 0.38 0.53
South Africa 0.09 0.09 0.00 0.91
Brazil 0.43 0.22 0.21 0.57
India 0.26 0.08 0.18 0.74
Additionally, in 2015-2018, Brazil trade with BRICS countries is more or less IT except
with India and China; where trade with India is HIIT (13%) as well as VIIT (33%) and
trade with China is VIIT nearly 46%. Further, China trade with BRICS countries takes
place mostly with Brazil and India, such as trade with Brazil is VIIT nearly 50% and
trade with India is HIIT (14%) and VIIT (31%) as well. India trade with BRICS countries
is dominated by IT. Nevertheless, IIT trade also observed between India and BRICS
countries, where HIIT accounts the highest such as Brazil (26%), China (17%), Russia
(2%) and South Africa (16%). In addition, VIIT also observed Brazil and China 21% and
22%, respectively.
On the other side, Russia trade with BRICS countries is approximately IT except India,
where its percentage share in trade flow is 51%. Nevertheless, Russia’s trade with BRICS
countries also observed VIIT such as trade with Brazil, China and India accounts for
15%, 27%, 41% respectively. And South Africa trade with BRICS countries is mostly IT.
However, South Africa’s trade with BRICS countries also reveals IIT, where, South
Africa’s VIIT with Brazil, China and India accounts for 24%, 27% and 11% respectively.
In addition, HIIT only takes place between South Africa and Brazil stands for 10% only,
during 2015-18 (see Table 12).
Table 12. Decomposition Trade Flow of BRICS Countries during 2015-2018
Country Name Partner Country Name AJ AW AJ-AW 1-AJ
IIT HIIT VIIT IT
China 0.52 0.05 0.46 0.48
Brazil India 0.46 0.13 0.33 0.54
South Africa 0.00 0.00 0.00 1.00
Brazil 0.54 0.04 0.50 0.46
China India 0.46 0.14 0.31 0.54
Country Name Partner Country Name AJ AW AJ-AW 1-AJ

IIT HIIT VIIT IT
Brazil 0.46 0.26 0.21 0.54
India China 0.39 0.17 0.22 0.61
South Africa 0.16 0.16 0.00 0.84
Brazil 0.17 0.03 0.15 0.83
India 0.49 0.08 0.41 0.51
South Africa 0.00 0.00 0.00 1.00
Brazil 0.34 0.10 0.24 0.66
India 0.14 0.03 0.11 0.86
Discussion
The GLI indices values is in line with Grubel and Lloyd (1971), Leitão (2011).
Further, the MIIT indices values are also in line with Brulhart (1994) and Thomand
and McDowell (1999).
In addition, the value of both the static and dynamic indices should not be greater than
one have been observed from the analysis, which is in line with literature of IIT. For
example, Proença and Faustino (2015), Filipowicz (2016), Şahbudak and Şahin (2016),
Maxir and Masullo (2017), Mutambara (2017), Dwesar and Kesharwani (2019) and
Varshini and Manonmani (2019). Hence, the results are very robust and reliable.
Conclusion and policy implications

In this article, we tried to investigate Intra-Industry Trade among the BRICS countries.
IIT measured by using Grubel and Lloyd IIT Index for static analysis and Thom and
McDowell (1999) MIIT index for dynamic analysis. In addition, the decomposition of IIT
carried out to distinguish between Horizontal and Vertical IIT at Industry level (two
digits- SITC level data). The unit of analysis selected at one- digit and two-digit SITC
Industry level for GL IIT index. Further, to conduct MIIT analysis, industry is defined at
two-digit-SITC level data by aggregating four-digit SITC sub-industry level data of IIT of
BRICS countries. Further, study analysed the Pre and Post-BRICS trade pattern of IIT.
Thereafter, this study emphasises that do emerging economies IIT among themselves? On
the basis of estimated results of this study revealed that IIT occurs at higher level of
aggregation. This signifies that developing countries are trading in the same Industry for
love for variety and cost effectiveness. In addition, evidence on MIITI shows that BRICS
countries’ IIT increased over the period. Besides, BRICS countries get benefited from IIT
over the years and since BRICS inception as well, which is validated by MIITI values.
Hence, the empirical result contradicts conventional Krugman (1979, 1985) hypothesis of
IIT takes place in developed nations (industrialist nations).
With respect to the policy implications, BRICS countries should focus on opportunities of
trade complementary of intermediates products. This will enhance cost effectiveness of
product development or production. Further, this will promote innovation in the BRICS
region. To achieve this, countries needs to conduct constructive trade dialog among the
BRICS countries.
Acknowledgements
This article is based on a chapter of author’s PhD thesis. Author would like to offer
sincere thanks to Biswajit Mandal, Associate Professor, and Prof. Sarbajit Sengupta,
Professor of Economics, Department of Economics & Politics, Visva-Bharati (A Central
University), Santiniketan, West-Bengal, India. Financial support from Indian Council of
Social Science (ICSSR), New Delhi, in form of Doctoral Fellowship is greatly
acknowledged. The usual disclaimer applies.
References
Aditya, A. and Gupta, I., 2019. Intra Industry Trade: Is it Horizontal and Vertical, Economic and
Political Weekly, Vol. No. 25, pp. 29-38.
Al-Mawali, N., 2015. Country-Specific Determinants of Vertical and Horizontal Intra-Industry
Trade Of South Africa: An Empirical Investigation, South African Journal of Economics,
Vol. 73:3.
Anisul, M.I., 2018. Inter- and Intra-industry Trade Relations between Bangladesh and India:
Empirical Results, FIIB Business Review, Sage Publications 7(4) 1-13, DOI:
10.1177/2319714518805182
Balassa, B., Tariff Reductions and Trade in Manufactures Among the Industrial Countries,
American Economic Review, 56, 3, 1966, pp. 466-73.
Bhalla, A. and Bhalla, P., 1997. Regional Blocks: Building Blocks or Stumbling Blocks?, London:
Macmillan.
Boyrie, M.E. and Kreinin, M., 2012. Intra-Industry Trade Revisited: A Note, Open Econ Rev23,
pp. 741-745DOI 10.1007/s11079-011-9212-6
Brülhart, M., 1994. Marginal Intra-Industry Trade: Measurement and Relevance for the Pattern of
Industrial Adjustment, Weltwirtschaftliches Archiv, 130 (3), pp. 600-613.
Casson, M and Pearce, R.D., 1988. Intra-Firm Trade and The Developing Countries, Ch. 8, edited
by David Greenaway, Macmillan Education Ltd., pp. 132-156.
Caporale, G.M., Sova, A., and Sova R., 2015. Trade flows and trade specialisation: The case of
China, China Economic Review, Vol. 34, pp. 261-273.
Czerewacz-Filipowicz, K., 2017. The Russian Federation RTAs in the Light of Global Value
Chains, Procedia Engineering, Elsevier, 182, pp. 120-126.
Dwesar, R. and Kesharwani, A., 2019. Examining Intra-Industry Trade between India & China: Is
India on the Right Track? Theoretical Economics Letters, 9, 1834-1851,
https://doi.org/10.4236/tel.2019.96117
Eaton, J. and Kierzkowski, H., 1984. Oligopolistic competition, product variety and international
trade. In H. Kierzkowski (Ed.), Monopolistic Competition and International Trade, Cambridge:
Cambridge University Press.
Ferto, I., 2008. Dynamics of intra-industry trade and adjustment costs. The case of Hungarian food
industry, Applied Economics Letters, 15:5, 379-384, DOI: 10.1080/13504850600689949
Falvey, R. and Kierzkowski, H., 1987. Product quality, Intra-industry trade (Im) Perfect
competition, in H. Kierzkowski (Ed.), Protection and Competition in International Trade.
Essays in Honor of W.M. Corden, Oxford USA, Basil Blackwell, pp. 143-116.
Grubel, H.G., and Lloyd, P.J., 1971. The Empirical Measurement of Intra-industry Trade.
Economic Record 47 (December): pp. 494-517.
Greenaway, D., Hine, R. and Milner, C., 1994. Country Specific Factors and the Pattern of
Horizontal and Vertical Intra-industry Trade in the UK. Weltwirtschaftliches Archiv, 130,
No. 1: pp. 77-100.
Greenaway, D., Hine, R.C., Milner, C. and Elliott, R., 1994. Adjustment and the Measurement of
Marginal Intra-industry Trade. Weltwirtschaftliches Archiv, 130, No. 2: pp. 418-427.
Heckscher, E., 1919. The effect of foreign trade on the distribution of income,
EkonomiskTidskrift21, 497-512. Reprinted (1991) in Flam, H. and Flanders, M. (Eds.).
Heckscher-Ohlin Trade Theory, Cambridge, MA: MIT Press, 43-69.
Helpman, E., 1981. International trade in the presence of product differentiation, economies of
scale and monopolistic competition: A Chamberlin-Heckscher-Ohlin approach, Journal of
International Economics, Elsevier, vol. 11(3), pp. 305-340
Helpman, E. and Krugman, P.R., 1985. Market Structure and Foreign Trade. Cambridge, MA,
MIT Press.
Kierzkowki (Ed.), Monopolistic Competition and International Trade, Oxford USA: Oxford
University Press, pp. 69-83.
Krugman, P., 1979. Increasing returns, monopolistic competition, and international trade. Journal
of International Economics 9 (4), pp. 469-479.
Krugman, P.R., 1980. Scale economies, product differentiation, and the pattern of trade. American
Economic Review 70(5), pp. 950-959.
Krugman, P.R., 1981. Intra-industry specialization and the gains from trade. Journal of Political
Economy, University of Chicago Press, v. 89, pp. 959-973.
Kubendran, N., 2020. Trade relation between India and other BRICS countries: A
multidimensional approach using Gravity Model and Granger Causality, Theoretical and
Applied Economics, Vol. XXVII, No. 1(622), Spring, pp. 41-56.
Leitão, N.C., 2011. Intra-Industry Trade in Tourism Services, Theoretical and Applied Economics,
Vol. XVIII, No. 6(559), pp. 55-62.
Lohani, K.K., 2020. Trade Flow of India with BRICS Countries: A Gravity Model Approach,
Global Business Review, Sage Publications, pp. 1-18. Forthcoming.
Maxir, Henrique dos Santos and Masullo, Liamara Santos, 2017. The Brazilian Insertion into the
International Trade of Forest Products Chain, RevistaÁrvore. 41(3): E410318,
HTTP://DX.DOI.ORG/10.1590/1806-90882017000300018.
Marzábal, Ó.R., Carreira, M. del C.S., Arévalo, J.A.L. and Ovando E.A., 2016. Progress in the
Pattern of Intra-industrial Trade between the European Union and Latin America: The Cases
of Brazil and Mexico, EU-LAC Foundation, Germany, www.eulacfoundation.org, DOI:
10.12858/0216EN1
Mhaka, S. and Jeke, L., 2018. An evaluation of the trade relationships between South Africa and
China: An empirical review 1995-2014, South African Journal of Economic and
Management Sciences 21(1), a2106. <https://doi.org/10.4102/ sajems.v21i1.2106>
Mui-Yin Chin, Sheue-Li Ong, Chew-Keong Wai and Chin-Hong Puah, 2019. Vertical intra-
industry trade and economic size: The case of Malaysia, The Journal of International Trade
& Economic Development, DOI: 10.1080/09638199.2019.1696878
Nicolette Cattaneo and Jen Snowball, 2019. South Africa’s trade in cultural goods and services
with a focus on cultural trade with BRICS partners, International Journal of Cultural Policy,
25:5, 582-601, DOI: 10.1080/10286632.2019.1626845
Ohlin, B., 1924. ‘The theory of trade’, Reprinted (1991) in Flam, H. and Flanders, M. (eds.)
Heckscher-Ohlin Trade Theory. Cambridge, MA: MIT Press, pp. 75-214.
O’Neill, J., 2001. Building better global economic BRICs. Global Economics (Paper no. 66).
Goldman Sachs. <https://www.goldmansachs.com/insights/archive/archive-pdfs/build-better-
brics.pdf>
Proença, I. and Faustino, H.C., 2015. Modelling bilateral intra-industry trade indexes with panel data:
a semi parametric approach, Comput Stat, 30: 865-884, DOI 10.1007/s00180-015-0556-z
Ricardo, D., 1817. The Principle of Economy and Taxation, London: John Murray.
Şahbudak, E. and Şahin, D., 2016. Measurement of Vertical And Horizontal Intra-Industry Trade
In Agricultural Food Products: The Case Of China And Brazil, The Journal of Academic
Social Science Studies, No. 42, pp. 145-154, Doi number: <http://dx.doi.org/10.9761/
JASSS3280>
Smith, A., 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. 3 vols. Dublin:
Whitestone.
Thom, R. and McDowell, M., 1999. Measuring Marginal Intra Industry Trade, Weltwirtschaftliches
Archiv, Vol. 135 (1), pp. 48-61.
Tomasz Brodzicki, Tomasz Jurkiewicz, Laura Márquez-Ramos and Stanisław Umiński, 2020.
Patterns and determinants of the horizontal and vertical intra-industry trade of regions: panel
analysis for Spain & Poland, Applied Economics, 52:14, pp. 1533-1552, DOI:
10.1080/00036846.2019.1676871
Tsitsi Effie Mutambara, 2017. How has trade between South Africa and China evolved over the
past decade?, Transnational Corporations Review, 9:2, pp. 97-111, DOI:
10.1080/19186444.2017.1326721
Varshini, N.M. and Manonmani, M., 2019. Pattern of Trade and Trade Advantage in
pharmaceutical industry in India, Journal of International Economics, Institute of Public
Enterprises, Vol. 10 (1), pp. 50-60.
Wang, J., 2009. The analysis of intra-industry trade on agricultural products of China, Front.
Econ. China, 4(1): 62-75 DOI 10.1007/s11459-009-0004-5
Yugo Konno, 2016. Evaluating Russia’s trade patterns, Post-Communist Economies, 28:3, 300-313,
DOI: 10.1080/14631377.2016.1184427
Appendix
SITC (Standard International Trade Classifications)-one digit. It includes ten sectors such as total
all products, [0] Food and live animals, [1] Beverages and tobacco, [2] Crude materials, inedible,
except fuels, [3] Mineral fuels, lubricants and related materials, [4] Animal and vegetable oils, fats
and waxes, [5] Chemicals and related products, n.e.s., [6] Manufactured goods, [7] Machinery and
transport equipment, [8] Miscellaneous manufactured articles, [9] Commodities and transactions,
n.e.s.
Fossil fuel consumption, economic growth and CO2 emissions.

Causality evinced from the BRICS world
Rochna ARORA
Guru Nanak Dev University, India
Rochna.arora17@gmail.com
Dr. Baljit KAUR
Guru Nanak Dev University, India
baljit17_kaur@yahoo.co.in
Abstract. The present study looks at the long term relationship among fossil fuel consumption,
economic growth and CO2 emissions in the BRICS economy. The data runs from the period 1990-
2014. The unit root tests prove that the data contain unit root following which Pedroni’s and
Kao’s cointegration test is applied. The results of cointegration prove that there is a long term
relationship that exists. The long run estimates from DOLS and FMOLS show that increase in
fossil fuel consumption will increase growth but at the same time with more carbon emissions the
level of growth will decrease. Additionally increase in economic growth will reduce the amount of
CO2 emissions thus favouring Kuznets inverted U hypothesis. Finally causality results are arrived
at by using Dumitrescu-Hurlin panel causality technique. The results from causality tests show
that unidirectional causality from energy to GDP; bidirectional relationship between emissions
and GDP and unidirectional causality from fuel consumption to environmental degradation. The
causality directions serve important policy implications for the government of the emerging
economies to focus more on non-conventional sources of energy so as to keep the environment in
the best of its health and at the same time make growth sustainable in the long run.
Keywords: energy consumption, economic growth, emissions, causality, BRICS.
JEL Classification: C23, O44, O57.

132 Rochna Arora, Dr. Baljit Kaur
1. Introduction
The concept of sustainable development was first discussed in the Brundtland Report
(1972). The growth thereafter was always questioned whether it is sustainable or not.
Numerous regulations and summits found place in the economic domain addressing the
issue of environmental cost of growth and development. Though there does exist a
negative relation between energy consumption and CO2 emissions for almost all the
economies of the world but at the same time energy consumption in all its forms supports
economic growth. Energy is regarded as a precondition for economic growth. The U.S.
Energy Information Administration (EIA, 2013) conjectured the momentous role played
by energy in the process of economic growth. It stated: A country's economy and its
energy use, particularly electricity use, are linked. Short-term changes in electricity use
are often positively correlated with changes in economic output (measured by gross
domestic product (GDP). However, the underlying long-term trends in the two indicators
may differ. All else equal, a growing economy leads to greater energy and electricity use.
The literature world across has corroborated these results for mixed set of economies
(Apergis and Payne, 2010; Campo and Sarmiento, 2013; Pradhan, 2010; Omri, 2013;
Mohanty, 2015 and many others).
The effect of energy consumption on economic growth can be regarded as a positive
externality but another dimension to this is the existence of even stronger but negative
externality in the form of environmental degradation. CO2 emissions from the consum-
ption of energy have taken a serious toll on the performance of economic growth of many
economies particularly the developing ones. For instance China has surpassed the state of
USA when it comes to the emissions of the most important greenhouse gas i.e. CO2. India
is also the third largest emitter of CO2 in the world lying below China and USA.
Interestingly China is touted as the world’s fastest growing economy too. It’s very
evident from this postulate that china’s growth is detrimental to the environment. Having
said that, one cannot imagine an economy’s existence without growth. But growth of
such nature would actually be illusionary. The Chinese economy attempted to estimate its
green GDP back in 2000’s but the exercise was abandoned because it abridged the value
of its GDP. Even then the role of green economy is stressed upon. Green Economy is
defined by United Nations Environment Program (UNEP) as way of operation of an
economy which ameliorates human’s well-being together with trimming down the
inequality and environmental risks so as to facilitate sustainable development. In Green
Economy, the main concern in not only on economic growth but it requires investing in
those projects which are more environment-friendly. The advancing and emerging
economies are not only seeing an increase in their economic growth but are also
experiencing environmental problems of increased pollution and global warming. Though
the contribution of developing countries in GHG emissions is lower than that of
developed countries yet it won’t be wrong to expect adoption of environmental friendly
modes of working from them.
The developing economies have always exhorted the effort to go for cleaner technologies
but at the same time there is a general consensus among all that cleaner technologies are
dearer and a misfit for them. It needs to be the responsibility of the developed world to
Fossil fuel consumption, economic growth and CO2 emissions. Causality evinced from the BRICS world 133
come forward in leading the world for clean technology. If the choice has to be made
between growth and environment, the developing countries will surely go for growth
instead of pursuing environmental friendly policies which will be jolting their growth
values. There clearly exists a trade-off between growth and environment conscious
policies. To achieve one, the other has to be sacrificed.
BRICS brings together five major emerging economies, comprising 43% of the world
population, having 30% of the world GDP and 17% share in the world trade. In 2019,
BRICS combined GDP will surpass (using PPP-adjusted GDP) that of G7 economies, and
in 2020, based on IMF forecasts, it will exceed the combined share of the world GDP for
the US + EU27 economies (see figure below). With the projections of BRICS economy
surpassing the other well grown economies in the near future, the economic activity will
definitely see a rampant change. Together with growth increases, the region has also
recorded increase in the CO2 emissions. The BRICS share of carbon dioxide emissions
from fossil fuel combustion in the global total increased from 27 per cent in 1990 to 42
per cent in 2018 (BP, 2019a).
For comparison, over the same period, the share of G7 countries (Canada, France,
Germany, Italy, Japan, the United States and the United Kingdom) in global carbon
dioxide emissions from fossil fuel combustion shrank from 42 per cent to 25 per cent.
The existing reserve of fossil fuels in these economies makes sense as to why there is
immense dependence on these fuels for meeting the energy needs for their operation. The
BRICS region comprises of 32.3% of the proved fossil fuel reserves with individual
countries holding Brazil (16.9), China (378.8), India (265.7), Russia (427.7) and South
Africa (26.1).
The present study can be differentiated from the existing literature in three respects.
1. Most of the studies delving into the energy-income relationship take up per capita
consumption of electricity as a variable representing energy statistics while the present
study will take up fossil fuel energy consumption as a representative variable.
2. The study differs from the existing literature by taking up the set of most important
emerging economies namely BRICS which display a potential to influence the World
economics. Thirdly, the study adopts modern panel econometric techniques in the
form of panel stationarity, panel cointegration techniques along with Fully Modified
Ordinary Least Squares (FMOLS) and Dynamic Ordinary Least Squares (DOLS)
techniques.
3. Apart from these, what differentiates the present paper from the already existing ones
is the application of an alternative to Granger causality test which has been the base of
almost all the causality studies conducted among various set of variables. Instead of
Granger causality tests, the study applies Dumitrescu-Hurlin panel causality technique
which is regarded as an alternative to Granger causality tests.
The present research paper is carried out to achieve the following mentioned objectives
1. To see if there exists any long term relationship among the three variables of fossil
fuel energy consumption, economic growth and CO2 emissions for the BRICS
economy.
2. To estimate the long run elasticity coefficients of each variable with respect to other.
3. To find out the direction of short term causality between all the variables, taking 2
variables each time.
The present analysis is divided into 5 sections including the present one. The second
section elaborates the existing literature talking about the relationship among the three
variables of concern. Section 3 describes the variables that form part of the study and also
details the methodology adopted for carrying out the research. Results are provided in
section 4. Conclusions and relevant policy implications are laid out in the last section i.e.
section 5.
2. Review of literature
The studies pertaining to the selected variables were initially bivariate in form with major
focus on energy-growth relationship. On the basis of the relationship, Ozturk (2010),
Squalli (2007) and Magazzino (2011) provide four hypotheses about the direction of
causality between energy consumption and GDP.
The first is the hypothesis of neutrality, which holds that there is no causality (in either
direction) between these two variables.
The second is the energy conservation hypothesis, which holds that there is evidence of
unidirectional causality from GDP growth to energy consumption.
Under the third hypothesis, which is known as the growth hypothesis, energy
consumption drives GDP growth.
The fourth hypothesis is the feedback hypothesis, which suggests a bidirectional causal
relationship between energy consumption and GDP growth. Numerous studies have been
conducted for various economies of the world each supporting one of the four
hypotheses. The list of studies that conform to the different hypotheses have been
presented in the following table.
Table 1. Studies on energy-income relationship
Growth Hypothesis Squalli (2007), OPEC countries
Apergis and Payne (2009, 2010), American economies.
Lean and Smyth (2010), ASEAN
Mohanty and Chaturvedi (2015)
Conservation Cheng (1999), India
hypothesis Ghosh (2002, 2009), India
Pradhan (2010), India
Odhiambo (2016) South Africa
Feedback hypothesis Belke (2011), OECD countries
Zhang and Xu (2012), Chinese economy
Omri (2013) MENA countries
Lao and Zheng (2014) SSA countries
Campo and Sarmiento (2013), Latin American countries
Osman (2016) GCC countries
Ahmad et al. (2016), India
Kirrikaleli (2018) 35 OECD countries
Source: Authors’ compilation.
The above mentioned studies talk about the bivariate relationship between the variables
of energy consumption and economic growth. The studies that explicitly took CO2
emissions as an additional variable in the model are presented here under.
Farhani and Rejeb (2012) examined the direction of causality among energy
consumption, Gross Domestic Product and CO2 emissions for set of 15 MENA countries
covering the period 1973-2008 by applying panel unit root tests, panel cointegration
models and panel causality tests. The findings of the study point to no causality between
GDP and electricity consumption and between CO2 emissions and energy consumption in
the short run. However a unidirectional causality is found in the long run from GDP to
CO2 emissions and energy consumption. Thus serving an important policy perspective
that MENA countries can go in for conservation policies without impeding the growth of
the economy. Another study for MENA countries was undertaken by Omri (2013) for the
period 1990-2011.
Using simultaneous equations model the causality worked out was bidirectional and thus
in favour of feedback hypothesis for energy and growth and energy and CO2 emissions as
well. Lao and Zheng (2014) explored the causal relationship among electricity
consumption, economic growth and carbon-di-oxide emissions for a group of 14 Sub-
Sahara Africa (SSA) countries from 1980-2009 using a panel cointegration and panel
Vector Error Correction modelling methods. The findings demonstrate that in the long
run electricity consumption has a statistically significant positive impact on CO2
emissions thus validating the existence of U-shaped Environment Kuznets Curve (EKC).
The panel causality tests indicate that there is short-run unidirectional causality from
economic growth to CO2 emissions and electricity consumption. In the long run the
causality becomes bidirectional running between electricity consumption to growth;
electricity consumption and CO2 emissions and economic growth to CO2 emissions. Thus
it implies that electricity consumption would promise growth together with environmental
degradation. For the environment protection suitable policies need to be initiated without
retarding the growth of the economy.
3. Database and methodology

A dataset on the 5 BRICS countries, namely Brazil, Russia, India, China and South
Africa is used to study the linkage among the three variables of fossil fuel consumption,
CO2 emissions and economic growth. Annual data for 1990-2014 is extracted from the
dataset on the World Bank website. Table 1 clearly highlights the variable given by their
abbreviations EC denotes the fossil fuel energy consumption (% to the total), CO2 and EG
indicate the amount of emissions (kg per 2011 GDP) and the GDP constant at 2011 U.S.
dollars respectively. These variables which form a part of the study are expressed in their
natural logarithmic forms.
Table 2. List of variables
Variables Representation
CO2 emissions (kg per 2011 PPP $ of GDP) CO
Fossil fuel energy consumption (% of total) EC
GDP per capita, PPP (constant 2011 international $) EG
To study the long term relationship and the causality pattern among the stated variables
certain technical pre requisites are required to be met starting with the stationarity
properties of the selected variables. If the variables are non-stationary at level then first
differenced variables are used for the empirical assessment. Most of the variables
normally become stationary at first level. If the variables fail to establish stationary at first
difference, second difference will be computed. The level at which variables become
stationary is important because that marks the order of integration of the study. After
establishing the non stationarity of the variables at level and stationarity properties at first
difference (unit root stationary), the cointegration properties of the model will be tested.
The existence or non-existence of cointegration is of high importance because that will
drive the future course of modelling for meeting the research objectives. If the variables
are cointegrated, meaning that they share a long run relationship, their relationship will be
quantified.
Lastly causality tests will be carried out to clearly mark off the direction of relationship
that exists between the two variables at a time. The stationarity tests available in the
literature are Levin, Lin, and Chu (2002), Breitung and Candelon (2005), Im, Pesaran,
and Shin (2003), Maddala and Wu (1999), and Choi (2001) unit root tests. The
stationarity tests can be conducted under three models: intercept, trend and intercept and
none of the two. The present analysis is based on intercept and trend model. There exists
a starking difference between these multiple panel unit root tests. The tests of Im,
Pesaran, and Shin (IPS), Augmented Dickey and Fuller-Fisher (adf-Fisher) and
Phillips and Perron-Fisher (PP-Fisher) (Im et al., 2003, Maddala and Wu, 1999, and
Choi, 2001) consider single unit root along with different autocorrelation coefficients for
different cross sections but Levin-Lin-Chu (LLC) and Breitung unit root tests (Levin et al.,
2002, and Breitung and Candelon, 2005) allow common unit root along cross sections. For
the same order of integration I (1) variables cointegration linkage is investigated using
Pedroni (1991) test. But this test can only be applied if the order of integration of all the
selected variables is same. Pedroni (1999) defines seven statistics divided into 2 groups:
within dimension and between dimension. The former includes 4 test statistics which are
Panel v-Statistic, Panel rho-Statistic, Panel PP-Statistic and Panel ADF-Statistic while the
latter category includes Group rho-Statistic, Group PP-Statistic and Group ADF-Statistic.
Pedroni (1999) also states that out of these seven statistics the two most important ones
are Panel ADF-Statistic and Group ADF-Statistic. The null hypothesis in these tests states
no cointegration which needs to be rejected. The rejection of Null hypothesis will be true
if four statistics out of seven reject the null hypothesis. But in case of conflicting results
the final conclusion for rejection of null hypothesis will be dictated by Panel ADF-
Statistic and Group ADF-Statistic. If these two reject the null hypothesis that would
ensure the existence of long run relationship among the variables in consideration. The
cointegrating equations will be run by considering all variables in turn dependent
variable. With three variables under study there will be three cointegrating equations. The
equation specification is given under:
lnEGit = αi+ δit + β1i ln ECit + γit lnCO+ Ɛit [1]
lnECit = αi+ δit + β1i ln EGit + γit lnCO+ Ɛit [2]
lnCOit = αi+ δit + β1i ln EGit + γit lnEC+ Ɛit [3]
In these equations, αi represents the country specific impacts, δi is representative of the
time trends in the analysis and lastly Ɛit is the residual term. The two subscripts also have
a significant meaning wherein i = 1, 2,…, N represent panel members and t = 1, 2,…, T
represent time periods.
The first equation studies the long run impact of energy consumption and CO2 emissions
on economic growth; the second equation looks into the impact of economic growth and
CO2 emissions on energy consumption and the last equation with CO2 emissions as the
dependent variable works out the long run impact of energy consumption and economic
growth on the dependent variable. Apart from Pedroni’s cointegration test, Kao’s residual
cointegration test is employed as a robustness check for cointegration in the analysis.
Once the cointegration tests verify the presence of long run relationship in all the three
models, the quantification of the results will be carried out using Pedroni’s Fully
Modified Ordinary Least Squares (FMOLS) and Dynamic Ordinary Least Squares
(DOLS) tests. These tests will help us in arriving the elasticity estimates of the variables.
After this Dumitrescu-Hurlin panel causality technique has been applied to discover
causality properties between energy consumption, economic growth and CO2 emissions
in the panel format. There is a specific reason involved in applying this particular
causality test rather than the conventional Granger Causality test. The reason behind the
application of Dumitrescu-Hurlin panel causality technique is that it makes no
assumption about homogeneity among the cross sections involved in the analysis while
the infamous Granger causality test is based on homogeneous cross sectional units.
4. Results
We first check the stationarity properties of the variables by employing the tests
mentioned in the previous section. The results of the stationarity tests are presented in the
adjoining table. We find out that all the variables of economic growth, fossil fuel
consumption and carbon-di-oxide emissions fail to reject null hypothesis of no unit root.
As a result when we find out the first difference of the variables and check for the
stationarity of the variables we see that variables become stationary. This means that all the
variables contain unit root and thereby deduce the order of integration to be 1 i.e. I (1).
Table 3. Unit root results
Variables LLC Breitung IPS ADF-Fisher PP-Fisher
At level EG -0.586 2.932 1.076 11.648 11.591
EC 0.233 0.694 0.018 9.956 5.780
CO 2.119 2.190 1.677 15.884 8.740
At first EG -4.076* -2.291** -3.186* 28.316* 49.506*
difference EC -4.419* -1.297*** -6.438* 52.594* 299.394*
CO -2.005** -2.436* -2.468* 22.881** 65.983*
* significant at 1%, ** significant at 5%, ***significant at 10%. Also the model is run using intercept and
deterministic trend.
Hereafter, since the variables contain unit root, we proceed with Pedroni’ long run
cointegration test by taking each variable as a dependent variable in turn. The results of
cointegration tests of both Pedroni as well as Kao have been reproduced in the ensuing
table.
Table 4. Panel cointegration results
Pedroni Residual cointegration test Kao residual
cointegration test
Within dimension Statistic Between-dimension Statistic t-statistic (ADF)
Model 1: EG-EC-CO
Panel v - statistic 0.0124 Group rho-statistic -0.627
Panel rho-statistic -1.946** Group PP-statistic -1.779**
Panel PP-statistic -3.959* Group ADF-statistic -3.820*
Panel ADF-statistic -3.931* -3.867*
Model 2: EC-EG-CO
Panel v - statistic 0.481 Group rho-statistic -3.609*
Panel rho-statistic -3.944* Group PP-statistic -2.167**
Model 3: CO-EG-EC
Panel v - statistic -1.873 Group rho-statistic -1.908**
Panel rho-statistic -2.346* Group PP-statistic -0.485
* represent 1% level of significance, ** represent 5% level of significance.
In all the models the null hypothesis of no cointegration is rejected after drawing ample evidence against it
from Pedroni’s as well as Kao’s test. Though in the third model Kao’s test fails to reject no cointegration
hypothesis, yet we find six out seven statistics of Pedroni claiming the same thing.
Table 5. FMOLS and DOLS results

Dependent variable Independent Variables
Model 1: EG-EC-CO EC CO
EG FMOLS 3.44* (8.15) -1.36* (-10.23)
DOLS 2.66* (5.83) -1.15* (-8.15)
Model 2: EC-EG-CO EG CO
EC FMOLS 0.17* (8.39) 0.19* (4.28)
DOLS 0.16* (7.45) 0.18* (3.31)
Model 3: CO-EG-EC EG EC
CO FMOLS -0.50* (-9.92) 1.43* (4.09)
DOLS -0.50* (-10.14) 1.63* (4.34)
*, **, *** signify 1%, 5%, 10% level of significance. The values in parenthesis denote t-statistic.
The outcome of the FMOLS and DOLS for all the three models is presented in the above
table. The results of both cointegration estimates are highly synchronised without any
discrepancy between FMOLS and DOLS coefficients. The signs and the level of
significance is same under both the criterion. The FMOLS estimates for energy
consumption is positive and significant for economic growth. This coefficient value
ranges from 3.44 (in FMOLS) to 2.66 (in DOLS). This means that a percent increase in
fossil fuel consumption will lead to increase in the level of economic growth to the tune
of 2.66-3.44. But on the other hand, increase in CO2 emissions will pull down the
economic growth value. This is evident from the negative coefficient of CO in the first
model. It shows that a percent increase in emissions of CO2 will hurt economic growth to
the level of 1.15-1.36. This symbolises a commensurate increase and decrease in
economic growth accompanied with the usage of fossil fuels which will cause the
emissions to grow. The net change in economic growth will be positive or negative
depending on the magnitude of both its determinants.
The conclusions from model 3 are even more insightful. This is because increase in
economic growth is causing the level of emissions to go down. This means that a percent
increase in economic growth will cause the emission level to decrease by 0.50 units. The
most plausible reason behind this can be switch to non-conventional sources of energy.
The renewable sources of energy in the form of wind, solar, tidal, hydro, geothermal
require a lump-sum investment which the economies can’t afford at an early stage. But
with more economic growth marking the scenes in the developing world could certainly
provide them with money for undertaking research in that direction. This research will
smoothen out the transition from conventional to non-conventional sources of energy for
meeting the energy needs of the country. Thus the research validates the working of
Kuznets inverted U hypothesis. Alongside this, a simplistic reasoning holds true that
more fossil fuel consumption will eventually lead to more emissions in the economy. In
the present case, a percent increase in energy consumption will cause the emission levels
to rise up to the tune of 1.43-1.63 units. This is why the increased usage of fossil fuels is
criticised. The findings of all the three models are in line with the existing literature.
Apart from estimating the long run coefficients using FMOLS and DOLS, we have
applied Dumitrescu-Hurlin panel causality test to find if there is any causality between
economic growth, energy consumption, and CO2 emissions in the short-run. The results
of causality are produced in following table.
Table 6. Causality results of Dumitrescu-Hurlin

Null Hypothesis: W-Stat. Zbar-Stat. Prob.
LOG_ENERGY does not homogeneously cause LOG_GDP 6.45047 3.68153 0.0002*
LOG_GDP does not homogeneously cause LOG_ENERGY 1.93158 -0.27908 0.7802
LOG_CO does not homogeneously cause LOG_GDP 6.05207 3.33234 0.0009*
LOG_GDP does not homogeneously cause LOG_CO 5.60938 2.94435 0.0032*
LOG_CO does not homogeneously cause LOG_ENERGY 2.91527 0.58308 0.5598
LOG_ENERGY does not homogeneously cause LOG_CO 4.78403 2.22096 0.0264**
* represent 1% level of significance, ** represent 5% level of significance.
The results of causality help us to reject the hypothesis claiming that there is no causality
from energy consumption to GDP. Thus there exists unidirectional causality from fuel
consumption to economic growth without any reverse effect for the BRICS countries.
This implies that fuel consumption will significantly contribute to improvement in the
growth levels of the selected countries. There exists bidirectional causality between
emissions and GDP levels. These results are significant at 1 percent level. Similar were
the findings from FMOLS and DOLS estimates. But additional those estimates provided
us with the sign of relationship between the variables which is not discussed at all by the
causality tests. Lastly there exists unidirectional causality between fuel consumption and
emissions level of CO2.
5. Conclusions and policy implications

The present study ventured into studying the nature of relationship and the direction of
causality between these variables. For this the data was culled from World Development
indicators for the BRICS nation spanning 1990-2014. While most of the studies talk
about bivariate relationship between electricity consumption and economic growth, some
have extended it to a tri-variate analysis involving CO2 emissions also into the model. As
per the existing knowledge the study on this aspect has been untouched for the BRICS
nations which represent the set of most important economies of the world. Thus the
present study fills in this gap in the literature. Using advanced econometric techniques the
results produced in the paper are highly important.
From strategic point of view, the results highlight a dire need to substitute conventional
sources of energy with non-conventional ones in order the save the environment from
being degraded to another level. Unless and until we do not surrender the use of fossil
fuels we won’t be able to correct the environmental issue that the world economies are
confronted with. But at the same time if an economy seeks to growth it will have to use
fossil fuels to generate electricity which is the backbone of any growing economy. But an
economy needs to be considerate enough to realise the importance of generating
electricity from sources other than fossil fuels once the ball of economic sets rolling.
What is expected from a growing economy is an understanding and readiness on its part
to accept the notion of developing the environmental friendly methods of generating
electricity.
References
Ahmad, A., Yuhuan, Z., Shahbaz, M., Bano, S., Zhang, Z., Wang, S. and Liu Y., 2016. Carbon
emissions, energy consumption and economic growth: an aggregate and disaggregate
analysis of the Indian economy. Energy Policy, Vol. 96, pp. 131-143.
Apergis, N. and Payne, J., 2009. Energy consumption and economic growth in Central America:
evidence from a panel cointegration and error correction model. Energy Economics, Vol. 31,
Issue 2, pp. 211-216.
Apergis, N. and Payne, J., 2009. Energy consumption and growth in South America: evidence
from a panel error correction model. Energy Economics. Vol. 32, pp. 1421-1426.
Belke, A., Dobnik, F. and Dreger, C., 2011. Energy consumption and Economic growth: new
insights into cointegration relationship. Energy Policy, Vol. 33, pp. 782-789.
Breitung, J. and Candelon, B., 2005. Purchasing power parity during currency crises: A panel unit
root test under structural breaks. Review of World Economics, 141(1), pp. 124-140.
Campo, J. and Sarmiento, V., 2013. The Relationship between Energy Consumption and GDP:
Evidence from a panel of 10 Latin American countries. Latin American Journal of
Economics, Vol. 50, No. 2, pp. 233-255.
Cheng, B., 1999. Causality between Energy Consumption and Economic Growth in India: An
Application of Cointegration and Error-Correction Modelling. Indian Economic Review, Vol.
34, No. 1, pp. 39-49
Choi, I., 2001. Unit root tests for panel data. Journal of international money and Finance, 20(2),
pp. 249-272.
Dumitrescu, E.I. and Hurlin, C., 2012. Testing for Granger non-causality in heterogeneous panels.
Economic Modelling, 29(4), pp. 1450-1460.
Energy Information Administration (EIA), 2013a. <http://www.eia.gov>
Ghosh, S., 2002. Electricity consumption and economic growth in India. Energy Policy, Vol. 30,
pp. 125-129.
Ghosh, S., 2009. Electricity supply, employment and real GDP in India: evidence from
cointegration and Granger causality tests. Energy Policy, Vol. 37, pp. 2926-2929.
Im, K.S., Pesaran, M.H. and Shin, Y., 2003. Testing for unit roots in heterogeneous panels.
Journal of Econometrics, 115(1), pp. 53-74.
Kao, C., 1999. Spurious regression and residual-based tests for cointegration in panel data. Journal
of Econometrics, 90(1), pp. 1-44.
Dervis, K., Abderrahmane, S., Mehmet, C. and Hasan, M.E., 2018. Panel cointegration: Long-run
relationship between internet, electricity consumption and economic growth. Evidence from
OECD countries. Investigación Económica, Vol. LXXVII, No. 303, pp. 161-176.
Levin, A., Lin, C.F. and Chu, C.S.J., 2002. Unit root tests in panel data: Asymptotic and finite-
sample properties. Journal of Econometrics, 108(1), pp. 1-24.
Maddala, G.S. and Wu, S., 1999. A comparative study of unit root tests with panel data and a new
simple test. Oxford Bulletin of Economics and Statistics, 61(S1), pp. 631-652.
Magazzino, C., 2011. Energy consumption and aggregate income in Italy: Cointegration and
causality analysis, Munich Personal RePEc Archive (MPRA), No. 28494.
Mohanty, A. and Chaturvedi, D., 2015. Relationship between Electricity Energy Consumption and
GDP: Evidence from India. International Journal of Economics and Finance, Vol. 7, No. 2.
Narayan, P.K. and Smyth, R., 2009. Multivariate granger causality between electricity
consumption, exports and GDP: evidence from a panel of Middle Eastern countries. Energy
Policy, 37, pp. 229-236
Omri, A., 2013. CO2 Emissions, Energy Consumption and Economic Growth Nexus in MENA
countries: Evidence from Simultaneous Equations Models, MPRA Paper No. 82501
Osman, M., Gachino, G. and Hoque, A., 2016. Electricity consumption and Economic growth in
the GCC countries: Panel data analysis. Energy Policy, Vol. 98, pp. 318-327.
Ozturk, I., 2010. A literature survey on energy-growth nexus. Energy Policy, 38(1), pp. 340-349.
Pedroni, P., 1999. Critical values for cointegration tests in heterogeneous panels with multiple
regressors. Oxford Bulletin of Economics and statistics, 61(S1), pp. 653-670.
Pradhan, R.P., 2010. Transport, Energy and Economic growth triangle in India: cointegration and
Causality. Journal of Sustainable Development. Vol. 3, No. 2.
Squalli, J., 2007. Electricity consumption and economic growth: Bounds and causality analyses of
OPEC members, Energy Economics, 29. pp. 1192-1205.
Zhang, C. and Xu, J., 2012. Retesting the causality between energy consumption and GDP in
China: Evidence from sectoral and regional analyses using dynamic panel data. Energy
Economics, Vol. 34, pp. 1782-1789.
Methods used in risk financing
Ana Maria POPESCU

notariat.dejure@gmail.com
Ștefan Virgil IACOB
stefaniacob79@yahoo.com
Alina Eliza DABIJA
University of Craiova, Romania
dabija_eliza@yahoo.com
Abstract. Given that in most risk management programs, risk control, which involves avoiding,
preventing and reducing risks, is not fully effective and over time several methods of financing the
losses that have occurred have been developed, such as would be detention or transfer.
The withholding assumes that the injured party is also the one that bears the financial consequences,
and the transfer involves the occurrence of a different entity from the injured party that bears the
direct financial consequences. This method transfers the risk to another entity and involves an
additional cost paid by the transferor to the one he accepts, namely the insurance premium or fees.
By transferring risk through insurance, the bank is offered related risk management services, better
defining its risk profile or can contribute to improving performance indicators and flows, thus
preventing the crisis situation.
Keywords: risk management, financing methods, efficiency, indicators.
JEL Classification: G20, G40.

144 Ana Maria Popescu, Ștefan Virgil Iacob, Alina Eliza Dabija
Introduction
From an economic point of view, it is not feasible for an institution to create a reserve
against losses characterized by low probability of occurrence and high impact, namely
catastrophic losses, losses that are very difficult to predict. Thus, an insurance company
that has access to the reinsurance market may be able to absorb losses that may threaten
the solvency of a single institution.
By insuring against operational risks, the bank has the possibility to reduce or eliminate
large fluctuations in the cash flow caused by operating losses, to improve the income and
performance indicators of the financial institution, even leading to an increase in its market
value, to avoid situations catastrophic and benefit from qualified monitoring services,
which can provide a better definition of the operational risk profile.
In order to reduce operational risk, the term operational risk must be included in the concept
of risk, operational risk management must be developed and implemented, business line
competencies and responsibilities must be defined, operational risk reporting must be
performed and the bottom-up management approach of risks.
The article is accompanied by tables and graphs that have the role of better highlighting
these aspects related to the efficiency of banking.
Literature review
In the literature it is considered that an operational risk in order to be insured must meet
three characteristics, namely: to comply with the law of large numbers, the occurrence of
the anticipated event to cause unforeseen damage, and the occurrence of the event and the
size of the loss to can be determined objectively. Notable works in this field, we mention
those of Awdeh et al. (2011) considered a number of issues regarding the impact of capital
requirements on banking risk. Anghelache et al. (2017) and Anghelache et al. (2016)
analyze banking risks, their management methods and have studied the fundamental
elements related to operational risk. Cipovova and Dlaskova (2016) are concerned with
credit risk management methods. Doerig (2000) and Hakens and Schnabel (2010) address
the impact of operational risks on financial services. Geiger (2000) is concerned with the
regulation and supervision of operational risk in the banking system. Hall and Howell
(2001) addressed some issues related to operational risk insurance under the Basel Capital
Agreement. Issues related to operational risk are addressed in their work by Kuritzkes
(2002) and Peters et al. (2009). Miller (2014), Thomas and Ware Preston (2008) and Wendy
(2001) are concerned with insuring operational risks in the banking system.
Methodology, data, results and discussions

The constraining factors regarding the risk mitigation methods are the following:
 the cost of risk mitigation;
 the time required to implement the actions;
 the need for resources;
Methods used in risk financing 145
 difficulties in changing conceptions, stubbornness, because the new procedures

inevitably involve people, difficulties in implementing technical solutions;
 ignorance that can affect the identification of risk.
The financing method corresponding to each institution is determined on the basis of
several criteria, namely:
 the legal provisions, depending on the type of risk, the transfer by insurance is
mandatory;
 the additional costs or those of the transaction determined as the difference between the
price paid by the transferor and the expected value of the payments from the institution
accepting the transfer.
Also, the organization's ability to bear and retain risks must be sufficient to bear the
maximum probable cost of risk, estimated, depending on the industry, on the basis of
financial indicators (e.g. 2% of net assets, 10% of gross profit, 5% of net cash flow).
If there are no transfer methods, withholding is the only way to finance the risk (e.g. events
that will occur with certainty). Also the much higher the degree of control over the risks
the more attractive their retention.
The activities that create risks will be monitored by insurers; for example, control activities
may be reduced if the remuneration is based on the profit recorded and the results of the
control activities. These are reflected in the profit only after a certain period of time, and
the balancing of the situation will be achieved through a monitoring of the risk control
activity.
By accepting the risk, the organization receives a compensation that it invests and from
which it obtains income, and the costs of a possible bankruptcy or the costs of attracting
funds in crisis situations are much higher than the transfer costs. The risk financing costs
for insurers are lower than for other organizations because the reserves set up for the
payment of claims by organizations that are not insurance/reinsurance companies are
considered non-deductible expenses. In the same vein, the additional costs are a
compensation for services that, in the absence of the transfer, the transferring organization
must do alone.
The Basel Committee recognizes the important role of insurance in reducing the financial
impact of an institution's operating losses, which may lead to the allocation of the minimum
capital required for these lower risks.
Insurance is an effective risk management tool used to reduce the economic impact of
losses used by the banking sector over several decades.
In the operational risk sizing model, an insurance contract can be incorporated by separately
evaluating the insurance and then deducting it from the aggregate loss. The effects of the
insurance contract are applied to each loss and then the net losses are added together. This
is much more realistic because it takes into account the overlaps or shortcomings of
individual insurance contracts, as well as the lack of risk exposure and compensation. It
will also allow large losses not to be covered if they occur very late after it has been used
or allow the incorporation of explicit stochastic extension such as loss due to the
counterparty (if the insurer cannot pay the obligations), payment insecurity (if the insurer
does not pay the compensation established in the contract, due to the misunderstanding
between the insured and the insurer regarding the causes of the loss, due to the nature of
the unobservable loss from outside) and the liquidity risk (due to the delay in paying the
insurance). The most natural incorporation of the insurance contract is made immediately
after the calculation of the capital, the insurance being applied on the risk profile of the
bank.
The role of insurance is to transfer the negative impact of risk from one entity to another,
so we can consider it practically a financing in case of an inconvenience that can cause an
acute decrease in liquidity of the financial institution leading to the loss of allocated capital,
and in case the occurrence of a very large event the negative impact in the banking sector
can spread in the insurance sector as well.
The Basel Accord stipulates that only a part of the operational risk can be insured, the rest
not being recognized due to the way the credit institution's operations and processes are
carried out.
Figure 1. Possibility of insuring operational risk
To understand the value of an insurance, it must be realized that the transfer of risk does
not involve the transfer of current risk but of economic impact and impact calculation (for
example, the purchase of fire insurance for a building does not involve immediate economic
impact, but insurance premiums are paid).
Figure 2. Effect of insurance on loss distribution
The insurance has the effect of diminishing the standard deviation (unexpected loss) on the
individual indemnities, as a cost for the institution in the formation of the predetermined
insurance premiums. Thus, the price of insurance must be equal to the expected value of
the loss for the policy period plus the additional costs and profits of the insurance
institution.
Thus, for the insurance company the risk is less expensive considering the effect of
diversification and development of risk management, resulting in the decrease of both the
standard deviation and the average related to the loss distribution curve. It is much more
advantageous for an institution to manage and estimate its own operational risk on a daily
basis, given that insurers manage losses with low probabilities of occurrence that require
high-accuracy industry data. Insurance reduces the uncertainty of the loss and thus the
institution will obtain constant cash flows, which will bring benefits, such as:
 reinvestment of added value, which would otherwise have been lost;
 reduces the financial impact of the loss;
 insurers provide loss control and risk management services:
˗ insurance companies perform external monitoring and risk investigation, the cost
and availability of insurance are incentives to reduce losses, and known causes of
risk determine decisions to assume or transfer risk;
˗ after facing losses due to operational risk, the institution will encounter difficulties
in increasing the new capital and thus the insurance payment will be considered a
valid fund.
Insurance companies can obtain information on the risks that a single institution could not
take and can monitor the risk management of institutions and can guide them in the
development of risk management programs. They also develop a comparative advantage in
the receivables process due to economies of scale and specialization gains and are
concerned about moral hazard, which leads to indifference on the part of institutions
regarding risk management once they have purchased insurance and transferred the risk.
In recent years, the insurance industry has been concerned with creating new products and
services that provide better coverage of operational risk, so there are insurance solutions
(multilateral, specific or limited risk approach), capital market solutions (securitization),
hybrid solutions and possible future solutions.
The multilateral approach (a single policy includes deductibility and a single limit that will
cover a multitude of losses established by the contract) involves the analysis of the controls
and risk procedures used by the institution. Thus, a multitude of possible losses are
identified and insured simultaneously with the commitment that the management of the
institution to maintain the analysis of controls and risk procedures.
The specific approach (several policies will be concluded each with deductibles and own
limits that will cover the multitude of losses established by the contract) involves the
analysis of controls and risk procedures used by the institution, and may even include a full
assessment of operational risk.
The limited risk approach (a single policy will cover a specific loss) involves the payment
of a premium over a period, the net present value of which is equal to the amount of the
loss. The benefit of the insured consists in the fact that the loss is borne during the period
in which the premium is paid.
Securing certain elements of operational risk (such as bonds for earthquakes affecting
Tokyo Disneyland, the guarantor and for earthquakes affecting Japan) involves the
purchase of bonds whose coupon may be attractive to local individual investors and thus
the risk will spread locally and not outside the affected area.
Hybrid solutions involve combinations between the multilateral or specific approach and
the limited approach.
Possible future solutions involve combinations between the multilateral or specific
approach and the limited approach (covering losses that would not otherwise have been
included) demonstrating increased insurance solutions and sufficient loss coverage, which
can be considered a reduction in the institution's risk. The multilateral approach is preferred
by large institutions as long as it has high deductibility’s and limits, but the specific
approach is preferred by any entity, although the required initial value may be quite high
for small institutions. However, the deductibility and limits can also be adjusted according
to the level of the applicant as well as according to different causal factors. Standardization
of insurance products requires only a single assessment by supervisors and can then be
used, but the main disadvantage is that it does not exactly fit the individual needs of the
entities that use them.
The FOA considers insurance as a viable measure to reduce operational risk, but there
should be concerns from supervisors about its use.
The insurance deduction is made within the limit of 20% of the operational risk capital
determined prior to the recognition of risk mitigation techniques. The following types of
insurance policies were used for operational risks:
 the complex bank insurance policy that covers risks such as fraud, counterfeiting, loss
of damage to property, dishonesty or failure to perform duties by employees;
 professional liability, which protects against losses suffered by third parties as a result
of the negligence or professional errors of the insured's employees;
 computer fraud, which covers losses caused by the malfunction of the computer
network, viruses, data transfer problems, fraudulent transactions;
 the employer's liability in case of violation of labour legislation;
 non-financial property that covers the usual risks that may affect the property of the
insured (fire, earthquake, etc.);
 unauthorized transactions that may represent computer fraud or fictitious records and
general liability.
The operational risk insurance procedures are presented in more detail below. Thus, the
complex bank insurance (Bankers Blanket Bond or BBB), covers direct financial losses
resulting from dishonest or fraudulent acts of employees, theft of money, shares or during
transport, forgery, destruction of invoices, shares or signatures.
Computer fraud (CC) policies cover direct financial losses resulting from the alteration,
destruction or falsification of electronic data, damage to programs, viruses or false
instructions transmitted through various types of communications.
Unauthorized Trade (UT) policies cover direct financial losses resulting from an
unauthorized activity performed by an employee.
Property insurance policies (P) cover physical damage to the insured property (tangible)
caused by fires, explosions, collisions and other disasters.
Policies to cover professional indemnities, errors and omissions of bankers (PI) cover debts
or compensation for damages and/or financial losses resulting from the activities of
employees.
General (multilateral) commercial liability policies (CGL) cover legal liability resulting
from personal injury, damage to the owners of a third party in an accident in the course of
business.
Employee practice liability policies (EPL) cover legal liabilities resulting from the
commission by which the institution of an improper practice towards employees including
discrimination harassment and/or coverage of financial losses.
Directors 'and Officers' Liability (D&O) policies cover legal liability resulting from
defective actions of directors 'and officers' actions, including misrepresentation,
mismanagement or material errors or omissions in disclosures of financial information.
Business interruption policies (BI) cover loss of profit and increase in costs due to the
suspension or interruption of business resulting from damage to property insured due to
fires, explosions, collisions, floods and other natural disasters.
Electronic insurance policies (El) cover legal liabilities resulting from electronic activities
related to the insured's internal (website, e-mail) including slander, defamation and
unauthorized copying, invasion of property, breach of security.
The most common elements insured are: loss of software, hardware, cost of programming
failure and emergency intervention services, data reset, consistent loss of financial or
activity.
Given the novelty of this risk, there are currently few institutions that perform proper
management of this type of risk, and the assertion of the existence of an optimal solution
is false, and can, at best, follow its main components. Operational risk instruments vary
from one financial institution to another and include: calculating the capital requirement to
cover this type of risk, creating scenarios against exposure, adjusting statistical
distributions, making cause-effect connections, dividing by risk classes. In order to achieve
the above, the institution must have a database containing information related to
transactions, events, checkpoints and reference, exceptions, variants, process maps.
The Basel II agreement provides for certain criteria for insurance eligibility. Insurance is
used to outsource the risk caused by losses of high severity and low frequency, such as
fraud, physical loss of assets, etc., by reducing the economic impact of operating losses,
which will be reflected in the capital requirement for this type of risk, thus encouraging
sound and prudent management.
The taxonomy provides the framework under which the capital requirement related to
operational risk is determined and provides the initial guide that helps to reduce the risk
through insurance and calculate a close compensation for recognizing the resulting risk
profile.
The connection between insurance and operational risk through a visual representation of
the convergence provided by standard insurance policies adjusted to the taxonomy universe
of loss risk events related to operational risk.
Three ways of adjustment are needed to change the taxonomy, namely it is proposed to
align with the description of loss events in several models depending on primary causative
factors (personnel, processes, systems and external events); it is recommended to include
additional activities to make the taxonomy more robust and it should be indicated whether
each activity is classified according to the definition of business risk.
The insurance application must help the easy understanding of its products because
standard insurance products have been developed in response to customer and market
needs, being maintained by competitive pressures, and the application of insurance
products is flexible to adjust the grouping for any category.
Given that insurance is not a perfect cover for operational risk, it must be considered a
residual risk, which requires the Basel Agreement to specify the minimum criteria for a
contract to reduce operational risk and the methodology for processing capital to explain
the residual risks associated with contracts.
Operational risk taxonomy involves:
 introduction of the initial level (processes, systems, personnel, external);
 intensifying the examples of the activity;
 considering whether each example of activity follows business risk;
 grouping of categories;
 the categories of internal and external fraud are renamed internal and external acts;
 computer fraud occurs in both the categories of internal and external documents;
 employee practices and workplace safety is incorporated into the category of employee
events;
 customers, products and business practices are incorporated in the employee risk
category;
 execution, delivery and process management are included in the process category;
 the destruction of physical assets has been replaced by the destruction or loss of assets.
Although credit institutions will use insurance products against this type of risk, they will
be tempted to reduce the likelihood of this risk, because the occurrence of such an event
can have a negative effect on reputation, and thus loss of customer confidence in the bank's
ability to removal of unwanted effects and thus may suffer much greater damage than those
due to operational risk.
Reinsurance is a very important aspect for the risk management associated with insurers,
and its strategy is a major factor in the external evaluation of the financial power and ability
to pay claims.
In the case of the first two approaches to determining the capital related to the operational
risk that is calculated by applying a certain percentage on the gross income obtained from
the activity carried out, it is not possible to distinguish between institutions that use
management and risk mitigation programs and those that do not. developed such systems.
Due to the fact that simple approaches are not risk sensitive, they do not allow capital
reduction by concluding insurance contracts or other risk mitigation techniques.
The methodology for reducing capital with the help of insurance is a correct, consistent and
right way. Thus, the following are some calculation methods, depending on the approach
used to determine the capital requirement.
Table 1. Capital reduction approach
Capital approach Approaching capital reduction
Approaching Premiums Approaching Limits
Basic Indicator Approach Capital decrease = Amount premium Capital decrease = the sum of the limits of the
insurance policies multiplied by a fixed insurance policies less the insurance premiums,
percentage and a loss reduction rate multiplied by a factor related to the coverage of
gradual transfer of risk each policy
Standardized Approach They are similar to those presented above except that the decrease in capital is determined after the
addition of capital for each line of business.
Internal Measurement The first is an indicator The reduction in capital associated with a policy is a
Approach for risk transfer and is multiplied by the proportion of the limits that cover unexpected losses
gradual reduction of the severity of the loss adjusted by reducing the expected loss through
and by a secondary factor related to the insurance. If the policy does not cover the risk
induction data. segment, it is necessary to introduce an additional
coverage factor to adjust the residual risk.
Loss Distribution Approach Based on historical data, the distributions of frequencies and severities of total loss are established. It
simulates the reduction of the high quantum of the loss distributions due to the insurance, obtaining
the net capital, being able to incorporate the complex or alternative insurance structure.
In the works performed by insurance companies, it is recommended to explicitly include

insurance for each approach to reflect the risk profile of each institution and to encourage
prudent and healthy risk management. Thus, the following are highlighted:
 the desire to create a level of the field, the basic and standardized approaches are less
sensitive in terms of risk, and the explicit inclusion of insurance will allow them a single
way to control the risk and recognition of efforts as well as those using advanced methods;
 the processing of the insurance is inconsistent along the different business lines
considering the explicit and implicit observance of the decrease, but also the potentially
different criteria of the transmission of the insurance of the decrease of the capital;
 the main concern is whether insurance to reduce capital has been incorporated.
The following are the different approaches to reducing capital, highlighting the benefits of
insurance in reducing risk. Thus the basic formula:
Kn = Kg - KRT (1)
Banks that use the basic approach are banks that do not operate internationally. The
approach of premiums and limits determines the capital requirement at the company level
and provides a partial reduction of the capital requirement for all standard insurance
policies. The Premium approach is based on the reduction of capital for insurance contracts
by paying premiums for policies, which are directly correlated with the amount of risk
transferred.
So:
KRT = P × λ (2)
The solution of this approach is the calibration of the factor λ.
Among the advantages of this method we can list: very simple calculation formula; the
insurance premium is a measure of the risk posed by market forces; avoid covering
potential arbitrage high limits/low probabilities of approaching limits. However, this
method has a number of shortcomings, such as:
 the differences in the value of the insurance limits are not explicitly taken into account;
 the efficiencies of the premiums paid are not taken into account (premiums paid
according to the favourable risk tail are more efficient in reducing the risk than what the
supervisors want 99.9%);
 determining the value of λ;
 the former fluctuate depending on market cycles.
The approach of the limits is based on the decrease of the capital for the insurance contracts
by the difference between the level of expected losses and the limits of the individual
insurance policies purchased by the institution.
The limits represent a measure of the maximum amount of risk that can be transferred to
the insurer; the premiums paid representing the proportion of the risk applied to the
expected loss. Policy limits lower than the insurance premium must be that the value of the
policy limits is related to the unexpected losses that are transferred through the policies.
Thus, the value of the risk transfer reduction is determined according to the formula:
𝐾 ∑ ∈ 𝐿 𝑃 𝐶𝐵 𝐶𝑅 (3)
This method has a number of strengths, such as: it provides higher limits for risk reduction
that reduce the amount of economic capital required; Recognizes that the reduction in
economic capital is more appropriate to come from higher coverage limits; applies both to
specific standard risk policies, but also to multilateral ones. At the same time, it has a
number of weaknesses, such as: it requires a differentiation between aggregate and single
loss limits; can cover arbitrage opportunities for high limits, extremely low hedging
probabilities if the limits are bought at values higher than unexpected losses.
The calculation methods for the Standardized approach are identical, except that the
decrease in capital by transfer of risk is determined after the sum of the capital for each line
of business, it is not necessary to allocate insurance at the level of the line of business.
In the case of approaching the internal evaluation, two methods will be distinguished for
diminishing the capital through insurance, namely: approaching the premiums and
approaching the limits.
The premium approach in which the insurance premium is considered an indicator of the
risk transfer exposure that multiplies with the expected loss reduction determined by the
bank's data, multiplied by a second factor, a sector factor that describes the relationship
between expected loss reduction and insurance. The insurance premium is used as an
indicator of the exposure to the transferred risk, multiplied by the reduction of the expected
loss. Thus the capital after insurance is determined according to the formula:
𝐾 𝐸𝐼 𝑃𝐸 𝐿𝐺𝐸 𝛾 𝐸𝐿 𝛾 𝑈𝐿 (4)
The capital reduction is given by the relationship:
𝐾 𝐸𝐿𝑅 𝛾 (5)
The net capital, which includes insurance, is calculated with the relationship:
𝐾 𝐾 𝐾 (6)
The approach has a weak point such as the factor 𝛾 which must be determined at the
industry level, but also a number of advantages: the loss coverage at individual or aggregate
level can be calculated using a single formula, the policy coverage is automatically taken
into account implicitly by the insurance price, the methodology is consistent with
determination of gross capital.
Approaching the limits within which the capital decrease is based on the limits of the
insurance policy. If the policy does not fully cover the risk segment (k), then an additional
coverage factor (CB) must be introduced to adjust the residual risk.
Both approaches are based on the approximation of the risk transferred to the insurer by
the premium paid, respectively the difference between the policy limit and the premium
paid. In both cases the value of the insurance is determined separately and then deducted
from the aggregate loss.
The benefit of implementing the loss distribution approach is the real replication of the
bank's risk profile, including the reduction of the risk affected by insurance, being
necessary a development of sophisticated models and the collection of substantial data.
Incorporating insurance into the aggregate loss distribution changes the net distribution of
the loss by reducing the severity of the loss from the risk transfer through insurance, but
the frequency is not affected. The basis of the LDA model is when the frequency and
severity curves are combined by simulation, each individual point of loss can be compared
with the specific insurance policy purchased by the institution, the policy limit and the
deductibility.
The transfer of risk through insurance products changes the distribution of the aggregate
loss by reducing the severity of the loss exceeding the deductible value of the policy, but
does not affect the frequency of the loss. By combining the frequency and severity
distribution by simulation, each individual loss can be compared with the purchased
insurance policies and the corresponding limits and deductibility policy.
To incorporate the insurance in the LDA model, the risk transfer policies must be applied
to the risk classes and business lines, and for each individual policy the value of the policy
limit, deductibility or retention and their type must be applied.
Insurance, as a method of financing operational risks, has a number of shortcomings.
Unlike the banking sector, the insurance market has insufficient capitalization to meet the
growing demand, which can lead to the situation where the insurer cannot provide
compensation in case of a large event. In reality, large losses are not covered by insurance,
so there is a discrepancy between the desired theoretical coverage and the coverage valid
in the market, illustrated in the following graph:
Figure 3. Theoretical coverage versus actual coverage
Each insurance contract has conditions, limits and exclusion clauses, and thus there may
be differences between the amount of the loss and the amount of compensation. The
specific risk of the insurance creates gaps between the coverage compared to the value of
the capital. The definition of coverage may be inconsistent with the internal operational
risk measure or the credit institution's risk management program.
The insured amount is difficult to determine due to the insufficiency of historical data on
operating risk losses and risk classification. By choosing risk financing through insurance,
institutions may be tempted not to improve their operational risk management, or on the
contrary it can be considerably improved, as insurers will select their clients according to
the application of appropriate operational risk management standards. One way to reduce
moral hazard is to set a high level of franchise. Due to the information asymmetry,
insurance companies do not know exactly whether or not credit institutions are safe, so they
are tempted to increase premiums, which will lead to lower demand for insurance, and
reduce product categories.
Figure 4. Insurance coverage
It is necessary to correctly frame the bank's exposure because in its absence, the banking
institution, in case of setting the risk premium only according to frequent losses and low
severity, will only aim to reduce them, and in case of significant loss could cause significant
damage.
Figure 5. Exposure to operational risk
If the insurer cannot rigorously assess the risks related to each client, he charges the same
premium for all policyholders intervening the issue of adverse selection. If the insurer
incurs very large losses, it may withdraw from the market and no longer offer insurance to
the banking sector against operational risks. For the rigorous risk assessment and for
determining the insurance premium, the insurer can use the techniques of the rating
agencies, which consist in:
 business risk assessment (aspects such as management quality, strategy, market position
are taken into account);
 financial risk assessment (the financial situation of the organization over a longer period
and the medium-term projections are analysed, and these indicators are compared with
the sector average);
 or the overall assessment of the organization (includes the potential impact of external
factors).
Insurers can use reinsurance to transfer some of the risks they have taken out, which is a
dissatisfaction to the insured institutions because:
 the collection of compensation requires a long time (because the insurer or reinsurer can
or is able to honour the payment only after obtaining the necessary funds from the next
reinsurer, in some cases the reinsurer pays only after the insurer has paid the indemnity,
sometimes creating liquidity problems for the insurer);
 there may be counterparty risk against a reinsurer that the institution does not know (if
the reinsurer goes bankrupt, the bank has the legal right to appeal against the original
insurer, but if the initial insurer is the bankrupt, the bank cannot go against the
reinsurer);
 concentration of risk on an unknown counterparty (which, if it knew, the bank would
not accept as a business partner).
Delays in payment of claims may occur if the exact value, time, place and cause of the loss
cannot be determined, or when the risk produced is covered by several insurance contracts
and may even lead to the loss of the entire activity of the insured. Delays in payment of
compensation are almost inevitable, and so this seems to be the main disadvantage of
insurance, compared to capital, which is valid almost instantly. Thus, as a solution to this
inconvenience is that the insurance policies include an advance payment without delay
(FIORI mechanism of acquisition of shares), which can be used only by very large
institutions and thus will limit the incentives to use insurance.
As insurance becomes a normal practice, in the case of paying a very high indemnity, the
insurer's bankruptcy may occur and thus the rest of the insured institutions would remain
exposed, which will lead to the manifestation of systemic risk. In case of catastrophic losses
the desired coverage will be different from the possible one. Against operational risks, as
self-insurance alternatives, mutual funds for self-insurance, securitization or risk plans with
limited duration can be used.
Mutual self-aspiration funds are agreements concluded by several institutions to set up
funds that are used to cover possible losses, which insurance companies do not accept or
for which they request exaggerated insurance premiums. The disadvantages of these funds
are the occurrence of cases of adverse selection, moral hazard, or cases where an institution
does not want to save a rival. According to the British Bankers' Association, for the fund's
solution to be effective, it must have at least 30 members.
The transformation of assets into securities traded on the market transfers operational risks
to investors in the capital market, offering a lower counterparty risk and an upper hedging
limit, due to the dispersion of risk and the large volume of resources available for
investments (e.g. issuing of obligations).
Limited-risk risk plans are self-financing programs through which the institution sets up an
account managed by an external agent through which it will periodically pay the premiums
that will be capitalized. If the available money in the account is exhausted, the external
agent can offer the bank a financing line.
Figure 6. Impact of operational risk on capital before and after using insurance for operational risk
It is estimated the expected loss, the average loss distribution, represented on the Ox axis
of point A, and all losses after this point will be absorbed by income. Point B on the Ox
axis represents the maximum possible loss, so provisions will be made so that it can absorb
the losses between points A and B. Point C is the point where the company becomes
insolvent.
The message from the Basel Committee shows that insurance protection is useful, but does
not relieve banks of the responsibility of creating internal risk control systems,
recommending that they also adopt counterparty risk limitation policies.
Conclusions
From the analysis made by the authors in the article Methods used in risk financing, a series
of theoretical and practical conclusions can be drawn. Thus, risk management and control
programs are not fully effective, so methods of financing losses through retention or
transfer have been developed.
Risk minimization methods depend on the resources available, the cost of risk mitigation,
the concepts and the time allocated to the implementation of the processes.
Another conclusion that emerges from the study undertaken by the authors is that the Basel
Committee on Banking Supervision recognizes the benefits of insurance, as an effective
risk management tool, to reduce losses incurred by banks due to operational risk, which
materializes in a much better definition of risk-taking.
Another conclusion is that adequate risk management leads to improved performance
indicators and cash flows and thus to an allocation of the minimum capital needed to cover
losses.
A final conclusion is that there are a number of disadvantages of insurance such as poor
capitalization of this market, difficult determination of the insured amount, partial or late
payment of compensation, withdrawal of insurers from the market which led to the
emergence of alternative systems such as securitization, self-financing programs or self-
insurance mutual funds.
References
Anghelache, C., Anghel, M.G., Diaconu, A. and Lilea, F.P.C., 2017. Operational risk – model of
analysis and control, Romanian Statistical Review, Supplement, 11, pp. 102-107.
Anghelache, C., Anghelache, G.V., Anghel, M.G. and Niţă, G., 2016. General Notions on banking
Risks, Romanian Statistical Review, Supplement, 5, pp. 13-1.
Awdeh, A., Moussawi, C. and Machrouh, F., 2011. The Effect of Capital Requirements on Banking
Risk. International Research Journal of Finance and Economics, 66, pp. 133-146.
Cipovova, E. and Dlaskova, G., 2016. Comparison of Different Methods of Credit Risk Management
of the Commercial Bank to Accelerate Lending Activities for SME Segment. European
Research Studies, 19 (4), pp. 17-26.
Doerig, H.U., 2000. Operational Risks in Financial Services: An Old Challenge in a New
Environment. Switzerland: Credit Suisse Group.
Geiger, H., 2000. Regulating and Supervising Operational Risk for Banks, Swiss Banking Institute,
Working Paper no. 26.
Hakens, H. and Schnabel, I., 2010. Credit Risk Transfer and Bank Competition. Journal of Financial
Intermediation, 19 (3), pp. 308-332.
Harmantzis, F., 2002. Operational Risk Management in Financial Services and the New Basel
Accord, working paper, Stevens Institute of Technology.
Hall, S. and Howell, J., 2001. Insurance of operational risk under the new Basel capital accord, A
Working Paper submitted by Insurance Companies.
Kuritzkes, A., 2002. Operational Risk Capital: A Problem of Definition. The Journal of Risk
Finance, 4 (1), pp. 47-56.
Miller, P.G., 2014. The Role of Risk Management and Compliance in Banking Integration, New
York University Law and Economics Working Paper, 11, pp. 1-26.
Peters, G., Shevchenko, P.V., Mario, V. and Wüthrich, M.V., 2009. Dynamic Operational Risk:
Modelling Dependence and Combining Different Sources of Information. Journal of
Operational Risk, 4 (2), pp. 69-104.
Thomas, B.B. and Ware Preston, L., 2008. Making More Operational Risk Insurable, Transurance
Services III.
Wendy, D., 2001. Insurance of Operational Risk and the New Basel Capital Accord, Capital
Allocation for Operational Risk Conference Boston, pp. 14-16.
Is there a relationship between economic welfare

and innovation performance?
Evidence from selected European countries
Lejla TERZIĆ
University of East Sarajevo, Bosnia and Herzegovina
lejla.terzic.efb@gmail.com
Abstract. Investigating economic welfare and innovation performance have been the main themes
of dominant researchers in the area of economics. Despite the fact the imperfection of Gross
Domestic Product to evaluate welfare is generally acknowledged in economic theory, it is still
applicated as one of the crucial indicators for creation of economic policy directions. Interest
related to the constraints of GDP have increased to the several attempts in the past decennium to
create aggregate economic welfare indicators as substitutes or augmentations to the GDP. A
number of challenging indicators have been created, although GDP remaines to be applicated as
the traditional indicator of economic welfare. The outstanding of these prominent economic
welfare indicators are: the Genuine Progress Indicator, the Index of Sustainable Economic
Welfare, the Human Development Index and the Better Life Indicator. This paper's main goal is to
determine relationship between the prominent indicators of economic welfare and innovation
performance in selected European countries. In order to examine the intercorrelations between
economic welfare indicators and innovation performance several methodologies have been
applicated. Research was conducted via statistical software package SPSS 25. The research
results have indicated that the relationship between indicators of economic welfare in selected
European countries are significantly associated with their innovation performance.
Keywords: economic welfare indicators, innovation performance, relationship, selected European

countries.
JEL Classification: C8, E0, O30, O57.

160 Lejla Terzić
1. Introduction
The idea for this investigation came from revealing the mystic tenacity of welfare
measurement that variously influenced innovation performance in the selected European
countries, a question that had not sustain appropriate deliberation in last decade by policy
creators and government. Gross Domestic Product (GDP) is an indicator of economic
achievement. Although GDP was not at all designed to be an indicator of well-being, it is
still commonly applied as indicator reversing the evident require for a scope of economic
welfare. The feasible relevance of indicators of public welfare can be emphasized. Policy
procedures alternatives by authorities and entire community, indicators of economic
growth and comparations among different countries, all indicate to measurements of
particular and common welfare. It is occasionally debated that modifications in GDP are
extremely associated with differences in economic welfare, although this disregards the
issue that whenever what we evaluate is significant, thinking out literally evaluating
substances it would not gain remarkably consideration.
This article is not concentrated as an exhaustive survey of the actual methods for
evaluating welfare and innovation, although it assures an overview of the prominent
crucial indicators in the economic welfare measurement. This paper has considerated
some new improvements in methods to evaluating welfare that were conffered by the
Index of Sustainable Economic Welfare (ISEW), the Genuine Progress Indicator (GPI),
the Human Development Index (HDI) and the Better Life Indicator (BLI).
Modern times are notable by several threats to well-being that society needs to deal with.
These are the threats of climate diversification, increasing income disparities, excalations
of conflicts and human rights abuse. Even though peace and recognition for people
freedom cannot be integrated in a particular-value indicator of welfare, every country
requires an accumulated indicator of economic welfare that is apprehensive of the
questions of environmental hazard and ultimate discrimination. Furthermore, every
country needs an indicator that integrates non-market donors to welfare, which could be
debased by performances on the market. Between the prominent new indicators of
economic welfare, only Genuine Progress Indicator (GPI) trails every threat and
aggregates non-market donors to welfare. Accordingly GPI is the best indicator to trail
economic activities and welfare during the observed period. Nonetheless, exactly because
the GPI is more extensive than other beyond GDP indicators, that does not assure that it
would be properly embraced for economic policy application.
The primary goal of this paper is to explore the relationship between indicators of
economic welfare and innovation performance in selected European countries. This paper
incorporates four parts.
The first part of paper presents theoretical synopsis of the literature related to new
“Beyond GDP indicators” in measuring economic welfare and innovation.
In order to examine the intercorrelations between economic welfare indicators (the
Genuine Progress Indicator, the Index of Sustainable Economic Welfare, the Human
Development Index and the Better Life Indicator) and innovation (Global Innovation
Is there a relationship between economic welfare and innovation performance? 161
Index) several methodological measurement devices have been applicated in the second
part of paper.
The third part of the paper is dealing with gathered data (primary and secondary) and
research methodology.
The fourth part presents findings from Indicators of Economic Welfare and Global
Innovation Index and research applicated by statistical software package SPSS 25.
2. Theoretical synopsis of the literature: Economic welfare indicators and innovation

performance
Economic welfare research generally exhibits special concern in economic growth.
Economist Ding (2014) has explored the possible destructive influence of welfare
expense on economic growth while the other stream of researches debate that economic
welfare spending could be helpful as a crucial catalyst for increasing growth (Lindert,
2003).
Correspondingly, one more stream of economic thought, that is usually attributed to as
the economic school of wage-driven growth, concentrates on the influence of
consumption created by the income enhancement culminating from economic welfare
(Onaran et al., 2017). The welfare state can have a significant impact in creating the
innovative capacity of a country (Hall, 2015), and suggest that economic welfare may
play a hidden role in stimulating innovation performance.
Essential to the criticism of GDP is its mistreatment as a gauge of welfare. Its actual
application as a gauge of market activities is commonly approved and infrequently
investigated, nevertheless appeals for modifications to superior apprehension and
configuration of new products and services which infiltrate the world market, especially
the digital economics (Stiglitz et al., 2009; Coyle, 2014; Jorgenson, 2018).
Notwithstanding, associating market performances with economic welfare over policy
converse, along with the media communications emphasizes relevant knowledge
divergences in percepting of the non-market donors to well-being, functioning as
voluntary care labour and the surroundings, or the performances that decrease from
welfare, e.g. climate deterioration and resource deficiency.
Additionally, in what way economic activities are evaluated in domestic income as input
or output aftermaths in an inaccurate productivity impression. Aforementioned is because
state authorities-supported services (e.g. education, health care) are measured on the
foundations of inputs applied to assure these services (e.g. numbers of doctors, number of
professors...) comparatively then on the real made outputs (e.g. numbers of medical
therapies, number of conferences...). Subsequently, if productivity increases in the public
sector, then the GDP indicators undervalue economic growth.
Besides abovementioned experts have highlighted that GDP augments welfare by
describing each disbursement as the donor to economic welfare, beyond differentiating
among welfare-increasing and welfare-decreasing performances. Many analysts
162 Lejla Terzić
emphasize that GDP is inadequate to evaluate sustainability in circumstances of income

and well-being together. Economists Nordhaus and Tobin in 1972 created the Measure of
Economic Welfare (afterwards the Index of Sustainable Economic Welfare) – a
comprehensive indicator of the actual consumption of households each year, composing a
few modifications to Gross National Product.
Several specialists also debate that GDP is essentially section of specific political rules
which stimulate modifications, as a result of its attention on market performances.
Accelerated dependence on GDP as an indicator of welfare requests to concerns that
favors enlargement of market alliances, the era of creating profit-generating chances as
apprehended by the GDP growth. Furthermore, that requests to affect which attempt to
make negative externalities unseeable (Costanza et al., 2009). Anyhow, this element of
the GDP and its accordance with the capitalism aims can interpret its maintained
preeminence as an indicator of economic welfare (Felice, 2016).
The awareness on the imperfections of GDP as an indicator of economic welfare has been
increased inordinately in past decennium and has boosted attempts to create new
indicators so called “Beyond GDP indicators” (Stiglitz et al., 2009; Redermacher, 2015;
Durand, 2013, Hayden and Wilson, 2018). Several analysists have focused their attention
on improvement of the Genuine Progress Indicator methodology by creating a guide for
policymaking toward sustainability (Talberth, 2007; Beneria et al., 2015; Berik, 2018).
Genuine Progress Indicator (GPI) is an indicator created to take adequate account of the
nation's well-being, by integrating environmental and social components that are not
evaluated by Gross Domestic Product. GDP measures income, not decency, it also
evaluates economic welfare, but not desolation, and it rejects social cohesion and the
natural environment (OECD, 2019). GDP presents an unique comprehensive tool that is
relevant for comparative analysis between countries. Alternative indicator that tried to
envelop the social dimension previously attains, although the GDP is commonly
applicated as the basic gauge.
The most widely operated indicator is the Human Development Index (HDI). HDI is an
aggregated indicator that incorporates dispersed indicators for the following dimensions:
life assurance (gauge of nation's health and endurance); nation's education, and living
standard, as measured by the GDP per capita at purchasing power parity (GDP PPP pc).
The primary goal of the popular innovation performance indicator – Global Innovation
Index (GII), that was established by the World Intellectual Property Organization
INSEAD – WIPO, is to reveal the flatten to which economies are answering to the
innovation requires. The GII is created from the 84 components divided into eight
aggregated gauges that are grouped as five input gauges and three output gauges. The five
input gauges include: institutions and policies, infrastructure, human capacity,
technological sophistication, business markets and capital (INSEAD – WIPO, 2019).
These gauges represent variables that are boosting the innovation capacity. The three
output gauges assimilate competitiveness, knowledge and prosperity. The GII engages
creditable data constrained from various associations (e.g. the World Bank, Organisation
for Economic Cooperation and Development, etc.), and subjective information

constrained from the Opinion Survey.
3. Data and methodology

To explore relationship between indicators of economic welfare and innovation in
selected European countries several methodological accesses and indicators, established
by respectable international organisations, have been incorporated:
 the Genuine Progress Indicator and the Index of Sustainable Economic Welfare, as the
macroeconomic indicators of System of National Accounts. The GPI is evaluated by
the 26 indicators that could be separated into three primary dimensions: Economic,
Environmental, and Social dimension;
 the Index of Sustainable Economic Welfare, developed by William Nordhaus and
James Tobin;
 the Human Development Index, established by United Nations Development
Programme (UNDP) methodology;
 the Better Life Indicator, created by Organisation for Economic Cooperation and
Development, includes several „dimensions” of well-being: Housing, Income, Jobs,
Community, Education, Quality of environment, Governance: (involvement in
democracy), Health, Life Satisfaction (level of happiness), Safety and Work-life
balance;
 the Global Innovation Index (GII), created by INSEAD – WIPO (World Intellectual
Property Organization) methodology.
The data that are used to calculate the GPI and ISEW for selected EU countries came
from the World Bank group national accounts database for abovementioned countries.
ISEW is an indicator that incorporates the traditional indicator of macroeconomic activity
(Gross Domestic Product – actually, the one part of it – personal consumption) with
supplementary environmental and social components included. It is necessarily to create
new groups of gauges that interfere the traditional quantitative measurement of economic
welfare to absorb qualitative components.
The formal expression proposed for the calculation of ISEW is shown in the following
formula:
ISEW = C(wpce) + G(ndge) + K + L(h) − N – D
where:
C(wpce) is the weighted private consumption expenditure;
G(ndge) is the non-defensive governmental expense;
L(h) represents the household labour;
K is the capital adjustment;
N is the reduction of natural environment and
D represents the defensive private expenditure on health care, education and social costs.
164 Lejla Terzić
Generational economic welfare is presented by the altered life satisfaction of actual and
new generations. Welfare is obtained if dEW (t ) / dt  0 . Economic welfare at time (t) is
presented by following formula:

 
EW (t )   U (C ( p))e  ( p t ) dp
t
where: EW(t) is economic welfare in the specific time (t),  represents life satisfaction
discount rate,   0 , C ( p ) represents consumption vector at time p.
Recently, the relevance of innovation for economic welfare has animated many explorers
to investigate its variables. Very popular model which describes the innovation
performance drivers in a specific economy is the knowledge – production function model.
This model, that indicates the connection among Research & Development (R&D) and
innovation scope, was afterwards examined by the national innovation system approach.
According to the national innovation system approach, connections between the
industries, academic communities, international organisations and government are
relevant components of a system that is regulated toward innovation performance
(Lundvall, 2006).
It accentuated the effect of innovation performance on the economic welfare. This model
may be demonstrated by separating specific economies into two groups. The first group
represents creation of outputs and the second division presents research and development
that leads to country's innovation performance. The following equations could be applied
to certain two groups of the economies:
Z  IPK z Lz H 1z    0 < α < 1; 0 <  < 1.

IP  IP  K IP LIP H IP
1  
0 < < 1; 0  < 1.
where:
variable z is the quantity of determinant used for production activities,
K and L represent capital and labor,
H is human capital,
variable IP presents the quantity of gauge implicated for Research and Development,

I P amplifies the innovation performance that is developed by the R&D division.
4. Findings from indicators of economic welfare and global innovation index:

A comparative analysis
Table 1 presents the indicators of economic welfare: the Genuine Progress Indicator, the
Index of Sustainable Economic Welfare, the Human Development Index, the Better Life
Indicator and the Global Innovation Index rankings in selected European countries in 2019.
The investigation was conducted in the following countries: Czech Republic, Hungary,
Poland, Slovakia, Croatia, Estonia, Latvia, Lithuania, Romania and Slovenia in the 2019.
Table 1. Rankings of the selected European countries in 2019, according to the Indicators of Economic
Welfare and Global Innovation Index
GPI ISEW HDI BLI GII
Czech Republic 9 7 26 22 26
Hungary 27 25 43 31 33
Poland 28 29 32 27 39
Slovakia 26 27 36 26 37
Croatia 20 22 46 47 44
Estonia 11 10 30 21 24
Latvia 25 24 39 32 34
Lithuania 33 32 34 28 38
Romania 41 42 52 54 50
Slovenia 15 12 24 20 31
Note: GPI – the Genuine Progress Indicator, ISEW – the Index of Sustainable Economic Welfare, HDI-the
Human Development Index, BLI-the Better Life Indicator, GII – Global Innovation Index.
Source: the World Bank Group national data base (WBG, 2019), UNDP Human Development Indicators
2019, OECD Better Life Index 2019, the Global Innovation Index Report 2019, INSEAD – WIPO (World
Intellectual Property Organization).
Table 2 declares the economic welfare and innovation performance rankings in selected
European countries in 2019. Czech Republic has accomplished the highest rank according
to Gross Domestic Product PPP per capita, the Genuine Progress Indicator and the Index
of Sustainable Economic Welfare. Slovenia is the best positioned country according to
the Human Development Index and the Better Life Indicator.
Table 2. Economic welfare and innovation performance rankings in the selected European countries in 2019
GPI ISEW HDI BLI GDP GII
PPP
p.c.
Czech Republic 1 1 2 3 1 2
Hungary 7 6 8 7 6 4
Poland 8 8 4 5 7 8
Slovakia 6 7 6 4 4 6
Croatia 4 4 9 9 10 9
Estonia 2 2 3 2 5 1
Latvia 5 5 7 8 8 5
Lithuania 9 9 5 6 3 7
Romania 10 10 10 10 9 10
Slovenia 3 3 1 1 2 3
Note: GDP PPP p.c. – Gross Domestic Product Purchasing Power Parity per capita.
Source: Author's own calculation.
Estonia has attained the highest position conferred to innovation performance in

comparison to the analyzed European countries. Croatia is the lowest positioned country
by GDP PPP per capita, while Romania is the lowest classified country according to
Genuine Progress Indicator, Index of Sustainable Economic Welfare, the Human
Development Index, the Better Life Indicator and the Global Innovation Index.
The relationship between economic welfare and innovation performance indicators are
presented in Table 3. Spearman’s correlation coefficients have revealed the
interconnection among crucial variables of economic welfare (GPI – Genuine
Progress Indicator, ISEW – Index of Sustainable Economic Welfare, HDI – Human
Development Index, BLI – Better Life Indicator) and innovation performance indicator
(GII – Global Innovation Index). The necessary data for this research were collected from
166 Lejla Terzić
constitutional and secondary sources. The investigation was carried out by applying the
SPSS 25 statistical software package.
Table 3. Relationship between different economic welfare indicators and global innovation index
GPI ISEW HDI BLI GDP GII
PPP
p.c.
GPI 1,000 .988** .588 .612 .406 .758*
ISEW .988** 1.000 .564 .576 .382 .782**
HDI .588 .564 1.000 .939** .818** .709*
BLI .612 .576 .939** 1.000 .818** .770**
GDP PPP p.c. .406 .382 .818** .818** 1.000 .673*
GII .758* .782** .709* .770** .673* 1.000
**Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
Source: Author's own calculation.
Investigation results have determined strong and significant relationship between the
Index of Sustainable Economic Welfare (ISEW) and the Global Innovation Index (GII)
conferred by correlation coefficients 0.782. Very strong positive intercorrelation is
revealed among Genuine Progress Indicator (GPI) and ISEW followed by correlation
coefficients 0.988. The positive interdependance is illustrated by the same correlation
coefficients 0.818 between the Better Life Indicator (BLI), the Human Development
Index (HDI) and Gross Domestic Product PPP per capita (GDP PPP pc), respectively.
Very strong positive correlation is revealed among BLI and HDI (0.939). The
intercorrelation determined among BLI and GII (0.770) illustrates that countries can gain
higher position regarding innovation performance by creating adequate climate for
economic environment, and approved circumstances for better life satisfaction. The level
of economic welfare between selected European countries by different “beyond GDP”
indicators, indicates that ISEW and BLI highly correlate with their innovation
performance, which is presented by high correlation coefficients.
Conclusion
The main goal of this paper was to determine the relationship between economic welfare
and innovation in the selected European countries. The investigation results have revealed
positive interconnections among the Index of Sustainable Economic Welfare (ISEW) and
the Global Innovation Index. In regard to the determined relationships, it can be
concluded that sustainable economic welfare strongly depends on innovation per-
formance. The results of this investigation confirmed that variations between selected
European countries may be intercorrelated with the variables of innovation performance,
human development and economic welfare. Apparently, the practices of the analyzed
high-ranked European countries are effective for the rest of the countries in boosting
economic welfare and innovation capability.
The utilization of the appropriate government policies could increase, along with
developing business environment, human development, higher innovative scope, better
life standard, and accelerated economic growth. The GPI, ISEW, HDI, BLI and GII
indicators are valuable devices to strengthen partnership among economists, governments
and respectable world institutions. Certainly, usage of prominent “beyond GDP”

indicators arouses remarkable new access into the mystic relationship between welfare
and innovation performance in the analyzed countries.
References
Anielski, M., 2001. Measuring the sustainability of nations: The Genuine Progress Indicator
system of sustainable well being accounts, The Fourth Biennial Conference of the Canadian
Society of Ecological Economics: Ecological Sustainability of the Global Market Place,
pp. 1-52, Available at <http://anielski.com/Documents/Sustainability%20of%20Nations.pdf>
[5 January 2020].
Benería, L., Berik, G. and Floro, M., 2015. Gender, development, and globalization: Economics as
if all people mattered, 2nd Edition, London and New York, Routledge.
Berik, G., 2018. To measure and to narrate: Paths toward a sustainable future, Feminist Economics,
Available at <https://doi.org/10.1080/13545701.2018.1458203> [5 January 2020].
Bleys, B. and Whitby, A, 2015. Barriers and opportunities for alternative measures of economic
welfare, Ecological Economics, Vol. 117, pp. 162-172.
Costanza, R., Hart, M., Talberth, J. and Posner, S, 2009. Beyond GDP: The need for new measures
of progress. The Pardee Papers No. 4, Boston University. Available at <https://pdxscholar.
library.pdx.edu/cgi/viewcontent.cgi?article=1010&context=iss_pub> [5 January 2020].
Coyle, D, 2014. GDP: A brief but affectionate history, Princeton University Press.
Ding, H, 2014. Economic growth and welfare state: A debate of econometrics. Journal of Social
Science for Policy Implications, 2(2), 165-196
Durand, M, 2013. The OECD Better Life Initiative: How’s life? and the measurement of well-
being, Review of Income and Wealth, Vol. 61, No. 1, pp. 4-17.
Easterlin, R.A, 1995. Will raising the incomes of all increase the happiness of all?, Journal of
Economic Behavior and Organization, Vol. 27, pp. 35-47.
Felice, E, 2016. The misty grail: The search for a comprehensive measure of development and the
reasons for GDP primacy, Development and Change, Vol. 47, No. 5, pp. 967-994.
INSEAD WIPO, 2019. The Global Innovation Index Report, 2019. Fontainebleau, Ithaca, and
Geneva.
Jorgenson, D.W, 2018. Production and welfare: progress in economic measurement, Journal of
Economic Literature, 56, 3, pp. 867-919.
Jorgenson, D.W. and Schreyer, P, 2017. Measuring individual economic well-being and social
welfare within the framework of the system of national accounts, Review of Income and
Wealth, 63, Supplement 2, pp. S460-77.
Hayden, A. and Wilson, J, 2018. Taking the first steps beyond GDP: Maryland’s experience in
measuring ‘genuine progress’, Sustainability 10 (462), pp. 1-24.
Hall, P, 2015. Varieties of capitalism. In R. Scott and S. Kosslyn (Eds.), Emerging trends in the
social and behavioral sciences, Wiley Online: John Wiley & Sons, pp. 1-15.
Helliwell, J., Layard, R. and Sachs J, 2018. World Happiness Report 2018, New York: Sustainable
Development Solutions.
168 Lejla Terzić
Lawn, P, 2003. A theoretical foundation to support the Index of Sustainable Economic Welfare
(ISEW), Genuine Progress Indicator (GPI), and other related indexes, Ecological Economics,
Vol. 44, pp. 105-118.
Lindert, P.H, 2003. Why the Welfare State Looks Like a Free Lunch? NBER Working Paper
No. 9869, Issued in July 2003.
Lundvall, B.A, 2006. Interactive learning, social capital and economic performance, in Foray, D.
and Kahin, B. (eds.), Advancing Knowledge and the Knowledge Economy, Harvard
University Press, US, pp. 63-74.
Mazzucato, M, 2018. The value of everything: Making and taking in the global economy, United
Kingdom, Allen Lane, Penguin Books.
Nordhaus, W.D. and Tobin, J, 1972. Is growth obsolete? in Economic Research: Retro-spect and
Pro-spect, Volume 5: Economic Growth, New York: National Bureau of Economic
Research, pp. 1-80.
New Economics Foundation. 2019. Happy Planet Index 2019: Methods Paper. Available at
<http://happyplanetindex.org/about> [5 January 2020].
Onaran, O ̈ ., Andersen, L., Cozzi, G., Dahl, S., Nissen, T., Obst, T. and Tori, D, 2017. An
Investment and Equality-Led Sustainable Development Strategy for Europe. University of
Greenwich, London.
Oulton, N, 2012. How to measure living standards and productivity, Review of Income and
Wealth, 58, 3 (September), pp. 424-56.
OECD, 2019. Better Life Index, Available at <http://www.oecdbetterlifeindex.org/about/better-
life-initiative/ accessed December 10, 2017> [5 January 2020].
Piketty, T, 2014. Capital in the Twenty-First Century, Cambridge, Mass.: Belknap Press of
Harvard University Press.
Radermacher, W.J, 2015. Recent and future developments related to ‘GDP and Beyond’, Review
of Income and Wealth, Vol. 61, No. 1, pp. 18-24.
Ravallion, M, 2012. Mashup indices of development, World Bank Research Observer, 27 (I),
pp. 1-32.
Stewart, F, 2014. Against happiness: A critical appraisal of the use of measures of happiness for
evaluating progress in development, Journal of Human Development and Capabilities,
Vol. 15, No. 4, pp. 293-307.
Stiglitz, J.E., Sen, A. and Fitoussi J-P, 2009. Report by the Commission on the Measurement of
Economic Performance and Social Performance. Available at <http://ec.europa.eu/eurostat/
documents/8131721/8131772/Stiglitz-Sen-Fitoussi-Commission-report.pd> [5 January 2020].
Talberth, J., Cobb C. and Slattery N, 2007. The Genuine Progress Indicator 2006: A Tool For
Sustainable Development, Oakland, CA, Redefining Progress.
UNDP, 2019. Human Development Indicators 2019, Washington, DC: UNDP.
WBG – World Bank Group, 2019. National databases, [online]. Available at
<https://data.worldbank.org/> [Accessed 15 Feb. 2020]
Analyzing Romania GDP:

Final consumption, gross investment, and net exports
influence compared to previously published models
Alexander I. VILLANUEVA
Texas Tech University, USA
alexvill89@yahoo.com
Abstract. This paper analyzes final consumption, gross investment, and net exports influence on
gross domestic product (GDP) for Romania from 1990-2014. The results show final consumption
and gross investment being the dominant influences during this time period, whereas net exports
had a minimal yet statistically significant effect. This paper also compares the model created against
a previously published model to determine which is best suited for this type of analysis.
Keywords: gross domestic product (GDP), consumption, investment, net exports, multiple linear
regression, Romania.
JEL Classification: C51, C52, E21, E22, F10.

170 Alexander I. Villanueva
Introduction
Gross domestic product (GDP) is one of various key macroeconomic indicators often
utilized in research when analyzing the health of a country’s economy (Wolla, 2018).
Reason being is that GDP is one measurement of a country’s economic activity, specifically
measuring the total value of all goods and services produced in a country during a given
year. Accordingly, GDP data is widely analyzed in determining whether an economy is
expanding or contracting. Based on which of the two a country is experiencing,
recommendations and decisions are then discussed, and used to map out economic policy
(Kira, 2013). The focus of this paper is regarding the analysis of Romania GDP post-
communism through the utilization of linear regression models.
Previous literature constructed simple linear regression models to determine possible
correlations between a variable of interest – chosen at the discretion of researcher – and
Romania GDP (Ioneci and Mîndreci, 2010; Anghelache, 2011). Simple linear regression
models are represented by the following equation:
𝑌 𝛽 𝛽𝑋 𝑈 (1)
where Yi is the dependent variable; Xi is the explanatory variable that influences the
dependent variable; and Ui is the error term representing unobservable factors other than
the included explanatory variable that affects the dependent variable. As stated previously,
prior research has applied Equation (1) to examine relationships between two specific
variables. Ioneci and Mîndreci (2010) analyzed the effect of foreign direct investment on
Romania GDP. Their analysis determined that there is a direct and strong relationship
between the two variables indicating investments positively affect Romania GDP.
Anghelache (2011) identified a relationship between Romania GDP and final consumption
(FC): an increase in FC (defined by Anghelache as the sum of private consumption and
public consumption) leads to an increase in Romania GDP.
Although simple linear regression models result in establishing whether two variables have
a relationship or not, there is a shortcoming due to the error term. This term, as stated
previously, represents factors other than independent variable that affects the dependent
variable. In the simple linear regression model, it is assumed the factors contained in the
error term are unobserved and therefore unable to be effectively utilized. However, if the
error term does contain an observed variable, then the simple linear regression model
suffers from omitted variable bias. Continuing to use a simple linear regression model with
omitted variable bias presents the following issue: the coefficient estimates produced are
either underestimated or overestimated, in which case the relationship once established
between two variables is flawed. Therefore, it is ideal to extend a simple linear regression
model into a multiple linear regression model to counteract omitted variable bias.
Extension into a multiple linear regression model requires adding one or more observed,
relevant variable(s) into Equation (1) resulting in the following new equation:
𝑌 𝛽 𝛽𝑋 𝛽𝑋 𝛽 𝑋 𝑈 (2)
with k independent variables, and where the definitions provided in Equation (1) still
applies. Using the multiple linear regression model in Equation (2) is more appropriate
Analyzing Romania GDP: Final consumption, gross investment, and net exports influence 171
when analyzing economic trends, such as a country’s GDP, that are influenced by more
than one variable (Constantin, 2017). This concept was previously realized by Anghelache,
Manole, and Anghel (2015) who provided further analysis on Romania GDP by utilizing a
multiple linear regression model. They investigated the impact of FC and gross investment
(INVMT) on Romania GDP, whereby they concluded changes in either of the two variables
has a positive and significant influence on Romania GDP (Anghelache et al., 2015).
The purpose of this paper is two-fold. First, to provide a deeper insight into the analysis of
Romania GDP by adding a third relevant variable to the model Anghelache et al. (2015)
presented in their paper. Second, to compare the new model against the model Anghelache
et al. (2015) presented. In comparing the two models, it will be demonstrated which of the
two is more beneficial when analyzing Romania GDP.
Methodology and data

The multiple linear regression model being used in this paper is an extension of the model
created by Anghelache et al. (2015): their model included FC, INVMT (both as explanatory
variables), and Romania GDP (defined as the dependent variable). This paper’s model will
include these variables with the addition of net exports (NETEX). The inclusion of NETEX
is predicated on the following economic foundation: GDP has been traditionally measured
by using consumption, investment, government spending, and net exports (Wolla, 2018;
Kira, 2013). Furthermore, exports and imports have been shown to have a positively
significant effect on a country’s GDP (Kartikasari, 2017). The following equation
represents the model used in this paper:
𝑅𝑜𝑚𝑎𝑛𝑖𝑎 𝐺𝐷𝑃 𝛽 𝛽 𝐹𝐶 𝛽 𝐼𝑁𝑉𝑀𝑇 𝛽 𝑁𝐸𝑇𝐸𝑋 (3)
where: FC, INVMT, and NETEX are the explanatory variables; Romania GDP is the
dependent variable.
The analysis of Equation (3) is performed over the period 1990-2014 for the country of
Romania. Yearly data for FC, INVMT, and Romania GDP were obtained from within the
Anghelache et al. (2015) paper. They obtained their data from the Statistical Yearbook of
Romania through the National Institute of Statistics. The authors then transformed the
data to reflect million Romanian Leu (RON) while also deflating the values to reflect
1990-RON value (deflation to represent 1990-RON values was conducted at the
discretion of the authors using the consumer price index). To generate the values for
NETEX, Romania’s exports and imports had to first be obtained followed by subtraction
from each other in the following manner:
𝑁𝐸𝑇𝐸𝑋 𝐸𝑥𝑝𝑜𝑟𝑡𝑠 𝐼𝑚𝑝𝑜𝑟𝑡𝑠 (4)
Data for Romanian exports and imports were obtained from the World Bank’s World
Development Indicators database. After computing NETEX from 1990-2014, the values
were transformed to be measured in million RON, and then deflated to represent
1990-RON value (deflation was calculated by: 𝐶𝑃𝐼 , where n
represents the given year to be converted). Figure 1 contains graphical representations of

the transformed data for the four variables.
Upon analyzing the data, there appears to be a positive relationship between Romania GDP,
FC, and INVMT based on the similar patterns the three graphs are experiencing from 1990
– 2014. For the first few years from 1990 to about 1994, these three variables experienced
declines followed by a gradual increase. Each of the three variables are shown to have steep
increases from about 1996 until 2008 when there is a drop-off. On the contrary, NETEX
experienced opposite progression. It is shown to have a steep decrease beginning about
2002 until 2008 when it sharply increases, followed by a slight increase, and then progress
to another sharp increase until it almost flattens out at the end of the data in 2014.
The movements observed from Figure 1 in Romania GDP, FC, INVMT, and NETEX can
be attributed to three specific events: 1 – the fall of communism in 1989 with a subsequent
movement towards a democratic, market-economy (Doroftei and Păun, 2013; Bălă, 2014),
2 – joining the European Union in 2007 (Zaman et al., 2016), and 3 – the financial crisis of
2008 (Belaşcu and Budac, 2016). After the fall of communism, there is a period of decrease
in Romania GDP, FC, and INVMT until about 1993 reflecting the difficulty in transitioning
away from communism. The following year until about 1996 there is observed growth with
these three variables followed by another decrease until about the end of 1999. This
decrease is a result from a change in political leadership, upon which a subsequent political
leadership change at the end of 1999/beginning of 2000 resulted in a lengthy, gradual
increase in Romania GDP, FC, and INVMT until 2008 (Bălă, 2014).
During the entire period of the data set, majority of the increase observed in FC has been
attributed to an increase from household final consumption (Savu and Voicu, 2017).
Additionally, Bălă (2014) stated the transition from communism resulted in a growth from
66% to over 83% in household final consumption as a proportion of GDP from 1990 to
2012. Regarding INVMT, this transition along with subsequent monetary policies helped
the country become more attractive to foreign investors in addition to domestic investors
(Ioneci and Mîndreci, 2010). The decrease in NETEX into the negatives is demonstrating
that imports have outweighed exports, indicating Romania has been dependent on imports
as it transitioned after 1989 (Zaman et al., 2016).
Additionally, the end of 2014 shows NETEX slightly increasing closer to zero. If continued
in this direction, exports will soon balance out imports, and eventually exports will
outweigh imports. Zaman et al. (2016) attribute this recent movement to Romania’s
involvement in the European Union upon which has resulted in steadily increasing
Romania’s openness, allowing more exposure to international markets. This openness has
also allowed Romania to become increasingly attractive to foreign investors, hence the
recent increase in NETEX (Carp, 2014) and the progression to more exporting and less
importing. Geamănu (2014) additionally attributes the recent movement in NETEX to
investments, specifically foreign direct investments, upon which improvements in
production has occurred. The drastic decrease (increase) in Romania GDP, FC, and
INVMT (NETEX) is attributed to the 2008 global financial crisis as European countries
were negatively affected in addition to the US (Belaşcu and Budac, 2016).
Figure 1. Romania GDP, FC, INVMT, and NETEX yearly development from 1990-2014
Romania GDP FC
180 160
160 140
140 120
120
100
100
80
80
60
60
40 40
20 20
0 0
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
INVMT NETEX
60 0
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
50 ‐10000
40 ‐20000
‐30000
30
‐40000
20
‐50000
10
‐60000
0 ‐70000
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
‐80000
Results
Equation (3) was estimated in STATA by use of ordinary least squares method to produce
the following results in Table 1. FC has a positive and statistically significant effect on
Romania GDP, indicating an increase of one million RON would increase Romania GDP
by about 963,000 RON. INVMT also has a positive and statistically significant effect on
Romania GDP. Its coefficient estimate indicates a one million RON increase would lead to
an increase in Romania GDP by about 1,227,000 RON. The NETEX coefficient is also
positive and statistically significant, indicating Romania GDP would increase by about 357
RON for every one million RON increase in NETEX. These results would signify there is
a direct and positive relationship between Romania GDP, final consumption, gross
investment, and net exports. The R-squared reported indicates 99.3% of the variation in
Romania GDP is explained by the three explanatory variables in the model.
Although the coefficient estimate for NETEX is significantly smaller than FC and INVMT,
it is still considered relevant. The Romania economy has been relying more on imports
rather than exports during this time period, due in part to the low value placed on products
manufactured within the country as well as transitioning away from communism (Doroftei
and Păun, 2013; Zaman et al., 2016). Thus, the NETEX graph in Figure 1 displays data
taking on a negative value throughout the time period. Additionally, the level of NETEX
is in the billions whereas Romania GDP, FC, and INVMT are within the millions. This
further indicates the high degree of reliance on imports during this time period, the inability
to establish a stable dependence on exports, and the lesser value of the coefficient estimate
compared to FC and INVMT. Overall, FC and INVMT are shown to be the dominant
factors behind Romania’s economy expanding after the fall of communism.
Table 1. Extended model
VARIABLES Romania GDP
FC 0.963***
(0.109)
INVMT 1.227***
(0.285)
NETEX 0.000357***
(6.70e-05)
Constant -1.760
(3.607)
Observations 25
R-squared 0.993
Standard errors in parentheses
*** p < 0.01, ** p < 0.05, * p < 0.1.
Comparison of two models

The secondary purpose of this research paper is to compare this paper’s extended model
against the model Anghelache et al. (2015) constructed in order to establish which of the
two is preferred when analyzing Romania GDP. Anghelache et al. (2015) included FC and
INVMT as their explanatory variables whereas this paper’s model used those same
variables with the extension of NETEX. From this point forward, Anghelache et al. (2015)
model will be referred to as ‘Model 1’, whereas this paper’s model will be represented as
‘Model 2’. Table 2 displays the results of each model’s regression estimates. The results
for Model 1 were obtained directly from within their 2015 paper.
The positive effect FC and INVMT has on Romanian GDP remained the same in both
models. However, the magnitude of these two explanatory variables were affected in
addition to the level of significance for INVMT with the addition of NETEX. The
coefficient estimate for FC decreased slightly (1.163 to 0.963) whereas INVMT’s
coefficient estimate increased considerably (0.325 to 1.23). Furthermore, the addition of
NETEX improved INVMT level of significance (p-value) from 0.3557 to 0.000 thereby
now becoming individually statistically significant. Considering these changes that
occurred with the inclusion of NETEX, it is assumed that Model 1 suffered from omitted
variable bias. Most notably the coefficient estimates for FC and INVMT were
overestimated and underestimated, respectively, in addition to the drastic improvement in
the statistical significance of INVMT from individually statistically insignificant to
significant. The relationship between the explanatory variables and Romanian GDP was
strengthened as evident by increases in both the R2 and adjusted-R2.
Based on Table 2, there is evidence Model 2 is preferred over Model 1 when analyzing
Romania GDP. However, to provide further support for this assumption, the Bayesian
Information Criterion (BIC) was calculated to assist in the comparison of the two models.
Table 2. Comparison of models
Model 1 Model 2
Constant
Coefficient -2.143838 -1.75998
Standard Error 5.40557 3.606827
t-Statistic -0.396598 -0.49
P-value 0.6955 0.631
FC
Coefficient 1.163113 0.9625384
Standard Error 0.152743 0.1086176
t-Statistic 7.614816 8.86
P-value 0.000 0.000
INVMT
Coefficient 0.324933 1.226545
Standard Error 0.344446 0.2852881
t-Statistic 0.94335 4.3
P-value 0.3557 0.000
NETEX
Coefficient 0.0003571
Standard Error 0.000067
t-Statistic 5.33
P-value 0.000
R2 0.98308 0.9928
Adjusted R2 0.981542 0.9918
The BIC is a statistic commonly utilized to identify which model is a better fit for its
specific data set. A key component of the BIC is the inclusion of a “penalty” which accounts
for irrelevant, random data that has the possibility of producing misleading results. The
preferred model will be the one with the smallest BIC. The BIC results are presented in the
table below.
Table 3. BIC results
Model 1 158.8614
Model 2 140.6773
As we can see from Table 3, Model 2 better fits the current data set. These results coincide
with what was determined from Table 2, indicating there is evidence Model 2 should be
the preferred model when analyzing Romania GDP.
Conclusion
This paper expanded upon previous studies regarding the analysis of Romania GDP,
specifically after its shift from communism towards a market-economy from 1990 – 2014.
Final consumption and gross investment were discovered to be the main factors driving the
expansion of Romania’s economy. Net exports, on the other hand, was observed to have a
minimal effect during this time period. However, Romania has been heavily dependent on
imports and less on exports during this period. This strategy is the rationale behind net
export’s minimal effect on Romania’s economy.
In addition to analyzing final consumption, gross investment, and net export’s influence on
Romania’s economy, this paper’s model was compared against the Anghelache et al. (2015)
model to determine which is better suited in analyzing Romania GDP. The results suggest
the model we created is preferred and should be the model utilized. It was also
demonstrated the model Anghelache et al. (2015) created suffered from omitted variable
bias. Future analysis should both expand the time period and include another relevant
variable such as government spending to further investigate Romania GDP post-
communism.
References
Anghelache, C., 2011. Analysis of the Correlation between GDP and the Final Consumption.
Theoretical and Applied Economics, 9(562), pp. 129-138.
Anghelache, C., Manole, A. and Anghel, M.G., 2015. Analysis of final consumption and gross
investment influence on GDP – multiple linear regression model. Theoretical and Applied
Economics, 3(604), pp. 137-142.
Bălă, R.M., 2017. Structure analysis of the evolution of private consumption in Romania. SEA:
Practical Application of Science, II: 2(4), pp. 181-188.
Belaşcu, L. and Budac, C., 2016. Considerations on the Impact of the Global Financial Crisis on
Economies from Eastern Europe. Ovidius University Annals: Economic Sciences Series, 16(2),
pp. 415-420.
Carp, C.L., 2014. The empirical analysis of the relation between FDI, exports and economic growth
for Romania. CES Working Papers, 6(1), pp. 32-41.
Constantin, C., 2017. Using the Regression Model in multivariate data analysis. Bulletin of the
Transilvania University of Braşov, 10(59), pp. 27-34.
Doroftei, I.M. and Păun, C., 2013. Monetary Policy and Central Bank Independence in a Former
Communist Country; Curious Evolutions in Romania. Journal of Eastern Europe Research in
Business and Economics, 2013(2013), pp. 1-14.
Geamănu, M., 2014. VAR analysis on Foreign Direct Investment in Romania. Theoretical and
Applied Economics, 21(4), pp. 39-52.
Ioneci, M. and Mîndreci, G., 2010. The Impact of the Foreign Direct Investment on Romania’s
Economy. Annals of the University of Petroşani, Economics, 10(2), pp. 199-206.
Kartikasari, D., 2017. The Effect of Export, Import and Investment to Economic Growth of Riau
Islands Indonesia. International Journal of Economics and Financial Issues, 7(4), pp. 663-667.
Kira, A.R., 2013. The Factors Affecting Gross Domestic Product (GDP) in Developing Countries:
The Case of Tanzania. European Journal of Business and Management, 5(4), pp. 148-158.
Savu, M. and Voicu, O.-L., 2017. The evolution of the final consumption in Romanian economy.
Analele Universităţii Constantin Brâncuşi din Târgu Jiu: Seria Economie, 1(1), pp. 206-209.
Wolla, S.A., 2018. How Do Imports Affect GDP? Page One Economics, 1-6.
<https://research.stlouisfed.org>
Zaman, G., Gheorghe, F.-V. and Simion, A.-E., 2016. Some characteristics of Romania’s external
trade in the period 1990-2014. Annals of the University of Oradea, Economic Science Series,
25(1), pp. 207-229.
Analysis of the natural movement of the population

under the spectrum of the health crisis
Constantin ANGHELACHE
actincon@yahoo.com
Mădălina-Gabriela ANGHEL
madalinagabriela_anghel@yahoo.com
Ștefan Virgil IACOB
stefaniacob79@yahoo.com
Abstract. The natural movement of the population refers in principle to births, deaths,
divorces and marriages. The reduction in the number of births or the maintenance at a lower level
was carried out under the rule of the pandemic crisis, when the couples were reserved in trying to
have an increased number of offspring.
In the same vein, the reduction in the number of births is due to the fact that the birth rate is
influenced by human fertility, a context in which it has been declining and probably under the effect
of the health and economic and financial crisis, which has been associated and continues at a fast
pace, will impose such a situation.
Mortality was also maintained in reasonable terms despite the fact that there were a number of
effects of the pandemic crisis, the total number decreased. A structure of deaths by age groups shows
that the most affected were those in the groups over 60-70 and even 80 years.
Keywords: birth rate, mortality, divorce, marriages, crises.
JEL Classification: J12, J13, J17.

178 Constantin Anghelache, Mădălina-Gabriela Anghel, Ștefan Virgil Iacob
Introduction
The pandemic crisis has an effect on the increase of deaths in case of comorbidity, with an
immediate effect on the increase of the natural increase of the population. The natural
increase of the population is less and less, because the older age groups reach critical
situations, while the samples, quotas, of young population is smaller compared to the period
of 10-20 years ago. It is found that in rural areas there are some situations in which the
number of deaths is much higher.
Regarding marriages, we will find that their number has decreased, of course due to some
measures imposed in the case of this strategy to fight infection, but also due to the
impossibility of future couples to legalize through marriage in the context of one of the
measures to combat infections it refers to the avoidance in any form of meetings,
assemblies or private manifestations over a particularly small number.
Divorce in turn consisted in some reduction also due to these conditions imposed by some
changes that were imposed in close correlation with the desire to limit their number.
The study is accompanied by some graphs and series of data, tables, which are suggestive
and ensure an easier understanding and interpretation of the data presented by the authors,
data they took from the National Institute of Statistics/Eurostat.
Literature review
The problem of a country's population is important from several points of view. First of all,
the population is the perspective of the evolution of a nation in terms of the source of
evolution of the source of the employed, active population, number of employees and even
unemployment. Many economists and researchers have paid attention to this aspect. A
number of analyses have emerged worldwide, some leading to the hypothesis that the
world's population is oversized, others that Europe's population is declining sharply or that
female fertility needs to be reconsidered. In this sense, we mention works such as
Anghelache et al. (2020) on the quarterly analysis of population evolution or Anghelache
and Anghel (2017) and Anghelache et al. (2018) who published studies on the analysis of
population evolution in Romania and more globally, with emphasis on the fact that
population evolution is demographic factor of population growth, factor of evolution of the
labour force or element of appreciation of the perspective of the evolution of humanity
through the prism of the relation between the global resources and the evolution of the
world population. In other words, a number of papers such as those of Anghelache (1996),
Anghelache et al. (2006), Baron et al. (1996), Ber Py (1989) analyse in their works various
aspects related to the methods of measuring and comparing macroeconomic indicators that
give essence to economic development, and Capanu and Anghelache (2001a, 2001b)
address in detail in their work macroeconomic indicators. Also, Pecican and Tănăsoiu
(1989) and Tövissi, et al. (1979) are concerned with some methods and models of analysis
of economic phenomena. Bijak et al. (2007) conducted a study on the population and
employment outlook in the 27 Member States of the European Union, with a view to 2052.
The expertise is somewhat bleak if we consider the resources, development programs and
Analysis of the natural movement of the population under the spectrum of the health crisis 179
the impact of international migration. Headey and Hodge (2009) analysed the effect of
population growth on increasing performance indicators at the national level. Melo et al.
(2008) dealt with the study of economic developments at the regional level, and Oster et
al. (2O13) addressed the issue of capital in general and the prospect of increasing
investment. Last but not least, Walker and Maltby (2012) dealt with the study of
demographic solutions on the achievement of a settlement strategy at European level.
Methodology
In order to understand the content of the demographic indicators used, we will briefly
see the main provisions of the methodology used by the National Institute of
Statistics/Eurostat in the calculations of the disseminated statistical variables. Thus, the
data on demographic phenomena were obtained by processing the information contained
in the statistical bulletins of live births, deaths, marriages and divorces prepared by
municipal, city and communal town halls, together with the registration of the
phenomena in civil status documents. The main concepts used have the following
content.
Born alive is the product of conception, expelled or completely extracted from the
mother's body, regardless of the duration of pregnancy and which, after this separation,
shows a sign of life (respiration, heart activity, umbilical cord pulsations or will-
dependent muscle contractions). The deceased is the person whose vital functions have
ceased definitively after some time has elapsed since birth.
Regarding the natural increase, this represents the difference between the number of live
births and the number of deceased persons, in the analysed period. Also, marriage is the
union between a man and a woman, concluded in accordance with the legislation of the
country, in order to establish a family and which results in rights and obligations
between the two spouses, as well as their children.
Divorce consists in the dissolution of a legally concluded marriage, by a final decision
of the court, of the civil status officer or of a notary public. The data refer to divorce
proceedings for which the dissolution of the marriage was allowed.
By the domicile of a person we mean the address at which he declares that he has the
main residence, entered in the identity document, as it is taken into account by the
administrative bodies of the state.
As for the habitual residence, it is the place where a person normally spends his daily
rest period, without taking into account temporary absences for recreation, holidays,
visits to friends and relatives, business, medical treatments or religious pilgrimage. The
habitual residence may be the same as the domicile or it may differ in the case of persons
who choose to establish their habitual residence in a locality other than that of domicile
in the country or abroad.
The number of live births includes live births whose mothers had, at their date of birth,
their habitual residence or residence for a period of at least 12 months in Romania, and
the number of deceased includes persons who had, at the date of death, their habitual
residence or residence for a period of at least 12 months in Romania.
Marriages include the legal act of marriage of persons who had, at the date of marriage,
domicile in Romania, as well as marriages of Romanian citizens who marry abroad and
are registered at the civil registry offices in Romania, while divorces include the
dissolution of marriages of persons whose divorces were concluded with judges, marital
status or notaries public, in accordance with the legal framework, as well as the divorces
of Romanian citizens who divorced abroad, transcribed in Romania.
The birth and mortality data for 2018 are final and are distributed after the date of the
demographic event. Also, the data on birth and mortality for 2019 are semi-final and are
distributed after the date of the demographic event. In the same vein, the data on
marriage and divorce for 2019 are final and are distributed after the date of the
demographic event. Also, the data regarding the demographic phenomena related to the
months of 2020 are provisional and are distributed after the date of registration of the
demographic event.
The provisional data on the demographic phenomena related to the months of a certain
year present the demographic phenomena, registered at the civil status offices in the
considered year, distributed after the date of registration of the phenomenon. Provisional
data for the year in question do not include phenomena produced in the previous year
and recorded late in the year in question (except January of each year).
By semi-definitive data regarding demographic phenomena such as births and deaths,
registered at the civil status offices, related to the months of the considered year, they
represent the absolute provisional data related to the demographic phenomena registered
at the civil status offices in each month of the considered year. redistributes by months
after the date of production, from which are excluded the phenomena registered late at
the civil status offices, in the considered year, but which took place in the year before it
and the demographic phenomena registered at the civil status offices, registered late in
Romania are added during the first month of the following year, but which occurred
during the year under review.
The final data on the demographic phenomena registered at the civil status offices,
related to the months of the considered year represent the absolute semi-final data related
to the demographic phenomena registered at the civil status offices of the considered
year, redistributed by months after the production date in the first nine months of the
following year, but which occurred in the year in question.
The revised data on demographic phenomena (births and deaths) related to the months
of the year considered represent the absolute final data related to demographic
phenomena (births and deaths) recorded in the year considered, to which were added the
demographic phenomena (births and deaths) recorded late in subsequent years the first
or second and the first 4 months of the third year, but which took place in the year in
question.
The final data with reference to the demographic phenomena (marriages and divorces)
produced in the considered year represent the total of the demographic phenomena
(marriages and divorces) registered in the reference year and distributed on months after
the date of the demographic events.
In the study, the authors used statistical methodologies, consisting of the system of
statistical indices and indicators, structural analyses, graphical representations,
statistical data series, etc.
Data, results and discussions

Based on the data presented by the National Institute of Statistics, we find that in September
2020 the number of registered births increased compared to September of the previous year
and compared to the previous month of the same year. Also, in September 2020 the number
of deaths and the number of deaths of children less than one year of age increased compared
to the same period of the previous year. The data are summarized in Tables 1 and 2.
Table 1. Natural population movement in Romania in 2019 (persons)
Year/ Increase Deceased
Live births1) Deceased1) Marriages2) Divorces2)
Month natural1) under 1 year1)
2019
January 17614 27388 -9774 5551 2196 99
February 14993 22357 -7364 5306 2672 90
March 15716 23665 -7949 6268 2884 105
April 15588 21338 -5750 5470 2686 103
May 16434 21264 -4830 13260 2788 89
June 16049 20529 -4480 14091 2657 106
July 18082 20305 -2223 16793 2153 114
August 16625 19899 -3274 23044 1579 101
September 16287 18982 -2695 15570 2466 91
October 14884 21079 -6195 11350 2909 82
November 13108 20482 -7374 6555 2827 79
December 12755 22433 -9678 5352 2380 86
1) Semi-finaldata distributed after the date of the demographic event.
2)Final data distributed after the date of the demographic event.
Source: INS communique number 291/10 November 2020.
Table 2. Natural population movement in Romania in 2020 (persons)
Year/ Increase Deceased
Live births1) Deceased1) Marriages1) Divorces1)
Month natural1) under 1 year1)
2020
January 15971 23352 -7381 4356 836 91
February 12127 21699 -9572 5864 2184 102
March 11857 22591 -10734 4053 1889 83
April 12098 21812 -9714 1409 1381 94
May 12873 20249 -7376 2488 1137 77
June 14786 21073 -6287 5389 2355 83
July 17415 22417 -5002 11168 1863 101
August 16804 23169 -6365 14929 1953 88
September 18036 21973 -3937 12471 2371 104
1) Provisional
data distributed after the date of the demographic event.
In the same vein, in September 2020 the number of registered marriages decreased
compared to September 2019 and compared to August 2020. Also, in September 2020,
18,036 children were born, with 1,232 more children than in August. 2020.
Regarding the number of deaths registered in September 2020, it was 21,973, with 1,196
fewer deaths than in August 2020, and the number of deaths of children less than one year,
registered in September 2020, was 104 children, increasing by 16 deaths compared to
August 2020.
The comparison with the previous month is affected by the fact that the statistical records
of demographic phenomena register a gap from one month to another, due to the fact that
there are demographic phenomena produced in the last 4-5 days of the current month. Also,
the mortality phenomenon registers a seasonal evolution, being more accentuated in the
winter months and registering the lowest values in July, August and September. The data
are structured in Table 3.
Table 3. Deaths recorded in the reference month, after the date of production (persons)
Deaths caused 2)
Months of the Recorded
in the reference in the month preceding two months or more before
year 2020 deaths 1)
month the reference month the reference month
January 23352 22434 757 161
February 21699 20607 1092 -
March 22591 21335 1153 103
April 21812 20593 1049 170
May 20249 19317 796 136
June 21073 19542 1292 239
July 22417 20980 1133 304
August 23169 21765 1108 296
September 21973 20371 1196 406
1)Provisional data distributed after the date of registration of the demographic phenomenon.
2)Provisional data distributed according to the date of occurrence of the demographic phenomenon.
Data on the evolution of the number of live births and deaths in July-September 2011-2020
are summarized in Table 4.
Table 4. Evolution of the number of live births and deaths in July-September 2011-2020
Year
2011 2012 20131) 20141) 20151) 20161) 20171) 2018 20192) 20203)
Month
Live births
July 18794 18860 19892 19323 19906 18905 20011 19416 18082 17415
August 17698 18147 19609 17477 17833 18916 19620 18028 16625 16804
September 18085 18067 20836 18824 19053 19963 19956 18766 16287 18036
Deceased
July 19353 20996 19146 19582 20593 19567 19863 20298 20305 22417
August 18660 18960 18450 19369 19411 19391 20508 20463 19899 23169
September 17567 17688 18894 18756 18696 18955 18734 19347 18982 21973
1) Data revised according to the INS Revision Calendar.
2) Semi-final data distributed after the date of the demographic event.
3) Provisional data distributed after the date of the demographic event.

Figure 1 shows the evolution of the number of live births and deaths, between January 2018
and September 2020, noting a decrease in the number of deaths in September 2020
compared to the previous one.
Figure 1. Evolution of the number of live-born and deceased, in the period January 2018 - September 2020
Figure 2 shows the deaths recorded in September 2020, by age groups.

Figure 2. Deaths recorded in September 2020, by age groups
According to the data presented, we find that in September 2020, almost two thirds of the
total number of deaths were registered for people aged at least 70 years, respectively 9,039
deaths, representing 41.1% of the total, were registered at age groups 80 years and over,
5,230 deaths, representing 23.8%, in people aged 70-79 years and 4,354 deaths, i.e. 19.8%,
in people aged 60-69 years. The lowest deaths were recorded in the age groups 5-19 years
(64 deaths), 20-29 years (104 deaths) and 0-4 years (131 deaths).
Regarding the natural increase, it remained negative in September 2020, with a surplus of live
births of 3,937 people. In the same vein, in September 2020, 12,471 marriages were registered
at the civil status offices, 2,458 less than in August 2020, and the number of divorces
pronounced by final court decisions was 2,371 in September 2020, 418 more than in August
2020. The data regarding the evolution of the number of marriages and divorces, in the period
September 2019 - September 2020 are presented graphically in Figure 3.
Figure 3. Evolution of the number of marriages and divorces, between September 2019 and September 2020
Regarding the number of live births registered in September 2020, it was higher by 1,749
compared to the same month in 2019. Also, the number of people who died in September
2020 was 2,991 higher than in September 2019, and as a consequence, the negative natural
increase increased in September 2020 compared to September 2019 (-3937 people in
September 2020, compared to -2695 people in September 2019). Also, the number of
children under one year of age who died was 13 times higher in September 2020, than the
one registered in September 2019. The evolution of the number of deaths by residence,
during January-September of 2019 and 2020 is presented in Figure 4.
Figure 4. Evolution of the number of deaths by area of residence, in the period January - September of 2019
and 2020
According to the data presented and analysed, we find that in September 2020, 11,146
people died in urban areas, of which 5,974 men and 5,172 women, and in rural areas the
death of 10,827 people, of which 5,885 men and 4,942 women. Compared to the similar
period in 2019, the number of people who died increased by 1,872 people, of which 1,092
men and 780 women in urban and rural areas by 1,119 people, of which 610 men and 509
women.
In the same vein, the number of marriages was, in September 2020, 3,099 lower than that
recorded in the same month of the previous year, and by final court decisions were
pronounced with 95 divorces less than in September 2019.
The effects of the health crisis on demographic phenomena are visible in the case of deaths,
especially in April, June, July, August and September 2020, when there were increases
compared to April, June, July, August and September 2019, of marriages which, since
March, have seen a decrease in the number of events compared to the corresponding month
in 2019. Also, the number of births has decreased every month compared to the same period
of the previous year, except for August and September when a number was recorded, higher
live births than the previous year.
Conclusions
From this article a series of theoretical and practical conclusions can be drawn, which must
be taken into account in the analysis of the trend that the natural movement of the
population will have in the next period.
First of all, the health crisis will temper the phenomenon of marriage due to the concrete
conditions. Of course, reducing marriage due to the health crisis or low fertility will reduce
the quotas of the young population. Romania has long been experiencing declines in the
natural increase of the population due to declining birth rates, along with higher mortality,
taking into account the older contingents, which are in the critical phase of leaving the stage
of life.
It also turns out that this health crisis will be able to increase the number of deaths, which
offers a negative perspective, that of the increase in stillbirth.
The natural movement of the population will be lower as a result of the declining incomes,
which do not encourage generations (young couples) to start a family, which is the most
appropriate framework for raising the birth rate.
In contradictory terms to some opinions that are manifested nationally in relation to the
world's population, the need to stabilize stillbirth in Romania becomes a problem, a context
in which couples should be encouraged to give birth to offspring and find additional ways
of health care for older generations, in order to extend the average life expectancy.
References
Anghelache, C., 1996. Măsurarea şi compararea dezvoltării economice, Economica, Bucharest.

Anghelache, C. (coord.), 2006. Analiză macroeconomică, Economica, Bucharest.
Anghelache, C. and Anghel, M.G., 2017. Analysis of population development – labour resources of
member states of the European Union. Management & Gouvernance, 17, January-June,
pp. 95-110.
Anghelache C., Avram (Burea), D. and Petre (Olteanu), A., 2018. Analysis of the Natural Movement
of Population and Labor Force Development. Romanian Statistical Review, Supplement, 2,
pp. 115-123.
Anghelache, C., Iacob, S.V. and Grigorescu, D.L., 2020. The analysis of the quarterly evolution of
the gross domestic product in 2019. Theoretical and Applied Economics, XXVII (1),
pp. 171-182.
Baron, T., Biji, E.M. and Tövissi, L., 1996. Statistică teoretică şi economică, Ed. Didactică şi
Pedagogică RA, Bucharest.
Ber Py, B., 1989. Statistique descriptive, Economica, Paris.
Bijak, J., Kupiszewska, D., Kupiszewski, M., Saczuk, K. and Kicinger, A., 2007. Population and
labour force projections for 27 European countries, 2002-052: impact of international
migration on population ageing. European Journal of Population, (23) 1, March 2007,
pp. 1-31.
Capanu, I. and Anghelache, C., 2001a. Indicatori economici – conţinut, calcul analiză, Economica,
Bucharest
Capanu, I. and Anghelache, C., 2001b. Indicatori macroeconomici – conţinut, metodologie de calcul
şi analiză economică, Economica, Bucharest.
Headey, D. and Hodge, A., 2009. The Effect of Population Growth on Economic Growth: A Meta‐
Regression Analysis of the Macroeconomic Literature. Population and Development Review,
35 (2), pp. 221-248.
Melo, P.C., Graham, D.J. and Noland, R.B., 2008. A meta-analysis of estimates of urban
agglomeration economies. Regional Science and Urban Economics, 39 (3), pp. 332-342.
Oster, E., Shoulson, I. and Dorsey, E., 2013. Limited Life Expectancy, Human Capital and Health
Investments. American Economic Review, 103 (5), pp. 1977-2002.
Pecican, E. and Tănăsoiu, O., 1989. Econometrie, ASE Bucharest.
Tövissi, L., Scarlat, E. and Taşnadi, Al., 1979. Metode şi modele ale analizei economice structurale,
Ed. Ştiinţifică şi Enciclopedică, Bucharest.
Walker, A. and Maltby, T., 2012. Active ageing: A strategic polity solution to demographic ageing
in European Union. International Journal of Social Welfare, 21 (s1), s117-s130.
On the relationship between economic growth and government

debt for Bulgaria. Test of the Reinhart-Rogoff hypothesis
Yu HSING
Southeastern Louisiana University, USA
yhsing@selu.edu
Abstract. Based on an extended production function, this paper finds that the turning point of the
government debt ratio for Bulgaria is estimated to be 45.2631%, suggesting that an increase in the
debt ratio beyond 45.2631% will cause the growth rate of real GDP to decline. This turning point
for Bulgaria is far less than the 90% turning point proposed by Reinhart and Rogoff. Therefore, the
Reinhart-Rogoff hypothesis does not apply to Bulgaria.
Keywords: fiscal policy, government debt ratio, Reinhart-Rogoff hypothesis, debt threshold.
JEL Classification: E62.

188 Yu Hsing
Introduction
Bulgaria has shown a declining trend of the government debt ratio in the long run from a
high of 79.406% in 1999 to a low of 14.086% in 2010. Even during the global financial
crisis, its debt ratios remained relatively low at 14.72% in 2008 and 14.579% in 2009.
Although its debt ratio rose to 27.357% in 2016, the Bulgarian government pursued fiscal
prudence to lower it to 19.156% in 2019. Relatively low debt ratios in recent years were
attributable to relatively low budget deficits as evidenced by less than 3% of government
borrowing or structural balance as percent of GDP since 2015.
Whether a higher government debt ratio would increase or reduce the growth rate of real
GDP has been studied extensively. The focus point is whether there would be a turning
point beyond which a further increase in the debt ratio would reduce the growth rate of real
GDP. Reinhart and Rogoff (2010a, 2010b) show that the turning point of the government
debt ratio is 90%. When the government debt ratio is below 90%, the negative impact of a
higher debt ratio on the growth rate of real GDP is weak. When the government debt ratio
is greater than 90%, a further increase in the debt ratio tends to reduce the growth rate of
real GDP.
The purpose of this paper is to test whether the Reinhart-Rogoff hypothesis may be
applicable to Bulgaria. This paper differs from most previous studies in several aspects.
First, a theoretical model based on an extended production function is presented. Second,
a quadratic function is employed to determine whether there may be a turning point. Third,
the GARCH process is used in empirical work in order to correct for potential autogressive
conditional heteroskedasticity.
Literature survey
Based on a sample of forty-four advanced and developing countries covering about 200
years, Reinhart and Rogoff (2010a, 2010b) find that there is a weak relationship between
the growth rate and the debt ratio when the debt ratio is less than 90% whereas a debt ratio
greater than 90% causes the growth rate to decline. This threshold for the debt ratio is
comparable in advanced and emerging economies. However, based on advanced
econometric techniques, Minea and Parent (2012) find that the threshold for the debt ratio
is 115%. Herndon, Ash and Pollin (2014) use the same data compiled by Reinhart and
Rogoff and find that for 20 advanced countries, the negative effect of the public debt ratio
above 90% on economic growth cannot be confirmed. According to Herndon, Ash and
Pollin, during 1946-2009, countries with public debt ratios over 90% recorded an average
growth rate of 2.2% instead of -0.1% as presented by Reinhart and Rogoff. The relationship
between the economic growth rate and the public debt ratio differs substantially by country
and period.
Using a sample of 18 OECD countries during 1980-2008, Cecchetti, Mohanty and
Zampolli (2011) show that the threshold of the government debt ratio is about 85%,
implying that a rising government debt ratio beyond 85% would have an adverse effect on
economic growth. They also indicate that if the corporate debt ratio is greater than 90% of
GDP, there will be a negative impact on economic growth.
On the relationship between economic growth and government debt for Bulgaria 189
Checherita-Westphal and Rother (2012) use a sample of 12 countries in the euro area to
examine the relationship between growth and government debt. They find a threshold or
turning point in the range of 90-100%, suggesting that more government debt reduces
economic growth if the debt ratio is greater than 90-100%. Based on the confidence
intervals, the threshold would start in the range of 70-80%.
Baum, Checherita-Westphal, and Rother (2013) study the relationship between growth and
government debt for twelve countries in the euro area. A higher debt ratio increases the
growth rate but has no effect on the growth rate when the debt ratio reaches about 67%.
When the debt ratio is greater than 95%, a higher debt ratio causes the growth rate to
decline.
Afonso and Jalles (2013) investigate the relationship between economic growth and
government debt for 155 advanced and developing countries during 1970-2008. According
to their results, the turning point is 59% for the euro zone and 79% for emerging economies.
If the debt ratio rises 10%, the growth rate would decline by 0.2% if the debt ratio is greater
than 90% of GDP and rise by 0.1% if the debt ratio is less than 30% of GDP.
Chirwa (2017) examines the relationship between growth, government debt and other
relevant variables for ten countries in the euro area. The threshold is found to be at the 70%
in the long run whereas government debt and growth have a negative relationship in the
short run.
Based on a sample 154 countries, Swamy (2015) finds that a 10-percentage point increase
in the government debt ratio leads to 2 to 23 basis point decrease in the average growth rate
and that growth and government debt have a nonlinear relationship.
Woo and Kumar (2015) reveal that if the initial debt ratio increases 10 percentage points,
the growth rate of real per capita GDP will decline about 0.2 percentage points. Higher
government debt ratios lead to larger negative effects. The negative effect is mainly due to
decline in labor productivity growth.
Velichkov (2016) evaluates the effect of government debt on economic growth for
Bulgaria. He reveals that more government debt promotes economic growth in the short
run but has a negative impact on economic growth in the long run. He does not present the
threshold for the debt ratio beyond which more debt would affect economic growth
negatively.
Lechtenberg (2017) studies the subject based on a sample for 10 individual countries.
Australia, Canada, Chile, Germany and New Zealand have had low and declining debt
ratios, and a higher debt ratio would not cause the growth rate of real GDP to decline. On
the other hand, debt thresholds are found for France, Greece, Italy, the UK and the US.
Beyond the debt thresholds, a higher debt ratio reduces the growth rate in Greece, Italy, the
UK and the US but increases the growth rate for France.
Shahor (2018) studies the relationship between growth and government debt for Israel
during 1983-2013. The relationship exhibits an inverted U-shape. The threshold or the
turning point of the debt ratio is 130% and greater.
190 Yu Hsing
Jacobs, Ogawa, Sterken and Tokutsu (2020) examine the relationship between economic
growth and public debt for 27 EU members and 4 OECD countries during 1995-2013. They
find that more public debt does not Granger cause economic growth. Instead, economic
growth Granger causes public debt. Slow economic growth causes more public debt. In
high-debt economies, slow economic growth increases public debt, which causes a higher
long-term interest rate, dampens interest-rate sensitive private spending, and increases
public debt. In addition, they show that the effect of economic growth on the debt ratio is
greater for high-debt economies and that the effect of the debt ratio on economic growth is
greater for low-debt countries.
The model
Extending Ram (1986, 1989), Goel, Payne and Ram (2008) and other studies, the growth
rate of real GDP can be expressed as:
𝑌 𝑤 𝐿, 𝐾 , 𝐷 (1)
where:
𝑌 – the growth rate of real GDP;
𝐿 – the growth rate of labor employment;
𝐾 – the growth rate of capital;
D – the government debt-to-GDP ratio.
Due to lack of the data for capital, the growth rate of capital can be substituted by the ratio
of investment spending (I) to gross domestic product (Ram, 1986, 1989).
𝑌 𝑧 𝐿 , 𝐼 ⁄𝑌 , 𝐷 (2)
In a linear form, the coefficient of 𝐿 measures the elasticity of output with respect to labor,
and the coefficient of 𝐼 ⁄𝑌 represents the partial derivative of Y with respect to K or the
marginal product of capital. The sign of the first two explanatory variables should be
positive, and the sign of the debt ratio is unclear. A lower and rising government debt ratio
for infrastructural improvements may be conducive to economic growth whereas a higher
and rising debt ratio may raise the interest rate, crowd out private spending, cause the
Bulgarian lev to appreciate, and hurt exports.
There may be a turning point or an inverted U-shaped relationship between 𝑌 and the
government debt ratio. That being the case, the following equation can be considered:
𝑌 𝑓 𝐿 , 𝐼 ⁄𝑌 , 𝐷, 𝐷 (3)
An inverted U-shaped relationship between 𝑌 and the debt ratio suggests that the sign of D
should be positive and the sign of 𝐷 should be negative.
The critical value (turning point) of the debt ratio corresponding to the maximum growth
rate of real GDP is given by:
𝐷∗ 𝛼 ⁄2𝛼 (4)
where: 𝛼 is the coefficient of D and 𝛼 is the coefficient of 𝐷 .
Empirical results
The data were collected from the World Economic Outlook and International Financial
Statistics published by the International Monetary Fund. The growth rate of real GDP is
expressed as a percent. The growth rate of labor employment is expressed as a percent.
Investment spending as a percent of GDP is used as the data for capital is not available.
Government debt is measured as a percent of gross domestic product. The sample ranges
from 1998 to 2019. The data for the government debt ratio before 1998 is not available.
Figure 1 exhibits growth rates of real GDP during the sample period of 1998-2019. Growth
rates were mostly positive except for -0.494% in 1999 and -3.586% in 2009 during the
global financial crisis. Figure 2 shows the scatter diagram between the growth rate of real
GDP and the government debt ratio. It seems that they may have a positive relationship
when the debt ratio is relatively low and a negative relationship when the debt ratio is
relatively high.
Figure 1. The growth rate of real GDP
GROWTH_RATE
8
‐2
‐4
98 00 02 04 06 08 10 12 14 16 18
Figure 2. Scatter diagram between the growth rate and the debt ratio
8
4
GROWTH_RATE
‐2
‐4
10 20 30 40 50 60 70 80
DEBT_RATIO
Table 1 presents the estimated regression and related statistics. As shown, the value of R
squared is estimated to be 34.71%, and all the coefficients have the expected signs and are
significant at the 1% level. Based on equation (4), the turning point or the critical value is
calculated to be 45.2631%, which is far below the turning point of 90% proposed by
Reinhart and Rogoff (2010). There may be several reasons why the turning point for
192 Yu Hsing
Bulgaria is much lower than 90%. First, the European Union recommends its member
countries to pursue a government debt ratio no greater than 60%. As a member of the
European Union, it is natural for Bulgaria to aim at a debt ratio below 60%. Second, the
Bulgarian government has kept the government debt ratio at a relatively low level in order
to reduce potential financial or political risk. Countries with huge sovereign debt might
face potential default and high political and financial risk. Third, a rising government debt
is expected to raise the interest rate and reduce consumption spending, investment
spending, and net exports.
Table 1. Estimated regression
Variable Coefficient Probability
Constant -4.253636 0.0000
Growth rate of employment 0.029125 0.0000
Investment/GDP ratio 0.025765 0.0000
Debt ratio 0.403475 0.0000
Debt ratio squared -0.004457 0.0000
R squared 0.347391
Akaike info criterion 4.443437
Schwarz criterion 4.790587
Sample period 1998-2019
If the debt ratio squared is not included in the estimated equation, all the coefficients have
the positive sign and are significant at the 1% level, and the value of R squared is 16.58%,
which is much less than 34.71% when the debt ratio squared is included in the estimated
equation in Table 1.
In comparison, the finding in this paper is different from and similar to some of previous
studies. Reinhart and Rogoff (2010a, 2010b) show that the threshold for the government
debt ratio is 90% whereas the threshold of the debt ratio for Bulgaria is estimated to be
45.2631%. Kumar and Woo (2015) and Swamy (2015) indicate that the growth rate and
the debt ratio have a negative relationship whereas this paper finds that the relationship
may be positive or negative depending upon the level of the debt ratio. The thresholds
reported by Cecchetti, Mohanty and Zampolli (2011), Minea and Parent (2012),
Checherita-Westphal and Rother (2012), Baum, Checherita-Westphal, and Rother (2013),
Afonso and Jalles (2013), Chirwa (2017), and Shahor (2018) are higher than the threshold
of 45.2631% estimated for Bulgaria. The estimated threshold of 45.2631% for Bulgaria is
close to the threshold of 47.4452% for Italy reported by Lechtenberg (2017).
Summary and conclusions

This paper has examined the impact of the government debt ratio on the growth rate of real
GDP for Bulgaria using an extended production function. The growth rate of real GDP is
specified as function of the growth rate of employment, the ratio of investment spending to
nominal GDP, and the government debt ratio. A quadratic form is used for the debt ratio to
test whether a turning point may exist.
The results show that the turning point is estimated to be 45.2631%. The growth rate and
the debt ratio have a positive relationship when the debt ratio is up to 45.2631% whereas
they have a negative relationship when the debt ratio is greater than 45.2631%. The
government debt ratio in 2019 was 19.156%, indicating that there would be room for the
Bulgarian government to engage in fiscal expansion to raise the debt ratio slightly to
stimulate its economy due to the worldwide pandemic crisis. The lower turning point for
Bulgaria also indicates that a criterion which is applicable to advanced countries may not
apply to emerging or developing countries. Hence, Bulgaria’s efforts to maintain fiscal
discipline are appropriate.
To stimulate the economy and enhance income convergence, the government may consider
increasing automatic fiscal stabilizers and public investment spending, allowing small
deficits in support of growth-oriented programs as long as the government debt ratio would
be kept at a relatively low level. The Bulgarian government may strengthen active labor
market policies, worker importation agreements, training programs to mitigate labor
shortages, and business involvement to streamline university curricula to address the
demand for skilled workers (International Monetary Fund, 2020).
References
Afonso, A. and Jalles, J.T., 2013. Growth and productivity: The role of government debt,
International Review of Economics and Finance. 25 (Issue C), pp. 384-407.
Baum, A., Checherita-Westphal, C. and Rother, P., 2013. Debt and growth: New evidence for the
euro area. Journal of International Money and Finance, 32, pp. 809-821.
Cecchetti, S.G., Mohanty, M.S. and Zampolli, F., 2011. The real effects of debt. Bank for
International Settlements, Bis Working Papers No. 352.
Checherita-Westphal, C. and Rother, P., 2012. The impact of high government debt on economic
growth and its channels: An empirical investigation for the euro area. European Economic
Review, 56(7), pp. 1392-1405.
Chirwa, T.G., 2017. Public debt and economic growth nexus in the Euro area: A dynamic panel
ARDL approach.
Goel, R.K., Payne, J.E. and Ram, R., 2008. R&D expenditures and US economic growth:
A disaggregated approach. Journal of Policy Modeling, 30(2), pp. 237-250.
Herndon, T., Ash, M. and Pollin, R., 2014. Does high public debt consistently stifle economic
growth? A critique of Reinhart and Rogoff. Cambridge Journal of Economics, 38(2),
pp. 257-279.
International Monetary Fund, 2020. Bulgaria: Staff concluding statement of the 2020 article IV
mission, February 14, 2020.
Jacobs, J., Ogawa, K., Sterken, E. and Tokutsu, I., 2020. Public Debt, Economic Growth and the
Real Interest Rate: A Panel VAR Approach to EU and OECD Countries. Applied Economics,
52(12), pp. 1377-1394.
Lechtenberg, L., 2017. The debt-to-GDP threshold effect on output: A country specific analysis.
Aisthesis, the Interdisciplinary Honors Journal, 8 (1), pp. 26-34.
Minea, A. and Parent, A., 2012. Is high public debt always harmful to economic growth? Reinhart
and Rogoff and some complex nonlinearities.
194 Yu Hsing
Ram, R., 1986. Government size and economic growth: A new framework and some evidence from
cross-section and time-series data. The American Economic Review, 76(1), pp. 191-203.
Ram, R., 1989. Government size and economic growth: A new framework and some evidence from
cross-section and time-series data: Reply. The American Economic Review, 79(1),
pp. 281-284.
Reinhart, C.M. and Rogoff, K.S., 2010a. Growth in a time of debt, American Economic Review,
100(2), pp. 573-78.
Reinhart, C.M. and Rogoff, K.S., 2010b. Debt and growth revisited.
Shahor, T. (2018). The impact of public debt on economic growth in the Israeli economy. Israel
Affairs, 24(2), pp. 254-264.
Swamy, V., 2015. The Dynamics of Government Debt and Economic Growth. Available at SSRN
2595106.
Velichkov, N., 2016. Effects of Government Debt on Macroeconomic Activity (The Case of
Bulgaria). Economic Alternatives, 1, pp. 24-32.
Woo J., and Kumar, M.S., 2015. Public debt and growth. Economica, 82(328), pp. 705-739.
f Theoretical and Applied Economics
Testifying the role of regulatory environment in trade

facilitation: Impact on intra-regional trade in South-Asia
Tanya GANDHI
Jamia Millia Islamia, New Delhi, India
Tanya.gandhi123@gmail.com
Shahid AHMED
Jamia Millia Islamia, New Delhi, India
sahmed@jmi.ac.in
Abstract. The intraregional trade in South-Asia region is only 6.1 percent of its total trade due to
a number of factors, including Regulatory environment governing international trade. This study
examines the impact of trade facilitation in Regulatory environment on bilateral exports within the
region. A panel gravity model augmented with a self-composed index for Regulatory environment
for both exporting and importing countries in South-Asia is estimated. It reveals that enhanced
regulatory environment of importing country impacts intra-regional trade positively whereas, that
of the exporting country impacts the same negatively. The estimates of trade potential within the
region are significantly encouraging. The simulations’ results reveal that by making collective
efforts in trade facilitation in regulatory environment, the intra-regional exports could increase
tremendously in a range of scenarios which could not be achieved even by phasing out of tariffs
within the region.
Keywords: trade facilitation; regulations; South-Asia; augmented panel gravity model; trade potential;
simulations.
JEL Classification: C23, F10, F15, F17, L51.

196 Tanya Gandhi, Shahid Ahmed
1. Introduction
Reducing tariff rates in consecutive WTO rounds, growing worldwide supply chain
management procedures, and continuing development of global trade have resulted in
increased discourse on the issues of “on-the border” and “behind-the border” trade
transaction costs and their repercussions on global trade. Trade transaction costs are
asserted to be an important element in describing the arrangement of foreign investment
and trade flows (Deardorff, 1998, 2014; Obstfeld and Rogoff, 2000). As the pace of
institutional globalization increases, the capacity of developing nations to integrate with
worldwide and regional markets is rapidly impaired by the cost of trade procedures incurred
by the private sector.
Economic conditions such as lopsided trade facilitation framework, trade-related
infrastructure deficiencies or a cumbersome regulatory environment create a negative
externality on business enterprises, increase the cost of trade procedures and deviate
industrial structure with detrimental consequences on trade and economic growth.
Therefore, a country’s institutions and regulatory environment has a lot to do with its
propensity to trade and trade performance.
A country's institutions are described as the provisions that shape its members ' social,
economic and political interplay. North (1990) states that the primary role of institutions is
to decrease ambiguity arising out of asymmetric information and transaction costs. This
promotes market interplay and generally improves market functioning. Whereas,
imperfectly devised institutions can escalate costs crucially and impede economic ventures
and specialization (Borrmann et al. 2006).
The political institutions of a country frame its economic institutions and a significant
component of these institutions is the regulatory environment of that country (Acemoglu
et al. 2001). During the last two decades, researchers, policymakers and international
organizations have significantly emphasized the role of a strong regulatory framework in
the development of a country. The trade related regulatory environment is particularly
important for an open economy that is in constant contact with the outside world and
depends upon it for optimal outcomes.
Trade openness or economic integration is strongly connected to regulatory environment
as it can magnify the repercussions of good or bad regulations. With rising integration of
world economies, factors of production and units of production are becoming increasingly
mobile. Therefore, regulatory environment of a country can either attract firms or lead them
to move to another country where regulations are more favourable so it can either be a
competitive advantage or disadvantage for a country. Hence, the consequences of
regulations on economic growth of a nation depends upon its level of economic integration
(Silberberger & Königer, 2016).
It is therefore a challenge to design regulatory environment of a country that ensures
quality, transparency, stringency and enforcement of policies along with providing
competitive advantage to a country in terms of trade. This is where Trade facilitation comes
into the picture. Trade facilitation is usually expected to decrease the cost of transit of goods
and services across borders and the procedural costs associated with enforcing, regulating
Testifying the role of regulatory environment in trade facilitation 197
and administering of trade guidelines (Staples, 2002). From a narrow perspective, trade
facilitation is defined as improving logistics for efficient movement of goods through ports
and easing the documentation related to international trade. The broader rendition of trade
facilitation spans the atmosphere provided for trade undertakings made up of transparency
of customs and regulatory environments, besides harmonisation of regulations as well as
requirements for trade with international standards (Engman, 2005).
In spite of the undeniable advantages of the trade facilitation reform, it is unlikely that
implementing these measures alone will bring notable improvements to export efficiency
in developing nations. To achieve this, an integrated strategic investment program directed
at strengthening the overall regulatory atmosphere along with tangible infrastructure will
be required, that could address the supply side barriers which prevent the economy from
responding to growing trade incentives. Integrative measures that increase the effectiveness
of the entire export supply chain are essential for a successful reform of trade facilitation.
Therefore, any decrease in trade costs arising out of improvements in trade facilitation
might produce comparatively lower impact on export outcomes if enacted without parallel
reforms aimed at relaxing the constraints on export production capability.
Despite soaring expectations of returns to trade facilitation in regulatory environment, the
empirical evidence of impact of reforms on trade volumes is limited and robust supporting
testimony of trade facilitation measures positively influencing trade volume is proved hard
to provide. As a result, nearly all the research pertaining to trade facilitation in regulatory
environment is concentrated on procedural improvements instead of trade performance
results. The purpose of this article is to tackle this limitation in the current literature by
supporting the trade facilitation argument with empirical analysis. In particular, the study
reports the outcomes of econometric tests reckoning the response of trade facilitation
measures on South-Asia’s intraregional trade. From a policy standpoint, the study's results
are aimed to assess the trade growth limitations, so that the prospective role of trade
facilitation reforms in regulatory environment in enhancing export performance compared
with other factors of trade development could be ascertained.
The Section 2 of this study narrates a review of the recent literature concerning trade
facilitation in regulatory environment apart from additional drivers of export performance
in South-Asia. Section 3 depicts the data used and methodology opted to fulfil the
objectives. Section 4 contains the discussion of results of the econometric model. Section
5 provides the methodology and the estimates of the trade potential among countries in
South-Asia. Section 6 discusses the various scenarios of improvement in the regulatory
environment of the region and simulation results of the same. Section 7 compares the
simulation results of tariff reduction and trade facilitation in regulatory environment in the
region. Finally, Section 8 concludes.
The estimation of economic outcome of trade facilitation is a significant analytical task.
The description and quantification of trade facilitation and the selection of a suitable
methodology for modelling the reaction of trade flows arising from trade facilitation is
quite challenging (Wilson et al., 2005). Djankov et al. (2010) noted that one extra day a
shipment is postponed on average lowers trade by around one per cent. Nordas et al. (2006)
examined the association among time taken to export and import, logistic facilities and
international trade to conclude that time delays lead to decreased volume of trade along
with decreasing likelihood of enterprises dealing in time-sensitive commodities for
entering export markets.
The significant part institutions play in dictating the success of reforms in policy measures
is widely recognised globally, and the ‘new’ Washington consensus provides enough
weightage to capacity building and institutional revamping (Rodrik, 2006). However, the
concept of institutions is still very abstract, since the notion of "successful governance"
itself is a multifaceted phenomenon. In a narrow framework, it could be said that
institutions determine the “rules of the game” for a community or, as per the popular
definition given by North (1990), are “the formal and informal constraints on political,
economic, and social interactions.” Through the aforementioned standpoint, “good”
institutions deploy a favourable framework which minimises uncertainty along with
improving productivity, thereby stimulating economic performance. In a broader spectrum,
institutions represent the discrete organizational bodies, procedural mechanism, and
regulatory environment that influences economic performance by endorsing superior
policy alternatives (IMF, 2003).
Based on the research of North (1990), numerous studies are now investigating the
influence of institutional features on international trade flows. In a study using gravity
model, Anderson and Marcouiller (2002) illustrate that robust institutions, specifically,
legal institutions efficient to enforce business agreements and unbiased devising and
implementing of state economic policy, promote growth in trade. Depken and Sonora
(2005) estimated the impact of economic freedom in the U.S. on exports and imports for
the years 1999 & 2000 using a gravity equation and observed that U.S. exports are
relatively more to the countries with better institutional quality. Iwanow and Kirkpatrick
(2007) used an augmented gravity model including variables pertaining to trade facilitation,
infrastructure and regulatory quality to measure their influence on export outcomes and
observed that 10 per cent growth in regulatory quality leads to 9-11 per cent increase in
exports.
Dollar and Kraay (2002) investigated the influence of institutional quality on external trade
and found positive high correlation between the two. Therefore, they further examined the
decadal changes in economic growth triggered by changes in institutions and trade. Their
study concluded that trade and institutions both influence economic growth in the long-run,
however growth in trade induces temporary benefits above institutional improvements in
stimulating economic growth. Jensen and Nordas (2004) also empirically verified that
overall level of openness is positively influenced by institutional quality of trading
countries. Chang et al. (2009) took a sample of large number of countries and ran growth
regression model using panel-data to conclude that the outcome of trade liberalization is
influenced a great deal by a comprehensive combination of institutions and policies.
Differences in behind-the-border infrastructure and regulatory quality along with on-the-
border procedures of trade facilitation supposedly cause considerable divergence in
transaction costs related to trade across countries. Levchenko (2007) states that divergence
in institutional quality of trading partners is a significant factor in determining trade
patterns which may originate comparative advantage on its own. On the other hand, another
study by De Groot et al. (2004) discerned that economies having similar attributes of
institutions tend to trade more.
Francois and Manchin (2007) examined the role of institutional quality, infrastructure,
trade preferences, and colonial and geographic backdrop in defining the structure of
bilateral trade. Emphasizing on those country pairs which do not actually trade and
controlling for other variables in this study, the authors suggested that institutional quality
and infrastructure are not just important determinants of volume of exports, but of the
possibility of exports happening at all as well. The study argues that both export
performance and the propensity to trade at all, is influenced by institutional quality.
Borrmann et al. (2006) examined the linkage between institutions, income and trade for a
sample of nations using ordinary least squares (OLS) approach and the findings of the study
depict that nations having poor status of institutions find it difficult to increase international
trade. Also, the regulatory quality (e.g., tax structure, market entry and labor market) plays
a greater role than good governance.
Apart from overall significance of institutions and regulatory quality, certain specific
aspects of institutions and regulatory environment are also studied by researchers to
measure their impact on trade. Bolaky and Freund (2004) singled out regulations that affect
business entry, bankruptcy and labour market flexibility and employed those in a growth
regression model that depicts that economies with good quality regulations experience rise
in growth rates due to international trade. By examining a particular aspect of institutions,
i.e. the impact of contract enforcement regulation on transaction cost of trade using a
gravity model, Ranjan and Lee (2007) reported that procedures of contract enforcement
influence the trade volumes of differentiated as well as homogenous products, largely for
differentiated goods.
3. Methodology
3.1. Data for regulatory environment indicator and its construction
To denote regulatory environment of a country, a Regulatory environment indicator (index)
is formulated in this study to evaluate the economy’s transparency, stringency,
implementation and quality of regulations and their impact on trade flows between that
country and its trading partner. This indicator is computed from data pertaining to each
South-Asian economy. Based on this indicator, a country can compare the level of its own
regulatory environment to that of other countries in South-Asia.
The most challenging aspect of research on trade facilitation is the lack of data series
corresponding to different measures of trade facilitation. Despite, the emergence of various
new data sources in the previous decade, the unavailability of historical data series,
inadequate coverage of LDC’s and discontinuation of several data series with time remain
the underlying problems in this research. Owing to these limitations, this analysis of South-
Asian region had to be constricted to five economies (India, Pakistan, Bangladesh, Nepal
and Sri Lanka) for which panel data of the variables used in computing the regulatory
environment indicator were available and the other member countries (Afghanistan, Bhutan
and Maldives) of the region were omitted. The sample of the economies taken in the study
is yet quite representative of the entire region as these are the largest five countries of
South-Asia with regards to trade volume and collectively represent 93.27 percent of South-
Asia’s overall intra-regional trade.
Also, the data series used in the computation of the indicator underwent major changes in
the methodology of their measurement in the year 2006 which makes the data preceding
2007 incomparable with the data of succeeding years, so the panel data analysed in this
study is from the year 2007 to 2016.
The data series employed in computation of regulatory environment indicator in this
analysis are drawn from two major databases – Worldwide Governance Indicators database
developed by World Bank (henceforth WGI) and Global Competitiveness Index database
of World Economic Forum (henceforth GCI). The structure of regulatory environment
indicator are as follows:
Regulatory environment for a country J is the average of five following “indexed inputs”:
 Transparency of government policy making (GCI).
 Stringency of environmental regulations (GCI).
 Enforcement of environmental regulations (GCI).
 Regulatory quality (WGI).
 Control of Corruption (WGI).
The description of the indices (“indexed inputs”) comprising above mentioned regulatory
environment indicator is provided in the Appendix.
The regulatory environment indicator in this analysis is computed with composite data of
five indices sourced from two different databases so as to restrict the impact of any
particular index on our composite index. However, the data belonging to the indices with
which the composite indicator is computed are measured on distinct scales and ranges (1
to 7, and -2.5 to 2.5). Therefore, the raw data must be normalized so as to make it
comparable.
Taking z-scores of the data series is one way of normalization but the limitation of this
technique is that it transforms the data and produces both positive and negative data inputs
and negative values are not compatible with log-log specification of the model that we
intend to use. So, here we decided on the normalization technique followed by Wilson et
al. (2003), which requires indexing of all the observations of a raw data series to the mean
of the raw data series’ values of each country pertaining to every year, producing an
“indexed input”. Therefore, an “indexed input” for a member country E (E = 1,2,3,4,5) is
computed as:
𝛱 𝛱 𝛱 /5
where 𝛱 indicates the raw data for a member country E. In this way, the countries
performing above the average of South-Asia shall have “indexed input” greater than 1 and
those performing below the average of South-Asia shall have “indexed input” less than 1.
These “indexed inputs” of each of the five indices used in the composite index are then
averaged with aggregation method of providing equal weights to each index to calculate
the regulatory environment indicator of trade facilitation.
Regulatory environment indicator constructed in this study finds its roots in Article X of
GATT pertaining to “Publication and Administration of Trade Regulations” rooted on
fundamental transparency obligation that calls for timely publication of regulations
governing exports and imports in order to provide a clear understanding of them to
stakeholders like governments, agents and traders of partner countries.
3.2. Data for trade flows and other variables
For the estimation of the model, the bilateral exports data for five economies in South-Asia,
namely, India, Bangladesh, Pakistan, Sri Lanka and Nepal is used in this study which are
in USD millions accessed through World Integrated Trade Solution (WITS) software from
the Commodity and Trade Database (COMTRADE) of the United Nations Statistics
Division.
The data for gross domestic product (GDP) and per capita GDP at constant prices is taken
from World Bank's World Development Indicators (WDI) database which makes them real
GDP & real per capita GDP and are expressed in USD millions.
The data pertaining to tariffs are drawn from World Trade Organisation (WTO) Tariff
database on Free-Trade Agreement duty rates for the agreement on SAFTA and weighted
average of all traded items are taken for each year. The data for tariff rates often has missing
values and to avoid a considerable fall in observations, the tariff data for the previous year
is taken in case of missing data for one year and the data is linearly interpolated or
extrapolated for a particular pair of trading countries when values for two or more
consecutive years remain missing.
The data for other variables used in the gravity model that are country specific, namely,
distance, and dummy variables including common language and common colony, are
obtained from Centre D’ Etudes Prospectives et D’ Informations Internationales (CEPII).
Missing values in trade data is an inevitable issue which is usually dealt with by replacing
all such values with minuscule values of trade flows, thereby depicting the values as nil
trade. Nevertheless, in case of logarithmic conversion of the series of data, the problem of
selection bias will arise. As suggested by Pusterla (2007), this problem could be curtailed
if all the missing values of trade data series are replaced with a very small figure and
subsequently it is added to one and then logarithmic conversion of this data series should
be done. This produces the final variable used in the model as log (Tie + 1), which will be
equal to zero in case Tie = 0. Tie represents trade data series here and it could be both exports
or imports.
By using the data series for GDP at current prices and GDP at constant prices, we calculated
the GDP deflator for each economy with their respective base years with the help of the
following formula:
GDP
GDP Deflator 100
GDP
With the help of the data series for GDP deflator for each economy with their respective
base years, we convert the nominal bilateral exports (NX from country I to country E in
year t into real bilateral exports 𝑋 by using the following formula:
NX
𝑋 100
GDP
3.3. The econometric model
Tinbergen (1962) and Pöyhönen (1963) propounded and developed gravity model used in
international trade research to examine bilateral trade flows by geographic distance
between the trading partners and their GDP (or GNP), on the premise of gravity equation
of physics given by Newton (1687). The theoretical foundations of this model were further
strengthened by Bergstrand (1985, 1989), Deardorff (1998), Anderson and Wincoop (2003)
and Helpman et al. (2008). Gravity model is one of the most frequently used technique of
modelling bilateral trade flows. The gravity equation hence derived from the newton’s
gravity equation is as follows:
𝑚 𝑚 𝐺𝐷𝑃 𝐺𝐷𝑃
𝐹 𝐺 ⇒ 𝐵𝑖𝑙𝑎𝑡𝑒𝑟𝑎𝑙 𝑇𝑟𝑎𝑑𝑒 𝛼
𝑟 𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒
In accordance with the standard regression analysis, the above equation is thereby
converted into linear form as follows:
log 𝐵𝑖𝑙𝑎𝑡𝑒𝑟𝑎𝑙 𝑇𝑟𝑎𝑑𝑒 𝛼 𝛽 log 𝐺𝐷𝑃 𝐺𝐷𝑃 𝛽 log 𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒 𝜇
In the standard gravity model specification, the logarithmic values of bilateral trade flows
are regressed on logarithmic values of distance between the exporting and importing
country pairs and the logarithmic values of their GDPs, and other variables explaining the
remaining variation (Maskus et al., 2001). Apart from distance and GDP, other factor
explaining bilateral trade flows could be per capita GDP, population, regional trade
agreements, common colony, common borders, common language or ethnic similarities.
The model used in this study includes primary economic variables of the gravity model,
i.e. the geographic distance between the country pairs and real GDP and real GDP per
capita of both importing and exporting countries. The standard gravity model specification
is further augmented with the corresponding tariff rate imposed by the importing country
on the exporting country and the regulatory environment indicator of trade facilitation
propounded in the study for both exporting and importing country along with the dummy
variables depicting common language and common colony. The augmented gravity model
used here is as follows:
Ln V β β ln 100 TARIFF β ln RE β ln RE β ln GDP

β ln GDP β ln GDPPC β ln GDPPC
β ln DISTANCE β D β D ε
where I is the importing country and E is the exporting country, β terms are coefficients
and t represents years of trading (i.e., t = 2007, ... 2016). The value of real bilateral exports
from country E to country I in the year t is represented as V . The term TARIFF indicates
the weighted average FTA tariff rate imposed by country I on country E in the year t. The
tariff variable is included in the model to decrease the excluded variable biases and it is
specifically important because in South-Asia (or SAFTA) FTA tariff rates are not
completely harmonized and they often vary across member countries and trading partners.
The term RE and RE depict the level of trade facilitation in regulatory environment
pertaining to the importing country I and the exporting country E in the year t. GDP and
GDP are the real GDP and GDPPC and GDPPC are the real GDP per capita of countries
I and E respectively for the year t. The geographic distance between the capital cities of
countries I and E is referred to as DISTANCE . Dummy variables D and D
reflect the impact of having a common language and being a common colony on the
bilateral trade flows between I and E. The dummy variables take the value 1 if country I
and E have a common language and were a common colony and 0 otherwise. ε depicts
the error term.
Typically, conventional gravity models utilize cross-section data for evaluating trade ties
during a specific timespan, be that as it may, it has its own negative points. On the other
hand, panel datasets i.e. cross-sectional variables recorded for various timeframes are
considered better than cross-sectional datasets while evaluating trade relations among
multiple pairs of trading partners in a time-frame as panel data models are the most efficient
models in separating time-invariant effects from the country-specific effects (Egger, 2002).
So, a panel gravity model is estimated in this study.
Additionally, the estimation of a panel regression model requires a decision among the
random effects model and fixed effects model to be used depending on the research
objectives and the variables included in the model. The fixed effects model has a limitation
that it cannot measure the time-invariant effects as it omits the time-invariant variables
before estimation, so if the objective is to measure the time-variant as well as invariant
effects in trade potential, then fixed effects model cannot be used and random effects model
has to be employed (Ozdeser & Ertac, 2010). Since, the distance and the dummy variables
pertaining to common language and common colony augmented in our regression model
are time-invariant variables which have fixed values over-time for a particular country pair,
fixed effects model cannot be employed in this scenario. Hence, a random effects model is
employed for estimation in this study.
4. Empirical results and analysis

It is significant to analyse the data series of “indexed inputs” used in composition of
regulatory environment indicator of trade facilitation first. Table 1 presents the descriptive
statistics for the regulatory environment indicator and the “indexed inputs”. The descriptive
statistics of the “indexed input” series and the aggregated indicator portray the range of
each “indexed input” and the indicator. It could be observed where an economy lies in that
range from highest to lowest value of the region. India has the highest level of transparency
and stringency in regulatory standards, whereas Nepal has the lowest. Sri Lanka is the best
performing country in terms of enforcement of these standards. Bangladesh has the highest
level of regulatory quality and control of corruption. Bangladesh is also the country having
the highest value of overall regulatory environment indicator, whereas Sri Lanka has the
lowest overall value.
Table 1. Descriptive statistics of regulatory environment indicator and its components
Indicator and components Source Sta Dev Min Economy Max Economy
Regulatory Environment
Transparency of government policy making WEF: GCI 0.09 0.85 Nepal 1.21 India
Stringency of regulatory standards WEF: GCI 0.16 0.77 Nepal 1.31 India
Enforcement of environmental regulations WEF: GCI 0.17 0.72 Bangladesh 1.33 Sri Lanka
Regulatory quality WB: WGI 0.46 0.09 Sri Lanka 1.79 Bangladesh
Control of corruption WB: WGI 0.41 0.24 Sri Lanka 1.64 Bangladesh
Aggregate Index 0.10 0.79 Sri Lanka 1.19 Bangladesh
Note: Mean of all indexed inputs is 1. WEF: World Economic Forum; GCI: Global Competitiveness Index
database; WB: World Bank; WGI: Worldwide Governance Indicators database.
Source: Authors’ computations based on data from sources mentioned.
Table 2 provides the correlation matrix of the “indexed inputs” pertaining to the aggregate
regulatory environment indicator. The correlation matrix indicates the optimum use of
multiple indices in quantifying regulatory environment indicator of trade facilitation to
restrict the influence of a single index in its quantification. The high value of correlation
(as high as 0.95) of the “indexed inputs” series indicates the robustness of the regulatory
environment indicator in regards to source of data and since the correlation values are not
perfect 1 in any case, it rationalizes the use of multiple indices (or “indexed inputs”) for
calculation of the indicator.
Table 2. Correlation matrix of indexed inputs of regulatory environment indicator
Indexed inputs of Regulatory Environment
Transparency of Stringency of Enforcement of Regulatory Control of
government regulatory environmental Quality corruption
policy making standards regulations
Transparency of government
1.00
policy making
Stringency of regulatory
0.77 1.00
standards
Enforcement of environmental
0.69 0.95 1.00
regulations
Regulatory Quality 0.61 0.88 0.92 1.00
Control of corruption 0.72 0.81 0.88 0.85 1.00
Source: Authors’ computations.
An augmented panel gravity model is estimated by integrating self-constituted trade

facilitation indicator (regulatory environment indicator for both exporting and importing
countries), tariff variable and dummy variables along with basic variables of gravity model,
i.e. GDP, per capita GDP of trading countries and the distance between them. Table 3
displays the regression results which exhibits that the value of overall R2 of the estimated
model is 0.79 which is quite high and it reflects that the regression model is successful.
Unlike other indicators of trade facilitation including transport efficiency, customs

environment and service-sector infrastructure which always yield positive outcomes for
bilateral trade flows, the regulatory environment indicator could yield positive as well as
negative outcomes for the importing or exporting countries. According to the “sanders”
and “greasers” hypothesis of corruption, quality regulations on one hand work as a
“greaser” for the wheels of the economy through faster movement of goods, whereas, on
the other hand the improved quality of regulations and their stringency and transparency
could work as a “sander” for movement of goods across borders.
Those who advocate the “greasers” hypothesis proclaim that corruption and flaws in
regulations facilitate trade by improving efficiency by enabling private sector entities to
evade burdensome regulations (Leff, 1964; Méon and Weill, 2010). Others who endorse
the “sanders” hypothesis state that rampant corruption and lenient and ambiguous
regulations repel traders and investors from the economy (Mauro, 1995). This debate is still
ongoing and a number of studies like Bardhan (1997), Pande (2008), Aidt (2009) are
dedicated to it. However, it has largely been inconclusive.
In our analysis, the estimated coefficients of regulatory environment indicator for both the
exporting as well as the importing countries are significant and of high magnitude but of
opposite signs, i.e. negative for exporting country and positive for importing country. This
suggests that regulatory environment will distinctively affect trade flows of individual
countries depending upon whether it is an importing country or an exporting country in a
specific scenario. From a policy perspective, this approach is much more beneficial as it
provides specific directions to work upon.
Table 3. Augmented panel gravity model results
Variable Coefficient P Values
Constant 1.249 0.961
Regulatory Environment of exporting country -5.214*** 0.000
Regulatory Environment of importing country 3.229** 0.014
Tariff -5.461* 0.056
Real GDP of exporting country 1.317*** 0.000
Real GDP of importing country 0.902** 0.015
Real GDP per capita of exporting country -1.243** 0.014
Real GDP per capita of importing country -0.630 0.207
Geographic Distance -0.919 0.493
Common Language dummy variable 2.744* 0.078
Common Colony dummy variable 2.447* 0.081
R2 within 0.1304
R2 between 0.8163
R2 overall 0.7942
Prob > Chi2 0.000
No. of observations 200 (20 groups)
Note: *Significant at the 10 percent level. **Significant at the 5 percent level. ***Significant at the 1 percent
level. All the variables (dependent and independent) are in logarithms.
Source: Authors’ calculation.
Coefficient of regulatory environment indicator of trade facilitation for exporting country

is (-5.21) which means that bilateral trade flows to importing country from exporting
country are negatively associated with regulatory environment of exporting country and
one percent improvement in regulatory environment indicator of trade facilitation of
exporting country will reduce bilateral trade flows by five percent. On the other hand,
coefficient of regulatory environment indicator of trade facilitation for importing country
is (3.22) which means that bilateral trade flows are positively associated with regulatory
environment of importing country and one percent improvement in regulatory environment
indicator of trade facilitation of importing country will increase bilateral trade flows by
three percent. As predicted, tariffs affect bilateral trade flows negatively and significantly.
On the whole, the analysis implies that trade facilitation in regulatory environment generate
opposite outcomes for a country in case of exports and imports. If a country ensures
improvement in its regulatory environment in terms of transparency, stringency,
enforcement and quality of regulations along with controlling corruption, then it leads to
increase in its imports as it facilitates the exporters from other countries to export their
goods to this country easily.
However, this improvement in regulatory environment of a country works the other way
around for its own exports to other countries as the transparency, stringency and strict
enforcement of regulations with low level of corruption could hinder the faster movement
of goods from the home country to the other countries. These results are in line with the
results of Van Beers and van den Bergh (1997) who examined the consequence of
environmental regulations on a nation’s imports and exports using a gravity model and
concluded that precisely defined environmental stringency variable significantly affects a
country’s exports negatively. Similarly, the results of the present study are supported by
the results of Lee and Weng (2013) who scrutinized the effect of bribery in a country on
firm exports of that country by following three-stage least squares method to test the
contrasting hypotheses. One of the hypotheses formulated by the study stated that bribery
leads to preferential treatment to firms by government officials which enhances their
efficiency to perform competently in foreign markets. Whereas, the other hypothesis stated
that bribery provides preferential treatment within domestic market and it reduces the
firm’s incentive to dig into foreign markets, thereby reducing exports. The study concluded
that bribery reduces the firm exports from home country instead of increasing them.
The estimated coefficients of basic variables also confirm to the ordinary features of gravity
model. The coefficients of real GDP of both the importing and exporting countries in the
model are positive and significant showing that exports and imports of a country tend to
rise with increase in the size of the economy. The coefficient of real GDP of exporting
country is 1.3 and that of importing country is 0.9 suggesting that the elasticity of trade
volume corresponding to GDP of exporting country is higher as compared to that of GDP
of importing country. The coefficients of real per capita GDP of both the exporting and
importing country are negative depicting that it negatively affects trade flows. The
estimated coefficient of distance in the model is negative (-0.91) as anticipated which
implies that trade flows are negatively associated with distance and a country is more
inclined to trade with countries that are closer than with the far ones. The dummy variables
depicting the common official language and common colonial history of trading partners
also possess positive and significant coefficients, i.e. 2.74 and 2.44, respectively. This
reflects that countries with same official language and common colonial background are
likely to trade more.
5. Estimating trade potential within South-Asia

For further analysis, the results of augmented panel gravity model for intraregional trade in
South-Asia displayed in Table 3 are used to estimate trade potential of each member
country in the study within South-Asia. Trade potential could be defined as maximum level
of trade possible between two countries which have liberalized their trade restrictions
(Kalirajan, 1999). Based on the estimated coefficients of augmented panel gravity model,
the trade potential (P) for each of the five countries with one another within South-Asia is
calculated. The recorded actual trade (A) represents the present volume of trade taken place
with ongoing set of restrictions and institutions. The difference between potential (P) and
actual (A) trade arises due to several institutional and socio-economic factors that impede
actual trade to rise to the highest point of export production frontier.
To evaluate whether a pair of two countries has unrealised or untapped trade potential
between them, the ratio of trade potential (P) and actual trade (A) is calculated, i.e. (P/A)
as per Batra (2006). If this ratio (P/A) exceeds 1 then we interpret this as unrealised or
untapped trade potential between a member country of South-Asia with its trading partner
country within South-Asia and that the trade can be expanded. On the other hand, if the
ratio (P/A) is less than 1 for a given country then it implies that this country has exceeded
its trade potential with its trading partner. So, this ratio (P/A) indicates whether a country
has potential for increasing trade with other countries or not.
Further, to estimate the actual figure of how much the trade between two countries could
be expanded we have calculated the difference between the potential trade (P) and the
actual trade (A), i.e. the magnitude of (P-A). This figure (P-A) depicts whether a country
has potential for expanding its trade with its partner country or it has exceeded its potential
for trade with the partner country. If (P-A) is positive, it indicates that the country has
unrealised trade potential and there exists an opportunity for this country to realise this
untapped trade potential with its trading partner. Contrarily, if (P-A) is negative, it indicates
that the country has already exceeded its potential for trade with its partner country. The
absolute figures denote the magnitude of possible expansion of export trade (if positive) or
the magnitude of over-trade (if negative) within two countries in US $ millions.
5.1. Analysis of estimated trade potential within South-Asia
The results of estimated trade potential within South-Asia are displayed in Table 4 and
Table 5. The results depict the export potential of each of the five countries of South-Asia
taken in analysis with the rest of the four countries within the region.
Table 4 displays the results of trade potential (measured by the ratio P/A) within the region.
The results depict that out of the five countries, Bangladesh is the only country which has
potential of expanding its exports to all the four countries as the ratio values are greater
than one for all the four countries. It means that Bangladesh is the country with the greatest
potential. On the other hand, Pakistan has unrealised export potential only with Nepal and
it is outperforming its export potential with the other three countries. Sri Lanka has
unrealised export potential with all the three countries (Bangladesh, Nepal and Pakistan)
except India. Nepal has export potential with Pakistan and Sri Lanka as it does not share a
common border with these two countries, whereas it is surpassing its export potential with
India and Bangladesh. Despite sharing a land border with Bangladesh and Pakistan, India
has untapped export potential with both the neighbouring countries, however it is exceeding
its export potential to Nepal and Sri Lanka.
Table 4. Trade potential of countries within South-Asia in ratios (P/A)
Bangladesh India Nepal Pakistan Sri Lanka
Bangladesh - 2.56 3.68 6.27 1.05
India 4.94 - 0.14 2.04 0.38
Nepal 0.49 0.06 - 4.81 2.13
Pakistan 0.70 0.73 9.87 - 0.13
Sri Lanka 1.05 0.51 1.05 1.23 -
Source: Author’s computations based on augmented panel gravity model results.
The results of actual trade potential (P-A measured in US$ millions) of member countries
with each other in South-Asia are given in Table 5. As revealed by the ratio analysis of
trade potential, Bangladesh has positive export potential with all the four countries of the
region. The magnitude of expanding exports from Bangladesh is also quite huge, i.e.
approximately US$ 1.36 billion out of which US$ 1.14 billion arises only out of its exports
to India. Not only this, India also shows tremendous gap between potential trade and actual
trade, i.e. US$ 16.92 billion with Bangladesh. This reflects the dire need of facilitating trade
between these two countries from both the sides, i.e. exporter’s side as well as importer’s
side. Facilitation of trade between these two countries could ensure major contribution to
intra-regional trade in South-Asia. Apart from this, India also has export potential with
Pakistan of US$ 1.35 billion. Nepal has positive export potential only with Pakistan of
about US$ 2.99 million which is a small amount owing to its economy size. Pakistan too
has a small magnitude of positive export potential only with Nepal, i.e. US$ 6.77 million.
Likewise, Sri-Lanka also has small positive export potential with Bangladesh of US$ 4.58
million, Pakistan of US$ 12.47 million and Nepal of US$ 0.06 million.
Table 5. Trade potential of countries within South-Asia in value (US$ millions) (P-A)
Bangladesh India Nepal Pakistan Sri Lanka
Bangladesh - 1,149.80 6.48 209.00 1.34
India 16,927.53 - -3,667.10 1,355.08 -2,486.05
Nepal -22.38 -395.73 - 2.99 0.23
Pakistan -146.94 -101.22 6.77 - -202.95
Sri Lanka 4.58 -403.17 0.06 12.47 -
The trade potential analysis provides us with a significant conclusion that trade facilitation
provides tremendous scope in South-Asia in increasing intra-regional trade and country
pairs like India-Bangladesh, India-Pakistan, Bangladesh-Pakistan and Sri Lanka-Pakistan
are the most promising ones if they facilitate trade across the borders for their regional
trading partners.
6. Estimating potential gains from trade facilitation in regulatory environment in South-Asia:
Simulation approach
In this section, we quantify the potential gains in intra-regional trade in South-Asia by
exploring few scenarios of trade facilitation initiatives concentrated on regulatory
environment. The objective is to estimate the gains in intra-regional trade arising out of
different scenarios of trade facilitation in regulatory environment, so as to inform the
policy-makers about the underlying potential of each initiative of trade facilitation in

increasing trade flows within the region.
6.1. Simulation design
The augmented panel gravity model operated in the previous section of the study enable us
to carry out simulations to quantify the gains in intra-regional trade in South-Asia resulting
from specific scenarios of improved regulatory environment indicator of trade facilitation.
The simulations conducted in this section are based on three different scenarios. First
scenario is the basic scenario which requires that out of all the member countries taken in
the analysis, the countries with below average Regulatory environment indicator should be
brought up to the regional average of the indicator and the countries already performing
above regional average of the indicator should not be altered, so that all the countries are
either at the regional average or above average of the Regulatory Environment indicator as
done by Wilson et al. (2003).
The emphasis here, is on below-average performers in the region since these countries bear
lower scores of Regulatory Environment indicator and need substantial efforts for capacity-
building as compared to above-average performers. The drawback of this scenario is that
it does not necessitate the improvement in regulatory environment of the countries which
are already performing above average which may get even better results for increasing trade
volumes if improved. Therefore, the next two scenarios root on uniformly improving trade
facilitation indicator of Regulatory Environment of all the countries in the region by a
common percentage.
Second and the third scenarios dictate that the Regulatory Environment indicator of trade
facilitation of all the countries taken in the analysis should be improved by ten per cent and
fifteen per cent of the original values, respectively. The important thing to note here is that
in all the three scenarios of improving Regulatory environment indicator, improvement
means increasing the value of the indicator for importing country and decreasing the value
of the indicator for exporting country at the same time. This is because the variable
Regulatory environment indicator has positive coefficient for importing country and
negative for exporting country. Therefore, improvement in regulatory environment for
importing country means increase in the value of the indicator and that for the exporting
country means decrease in the value of the indicator.
6.2. Simulation analysis
Simulations results for three scenarios of improved Regulatory Environment indicator of
trade facilitation to quantify the gains in intra-regional trade in South-Asia are displayed in
Table 6. The first scenario which required that the countries with below average Regulatory
environment indicator within the region are brought up to the regional average of the
indicator and the countries already performing above regional average of the indicator are
not altered, led to gain in total trade flows across the region by US$ 17.06 billion, which
accounts for 99.30 percent of the total intraregional trade in South-Asia. Such a
phenomenal change in the total intra-regional trade resulting from improving the regulatory
environment of only the below average countries has taken place partly because three out
of five countries taken in analysis were lying below regional average in actual scenario and
partly because of the large gap in the score of regulatory environment indicator of worst
performing economy and the regional average. However, the limitation of this scenario is
that the above average performing countries are not improving their regulatory
environment.
Table 6. Results for simulations done on regulatory environment indicator of trade facilitation
Trade facilitation Goal Change in trade flow
measure of Amount Share of total
simulation ($ billion) trade (%)
All Countries at Below average countries increase RE of importing countries up 17.06 99.30
South-Asia’s to South-Asia’s average and above average countries decrease
Average RE of exporting countries to South-Asia’s average
All Countries at 10% Increase RE of all the importing countries by 10% and decrease 52.22 303.95
Improved RE RE of all the exporting countries by 10%
All Countries at 15% Increase RE of all the importing countries by 15 % and decrease 90.75 523.86
Improved RE RE of all the exporting countries by 15%
The second and the third scenarios of simulations are based on uniformly improving trade
facilitation indicator of Regulatory Environment of all the countries by a common
percentage. In the second scenario, we improve the Regulatory Environment indicator of
all countries by 10 percent which brings all the member countries above regional average
except one country despite improvement in their Regulatory Environment indicator. In this
scenario, the gain in South-Asia’s total intra-regional trade is US$ 52.22 billion, which
constitutes 303.95 percent of total intra-regional trade. This indicates that when all the
countries of the region including the below average, above average and the best performer
in the region further improve their Regulatory Environment by merely 10 percent, then it
results in intra-regional trade becoming four-times of the actual trade.
In the third scenario, we improve the Regulatory Environment indicator of all the member
countries by 15 percent which brings all the five countries in the region above regional
average of Regulatory Environment indicator. In this scenario, the gain in South-Asia’s
total intraregional trade is US$ 90.75 billion, which claims a rise in 523.86 percent. This
gain is more than five-times the actual level of South-Asia’s intraregional trade. It depicts
that when all the countries in the region further improve their Regulatory Environment so
much so that all of them reach a level that is above current regional average, then the gains
in intra-regional trade are enormous.
The results of these simulations suggest that considering South-Asia already has a very low
share of intra-regional trade to that of its total trade, i.e. 6.1 percent, there exists enormous
scope for increasing this share through trade facilitation measures aimed at improving
Regulatory Environment of the countries in the region.
7. Tariff reduction versus trade facilitation

It is often assumed that tariff barrier to trade is the most significant barrier to trade since it
is the most tangible form of impediment to trade in terms of trade costs. However, in the
past few decades after phasing out of tariff barriers owing to WTO obligations and several
regional trade agreements, the emphasis has been shifted to other non-tariff barriers
including regulatory measures. In this section, we conduct yet another set of simulations to
calculate the gains in intra-regional trade in South-Asia arising from the scenarios of tariff
reduction so as to compare the gains with those arising from trade facilitation in regulatory
environment. The two scenarios of simulations examined here are reduction of weighted
average FTA tariffs in South-Asia by half and up to zero. This analysis helps us to quantify
the gains in intra-regional trade owing to reduction of tariffs by half and by completely
phasing out of tariffs from the region.
The results of the tariff simulations are provided in Table 7. The results indicate that if the
bilateral tariffs in the region applied by the member countries on each other are reduced by
half, then it leads to gain in intra-regional trade in South-Asia by US$ 23.14 billion which
accounts for 134.69 percent of total intra-regional trade. It is interesting to note that these
gains are comparable to the gains in intra-regional trade achieved by trade facilitation in
regulatory environment indicator of below average member countries of the region up to
the regional average. Now, the policy makers can decide as to which measure would they
be opting to generate these gains in intra-regional trade and it depends upon the cost to be
incurred on achieving this kind of Regulatory environment trade facilitation. Whichever
option out of the two demands least cost shall be chosen and followed.
Table 7. Results for simulations done on tariff rate imposed by member countries in South-Asia (or SAFTA)
Measure of Goal Change in trade flow
simulation Amount Share of total
($ billion) trade (%)
Half Tariff Decrease the actual bilateral Applied Tariff Rate imposed by 23.14 134.69
member countries on each other by half
Zero Tariff Decrease the actual bilateral Applied Tariff Rate imposed by 39.21 228.23
member countries on each other to zero
The next scenario of tariff reduction for simulation dictates complete phasing out of
bilateral tariffs for member countries of South-Asia. This scenario generates gains in intra-
regional trade of about US$ 39.21 billion which constitutes about 228.23 percent of the
actual intra-regional trade in South-Asia at present. If we compare this gain with the gain
in intra-regional trade resulting out of trade facilitation in Regulatory environment of all
the countries by 10 percent, then it depicts that gains from trade facilitation are larger than
gains in intra-regional trade achieved by completely phasing out of bilateral tariffs within
the region. This suggests that in case tariff abatement or phasing out is not possible then
trade facilitation could prove to be a favourable alternative policy. This also suggests that
reducing tariffs to increase the trade within the region are not sufficient enough to realise
full potential of intra-regional trade. Phasing out of tariffs is in fact less effective than trade
facilitation in regulatory environment and if the two measures are implemented
simultaneously, then the potential gains in intra-regional trade could be phenomenal.
8. Summary and conclusion

Being the least integrated region of the world with total intra-regional trade amounting to
just 6.1 percent of the total world trade of the region, South-Asia retains enormous scope
for collective gains in international trade and economic growth for all the member countries
of the region. Other than phasing out bilateral tariffs within the region, the commitment
towards capacity building in trade facilitation is inescapable to increase regional trade. The
formation of the South Asian Association for Regional Cooperation (SAARC) and its
further extension to South Asian Free Trade Area (SAFTA) was perceived as a positive
step in this direction. However, it could not yield the desired results due to the presence of
intangible barriers to trade within the region like regulatory barriers. The removal of these
barriers requires strategic efforts in the direction of trade facilitation.
There were three main objectives behind this study. The primary objective of the analysis
was to assess the impact of trade facilitation in regulatory environment on South-Asia’s
intraregional trade. To achieve this objective, an augmented panel gravity model was run
to estimate the bilateral trade flows within the region depending upon basic variables of
gravity model and regulatory environment indicator of trade facilitation for both importing
and exporting countries. The results of the regression model suggest that regulatory
environment indicator of trade facilitation is an important variable in determining the trade
flows within the region and there exists immense potential for gains in intra-regional trade
in South Asia emerging from collective efforts to boost capacity in trade facilitation in
regulatory environment. The regulatory environment indicator of the importing country
impacts trade flows positively, whereas, the regulatory environment indicator of the
exporting country impacts trade flows negatively. One percent improvement in regulatory
environment indicator of trade facilitation of exporting country will reduce bilateral trade
flows by five percent and that of importing country will increase bilateral trade flows by
three percent.
The second objective of the analysis was to estimate the trade potential of South-Asian
countries with each other. The results of trade potential analysis reveal that facilitation of
trade between India and Bangladesh from both exporter’s side as well as importer’s side
could generate gains in intra-regional exports by US$ 18.28 billion combined. Apart from
this, India and Bangladesh both have immense export potential to Pakistan of US$ 1.35
billion and US$ 209 million respectively. Therefore, the most promising countries in
South-Asia for increasing intra-regional trade through trade facilitation in Regulatory
environment are India, Bangladesh and Pakistan.
The third and the final objective of the study was to estimate the payoffs in South-Asia’s
total intraregional trade by running simulations on several scenarios of improving the
Regulatory Environment indicator through trade facilitation and reduction in tariffs within
the region and by comparing the outcomes of the two measures. The simulations results
indicate that if only the below average performing countries in South-Asia in Regulatory
Environment indicator improve their score for Regulatory Environment by trade
facilitation up to the regional average score then it can increase the intra-regional trade by
US$ 17.06 billion, which accounts for 99.30 percent of South-Asia’s total intraregional
trade. Further, if all countries in the region improve their Regulatory Environment indicator
by 10 percent and 15 percent through trade facilitation then the gains are nearly US$ 52.22
billion and US$ 90.75 billion, which constitutes 303.95 and 523.86 percent of total intra-
regional trade, respectively. Apart from this, slashing the bilateral applied tariffs in the
region by half could foster gains in intra-regional trade by US$ 23.14 billion, i.e. 134.69
percent of total intra-regional trade and complete phasing out of tariffs can generate gains
of US$ 39.21 billion, i.e. 228.23 percent of total intraregional trade in the region at present.
The overall analysis highlights the significant part to be played by India in carrying out
trade facilitation reforms in South-Asia as it represents 82.53 percent of the aggregate GDP
of South-Asia and could lead by example to promote the agenda of collective trade
facilitation in the region. Bangladesh’s economy seems to be having hostile regulatory
environment for India’s exports which needs to be corrected through trade facilitation.
Other significant deterrent in intraregional trade in South-Asia is posed by Pakistan as both
India and Bangladesh have immense potential for expanding exports to Pakistan. So,
Pakistan should focus on trade facilitation in regulatory environment for its own interest as
well as for larger gains to the region as a whole. Macroeconomic stability and bilateral
cooperation between India and Pakistan is also very important for encouraging regional
cooperation. It is obvious as per the statistics that regional integration can be promoted,
however, with coherent efforts of all the countries in the region to address barriers to trade
facilitation in regulatory environment.
References
Acemoglu, D., Johnson, S. and Robinson, J.A., 2001. The colonial origins of comparative develop-
ment: An empirical investigation, American Economic Review, Vol. 91, No. 5, pp. 1369-1401.
Aidt, T., 2009. Corruption, Institutions, and Economic Development, Oxford Review of Economic
Policy, Vol. 25, No. 2, pp. 271-291.
Anderson, J.E. and Marcouiller, D., 2002. Insecurity and the pattern of trade: An empirical
investigation, Review of Economics and statistics, Vol. 84, No. 2, pp. 342-352.
Anderson, J.E. and Van Wincoop, E., 2003. Gravity with gravitas: a solution to the border puzzle,
American economic review, Vol. 93, No. 1, pp. 170-192.
Bardhan, P., 1997. Corruption and Development: A Review of Issues, Journal of Economic
Literature, Vol. 35, No. 3, pp. 1320-1346.
Batra, A., 2006. India's global trade potential: The gravity model approach, Global Economic
Review, Vol. 35, No. 3, pp. 327-361.
Bergstrand, J.H., 1985. The Gravity Equation in International Trade: Some Microeconomic
Foundations and Empirical Evidence, Review of Economics and Statistics, Vol. 67,
pp. 474-481.
Bergstrand J.H., 1989. The Generalized Gravity Equation, Monopolistic Competition, and the
Factor-Proportions Theory in International Trade, Review of Economics and Statistics, Vol. 71,
pp. 143-153.
Bolaky, B. and Freund, C., 2004. Trade, regulations, and growth, The World Bank.
Borrmann, A., Busse, M. and Neuhaus, S., 2006. Institutional quality and the gains from trade,
Kyklos, Vol. 59, No. 3, pp. 345-368.
Chang, R., Kaltani, L. and Loayza, N.V., 2009. Openness can be good for growth: The role of policy
complementarities, Journal of development economics, Vol. 90, No. 1, pp. 33-49.
De Groot, H.L., Linders, G.J., Rietveld, P. and Subramanian, U., 2004. The institutional
determinants of bilateral trade patterns, Kyklos, Vol. 57, No. 1, pp. 103-123.
Deardorff, A., 1998. Determinants of bilateral trade: does gravity work in a neoclassical world?, in
The regionalization of the world economy (pp. 7-32), University of Chicago Press.
Deardorff, A.V., 2014. Local comparative advantage: Trade costs and the pattern of trade,
International Journal of Economic Theory, Vol. 10, No. 1, pp. 9-35.
Depken II, C.A. and Sonora, R.J., 2005. Asymmetric effects of economic freedom on international
trade flows, International Journal of Business and Economics, Vol. 4, No. 2, pp. 141-155.
Djankov, S., Freund, C. and Pham, C., 2010. Trading on time, Review of Economics and Statistics,
Vol. 92, No. 1, pp. 166-173.
Dollar, D. and Kraay, A., 2003. Institutions, trade, and growth, Journal of monetary economics,
Vol. 50, No. 1, pp. 133-162.
Egger, P., 2002. An econometric view on the estimation of gravity models and the calculation of
trade potentials, The World Economy, Vol. 25, No. 2, pp. 297-312.
Engman, M., 2005. The economic impact of trade facilitation, OECD Trade Policy Working Paper
no 21, OECD: Paris.
Francois, J. and Manchin, M., 2007. Institutions, infrastructure, and trade, The World Bank.
Helpman, E., Melitz, M. and Rubinstein, Y., 2008. Estimating trade flows: Trading Partners and
Trading Volumes, Quarterly Journal of Economics, Vol. 123, No. 2, pp. 441-487.
International Monetary Fund, April, 2003. World Economic Outlook: Growth and Institutions
available at <https://www.imf.org/external/pubs/ft/weo/2003/01/>
Iwanow, T. and Kirkpatrick, C., 2007. Trade facilitation, regulatory quality and export performance,
Journal of International Development, Vol. 19, No. 6, pp. 735-753.
Jansen, M. and Nordas, H., 2004. Institutions, Trade Policy and Trade Flows, Staff Working Paper
ERSD-2004-02, World Trade Organization, Geneva.
Kalirajan, K., 1999. Stochastic varying coefficients gravity model: an application in trade analysis,
Journal of Applied Statistics, Vol. 26, No. 2, pp. 185-193.
Lee, S.H. and Weng, D.H., 2013. Does bribery in the home country promote or dampen firm
exports?, Strategic Management Journal, Vol. 34, No. 12, pp. 1472-1487.
Leff, N.H., 1964. Economic development through bureaucratic corruption, American behavioral
scientist, Vol. 8, No. 3, pp. 8-14.
Levchenko, A.A., 2007. Institutional quality and international trade, The Review of Economic
Studies, Vol. 74, No. 3, pp. 791-819.
Mauro, P., 1995. Corruption and Growth, The Quarterly Journal of Economics, Vol. 110, No. 3,
pp. 681-712.
Maskus, K.E., Wilson, J.S. and Otsuki, T., 2001. An Empirical Framework for Analyzing Technical
Regulations and Trade, in Maskus, K.E. and Wilson, J.S. (eds.), Quantifying the Impact of
Technical Barriers to Trade: Can It Be Done? Ann Arbor: University of Michigan Press.
Méon, P.G. and Weill, L., 2010. Is corruption an efficient grease?, World development, Vol. 38,
No. 3, pp. 244-259.
Norda’s, H., Pinali, E. and Geloso Grosso, M., 2006. Logistics and time as a trade barrier, OECD
Trade Policy Working Paper No. 35.
North, D.C., 1990. Institutions, Institutional Change and Economic Performance, Cambridge
University Press: Cambridge.
Newton, I., 1687. Philosophiæ Naturalis Principia Mathematica. Retrieved from University of
Cambridge, Cambridge Digital Library. Retrieved from <http://cudl.lib.cam.ac.uk/view/PR-
ADV-B>
Obstfeld, M. and Rogoff, K., 2000. The six major puzzles in international macroeconomics: is there
a common cause?, NBER Macroeconomics Annuals, Vol. 15, pp. 339-390.
Ozdeser, H. and Ertac, D., 2010. Turkey’s trade potential with euro zone countries: A gravity study,
European Journal of Scientific Research, Vol. 43, No. 1, pp. 15-23.
Pande, R., 2008. Understanding Political Corruption in Low Income Countries, in T. Schultz and J.
Strauss (eds.), Handbook of Development Economics (Volume 4), Elsevier.
Pöyhönen, P., 1963. A tentative model for the volume of trade between countries, Weltwirtschaftliches
Archiv, Vol. 90, No. 1, pp. 93-100.
Pusterla, F., 2007. Regional integration agreements: Impact, geography and efficiency, Inter-
American IDB-SOE Working Paper No. 1. Paris: IDB-SOE.
Rodrik, D., 2006. Goodbye Washington consensus, hello Washington confusion? A review of the
World Bank's economic growth in the 1990s: learning from a decade of reform, Journal of
Economic literature, Vol. 44, No. 4, pp. 973-987.
Ranjan, P. and Lee, J.Y., 2007. Contract enforcement and international trade, Economics & Politics,
Vol. 19, No. 2, pp. 191-218.
Silberberger, M. and Königer, J., 2016. Regulation, trade and economic growth, Economic Systems,
Vol. 40, No. 2, pp. 308-322.
Staples, B.R., 2002. Trade facilitation: improving the invisible infrastructure, in Hoekman B.,
Mattoo A., English P. (eds.), Development, Trade and the WTO: A Handbook, World Bank:
Washington DC.
Tinbergen, J., 1962. Sharing the World Economy: Suggestions for an International Economic
Policy, New York: Twentieth Century Fund.
Van Beers, C. and Bergh van den, J.C.J.M., 1997. An Empirical Multi-country Analysis of the
Impact of Environmental Regulations on Foreign Trade Flows, Kyklos, Vol. 50, No. 1,
pp. 29-46.
Wilson, J.S., Mann, C.L. and Otsuki, T., 2003. Trade Facilitation and Economic Development: A
New Approach to Quantifying the Impact, World Bank Economic Review, Vol. 17 No. 3,
pp. 367-89.
Wilson, J.S., Mann, C.L. and Otsuki, T., 2005. Assessing the Potential Benefit of Trade Facilitation:
A Global Perspective, The World Economy, Vol. 28, No. 6.
Appendix
1. Transparency of government policy making (WEF) measures how easy is it for
businesses to obtain information about changes in government policies and regulations
affecting their activities in a country, ranging from 1=extremely difficult to 7=extremely
easy.
2. Stringency of environmental regulations (WEF) measures perceived stringency of a
country’s environmental regulations, ranging from 1=very lax to 7=among the world’s
most stringent.
3. Enforcement of environmental regulations (WEF) measures perceived rigorousness
of enforcement of a country’s environmental regulations, ranging from 1=very lax to
7=among the world’s most rigorous.
4. Regulatory quality (World Bank WGI) captures perceptions of the ability of the
government to formulate and implement sound policies and regulations that permit and
promote private sector development, ranging from -2.5=very low to +2.5=very high.
5. Control of corruption (World Bank WGI) captures perceptions of the extent to which
public power is exercised for private gain, including both petty and grand forms of
corruption, as well as “capture” of the state by elites and private interests, ranging from
-2.5 – very low to +2.5 –very high.
India and trade blocs: A gravity model analysis
Arjun SINGH
Panjab University Chandigarh, India
arjunmanhas25@gmail.com
Dr. S.P. PADHI
Panjab University Chandigarh, India
litton@pu.ac.in
Abstract. This paper analyses the trading relationship of India with major trade blocs which
includes EU, NAFTA and ASEAN weather the trade flow indeed affected by the GDP (Income),
Distance, tariff and per capita GDP by using gravity model on the panel data from the period of
1991 to 2017. The panel data is examined by the Multi-level mixed effect model with linear
regression and ML method of estimation is used for estimating the model. The study will attempt to
examine the trade bloc which has significant trade with India as per the gravity model is a
concern. The study found that the distance negatively affecting the trade flow for EU and NAFTA
which is as per theoretically expectations while in case of ASEAN its positively affecting the trade
flow which is not as per theoretical expectations, However the coefficient is not significant. GDP
is negatively and insignificantly affecting the trade flow between India and ASEAN and NAFTA
whereas GDP is negatively and significantly affecting the trade flow between India and EU.
Keywords: EU, NAFTA, Economic Integration, ASEAN, Gravity Model, Trade.

JEL Classification: F16, F100, F120, F140, F150.
218 Arjun Singh, Dr. S.P. Padhi
1. Introduction
Gravity model is a well established model to study India’s trade patterns with the
developed trade bloc’s viz., EU, NAFTA, and ASEAN and the present study is using it to
find out the trade bloc which has significant trade with India as per the model. Tinbergen
(1962) and Poyhonen (1963) first apply the gravity equation for the analysis of
international trade flows. At the outset, gravity model predicts why trade blocs emerge for
counties that are close with reference to developed status with respect to GDP,
preferences and technology and close geographical proximity helps (Zarzoso, 2003).
Trade blocs are mainly based on similar production function and preferences captured by
the similar GDP of the member countries i.e. similar level of GDP facilitate more trade
between the member countries and small distance between the member countries is a
bonus which further leads to increase in the trade. Trade blocs are based on the economic
arguments of trade creating and trade diverting which leads to the welfare gains for the
union or trade blocs along with these arguments they also based on the geopolitical
factors that can play an important role in the formation of regional groupings and the
formation of trade blocs. So, geography provides an encouraging factor for the integration
of the economies. As geographic distance and GDP both are core variables of the gravity
model. Gravity model predicts that trade would be high facilitated by policy to reduce
distance and tariff. Trade blocs are also about the reduction or no tariff rate between the
member countries. So, based on the somehow same preferences gravity model can
predicts the trade blocs.
Countries within trade blocs also have some advantages of greater competitiveness,
bargaining power and many other advantages with that translates into maintenance or
enhancement of their higher developed status and economic development (Doss and
Cabalu, 2000). As India is not a part of the developed blocs, therefore it would be at
unfavorable placed as the distance would be higher and the gravity-based trade
association with the developed countries would be lower. At the same time, given the
geographical distance, India recently has increasingly relied on globalization and imports
of capital and intermediate goods that can bring in similarly with respect to technology
and preferences and such increased developed status India. The assumption is: such
closeness of GDP can predict gravity model based higher trade flows with developed
countries and with the developed trade blocs as well. The other assumption that distance
is the proxy of the transportation cost along with the various other costs of doing trade
with the countries and distance has a negative relationship with the trade association.
India is doing trade with almost all major countries in the world and also has a bilateral
trade relationship with the individual countries and a focus can be on gravity model to
study India’s trade association with individual developed countries, India’s major trading
partners. In the present paper, on the other hand, an attempt has been made with the help
of gravity model to find out the significant trade with these trade blocs. This is because,
major advanced counties especially in the Europe, North America and South East Asia
operate through trade blocs and India’s closeness to trade blocs would further enhance
India and trade blocs: A gravity model analysis 219
relationship with the developed countries within the trade blocs via various trade related
agreements especially with the ASEAN members. Different counties within each trade
bloc could be specializing in different specialized tasks, products and industries; trade
blocs are base of such industrial differentiation that also underpins increased intra-firm
trade – an advanced status of trade. India’s closeness to trade blocs would inform on such
closeness to industrial differentiation – that defines advanced preferences and
technological choices.
When a country indulged in a trade bloc establishment it has many advantages which
mainly includes economies of scale, foreign direct investment, competition, market
efficiency and trade effects, The market size of India will increase and will also leads to
increase in the efficiency of the Indian firms as it increase the competition with the
joining of the trade blocs. India membership with the trade blocs will also leads to the
attendant trade expansion and barrier reductions which further induce economic growth in
India and the expansion of trade. The export of the India will also increase if India will
become the member of the trade blocs. There will be increase in the export of the labor
intensive goods and increase the employment and output with the membership of trade
blocs with India (Fukase and Martin, 2016).
Since 1991, when massive economic reforms were implemented in India this resulted an
increase in share of trade of India with the ASEAN countries as well as NAFTA
member’s countries along with some others countries as well. An increasing trend of
growth in the formation of regional integration and regional integration agreements
(RIAs) in all over the world has been observed since 1980s and each trade bloc members
has its own priorities and objectives and have different degrees of regional integration.
The emergence of international trade blocs mainly started with the creation of the
European Economic Area (EEA) in 1957 and after the enactment of the Maastricht Treaty
in 1993. In the Asia this has emerged with the ASEAN's creation in 1967. MERCOSUR,
also came into existence after the signed by major economies of the South America
(Argentina, Brazil, Paraguay, and Uruguay) in March 1991. In North America it has
emerged mainly after the free trade agreement between Canada and United States in 1988
and with the establishment of NAFTA in 1994.In the Africa in started after the formation
of Economic Community of West African States (ECOWAS) and many others trade
blocs (Sickert, 1995). With the formation of SAARC on 8th December 1985 India had
also joined the league of Trade blocs. India various regional trading arrangements
(RTAs), such as FTAs, Preferential Trade Agreements (PTAs), etc. with many countries
around the world, however how much these agreements are beneficial for India it is a
matter of concern given the competitiveness of India as compare to the partners. India has
signed multiple free trade and economic partnership agreements with some major trade
blocs especially with the south east economies like Japan, South Korea, and ASEAN
countries.
So, many Indians studies have used the gravity model to predict the trade between India
and different countries at the individual level and also with the individual trade bloc trade.
The study regarding the chronology of trade between India and EU since the seventies
has been analyzed by Bhattacharya (2005)his area of analysis consists of broad
parameters viz. Level of tariff barriers on India's exports, level of NTBs to India's
exports, revealed comparative advantage (RCA) and tariff equivalents (TEs) of India's
major export with the EU. By using the gravity model Bhattacharyya and Banerjee (2006)
try to find out the important factor in determining India's direction of trade. Bhattacharyya
and Mandal (2014) apply the gravity model at all HS 6-digit codes for trade between India
and ASEAN and found that intermediate goods will be more affected (both adversely and
favorably for India) than final goods under the scenario of a Free Trade Agreement (FTA).
So, previous studies on India mainly focus on the bilateral trade between India and
partner countries, India and EU studies on trade, India and ASEAN studies regarding the
trade in particular product code. They focus on the bilateral trade of India with Individual
trade bloc, direction of trade, and many others aspects of trade, but they wouldn’t focus
on the comparative analysis of the trade between India and developed trade blocs on the
basis of the gravity model. The present paper collaborate three main trade blocs and find
the trade pattern of India with these blocs on the basis of the gravity model as previous
studies didn’t focus on this objective. The study will attempt to examine the trade bloc out
of the three developed trade blocs which has significant trade with India as per the gravity
model is concern.
Nachane and Lakshmi (1997) also analyzed the likely consequences on India due to the
formation of two regional trading bloc’s viz. EC and the NAFTA on the basis of gravity
model and found that European community has strong trade creating effects, whereas
North America Free trade Agreement was strictly limited. As from the above studies we
can say that they attain their objectives by using the gravity model, however there is
difference in the model on the basis of the independent variable taken into consideration
except the core variables. In this present paper we took the feasible and adequate gravity
model variables along with other suitable variables for both India and developed trade
blocs as independent variable and trade value between the two as dependent variable
keeping into the concern of the multicollinearity problem in the model. The existing
studies of India regarding the bilateral trade especially who have used the gravity model
have different objectives to attain like to find the direction of trade, trade creating and
trade diverting effects, etc., however this paper have somehow different objectives to
attain which as per review are not able to achieve by the existing studies. So, in this paper
we estimate the gravity model results for India and EU, India and ASEAN and India and
NAFTA at Individual level and then compare all the coefficients of the independent
variable so that we able to find that with which trade bloc India have significant trade and
the pattern of India trade with these trade blocs. Multi-level mixed-effect model with
linear regression and ML method of estimation is utilized for the estimation of the result
in the present study. This paper does not taken into accounts all the member countries for
analysis from EU and ASEAN. It took only those members countries among the trade
blocs with which India have a significant trade at a country level on the basis of the trade
value.
As we know that there are different approaches to study trade patterns between the
countries or trade blocs. These approaches starts from Mercantilism to Smith’s Absolute
Advantage, later replaced by the more formidable theory of comparative advantage by
Ricardo, the twentieth century saw the propagation of revolutionary ideas by Hecksher
and Ohlin which proposed that based on the factor endowment and Factor intensities
affect trade between countries and explain the trade pattern. This theory has ignore
several other influences such as transport costs, economies of scale, external economies
etc., which too exert influence on the cost of production and also on the trade. The HO
theory investigates the pattern of international trade in a static sense which is not valid in
a current dynamic economic system. The theory overlooks the role played by product
differentiation in international trade. The gravity model has become known model for
empirical foreign trade analysis. The model has been widely used for flows such as
migration, foreign direct investment, and more especially to international trade flows.
Gravity model incorporated the distance variable as transport cost involved in the trade,
greater distance implies higher cost of trade and also a set of dummies incorporating
some type of institutional characteristics common to specific flows.
When we apply gravity model, it is often extended by taking in concern several other
qualitative variables such as language, tariffs, colonial history etc. Earlier model of
international trade like Hecksher and Ohlin did not incorporated it in the model.
Anderson (1979) derive the gravity equation from a model by assuming product
differentiation. He was the first person who gave a theoretical basis to the gravity
equation and provided it as one of the demand-side models. Bergstrand (1985, 1989) in
his couple of papers revealed about the gravity equations and also explored the theoretical
determination of bilateral trade on the basis of monopolistic competition models.
Helpman (1987) specify the gravity model as according to the increasing returns to scale
in the framework of differentiated product. Deardorff (1995) has proven the gravity
equation that it has been characterizes by many models in the trade and it can be justified
from the regards various standard trade theories. Eaton and Kortum (1997), and Deardorff
(1998) derived the Gravity model from a Ricardian, and Heckscher-Ohlin framework
respectively, whereas Helpman and Krugman (1985) derived it on the basis of the “New
International Trade Theory” framework. So, the essence of the earlier models of
international trade also incorporated in the gravity model under some assumption and
some conditions. Frankel, Stein and Wei (1995) used the gravity model to examine
bilateral trade patterns in the world and took land and sea routes as a measure of distance.
So this paper will analyze the variables which are included in the model viz., Trade flow,
GDP of the host country and GDP of the partner country, distance between the countries
and other gravity model variables incorporated in the model.
2. Gravity model
This present paper specifies a gravity model that is based on the Tinbergen (1962) and
Linneman (1966) bilateral model of trade. The gravity equation is a simple empirical
model for analyzing bilateral trade flows between geographical entities (Batra, 2006). The
basic form of gravity model explains that bilateral trade (Tij) is directly proportional to the
product of GDPi and GDPji.e. size of the economy and inversely related to the distance
between them.
Log(Tij)=α+βlog(GDPi×GDPj)+β2log(GDP/popi×GDP/popj)+β3log(Distij)
Apart from the basic core gravity model variables which are included in the model there
are other variables or you can say factors which can impact trade flows and thus results
the addition of dummy variables to the basic form of the model and this new model is
considered as augmented gravity model.
2.1. Augmented gravity model
In this augmented model we also include the dummy variables along with the primary
variables. The equation of the model is as follows:
Log(Tij)=α+β1log(YiYj)+β2log(PCIi,PCIj)+β3log(Dij)+β4(Borderij)+β5(Langij)+β6(RTA)+
+β6(Col))+γ1(landlocked) +γ2(Island)+ γ3(tariff)+ γ4(Exchange Rateij)
Where I and j denote countries and Tij denotes the value of bilateral trade between I and j.
Y represent the GDP and (Pop) represent the population. Distance (Dij) is the distance
between country i and country j. (PCIi) denotes the Per Capita Income of the reporting
country (PCIj) denotes the Per Capita Income of the Partner trade bloc, Borderij and
Langij, represent a dummy of border and common language respectively.
Col represents the Colonial links which is also a dummy variable represent the relation of
trade during the colonial period which is expected to ease of doing trade. Landlocked
represent the number of landlocked countries in the pair of the gravity model and (RTA)
denotes Regional trading arrangements as countries among the pairs often enter into
regional trading agreements to facilitate bilateral trade.
A typical gravity model revealed that bilateral trade flows between a pair of countries is
depend mainly on their economic size, geographical distance, populations, and some
qualitative factors also such as, common language, membership in RTAs etc. Gravity
models hypothesize that country’s production and its supply capacity has represent by the
exporting Country’s income while an importing country’s income represents the
country’s purchasing power or its absorption capacity. Various trade resistant factors such
as transportation costs and tariffs should be negatively related to trade flows.
So, the present paper deals mainly with India's merchandise trade with the developed
trade blocs which mainly include NAFTA, EU, and ASEAN member countries of the
Trade blocs over the period from 1991 to 2017.This study are based on the gravity model
on the basis of this model we will predict the trade flow of India with the developed trade
blocs. The model which we apply, the various variables which are taken into account, the
specification of the model are given in the below section.
3. Model specification and data

This paper seeks to find whether the trade flow indeed effected by the Income,
Geographical Distance in case of India and trade blocs from the time period of 1991 to
2017. In this study, the gravity model is specified and re-parameterized into a time series
and cross-sectional framework.
We have time-series data for the trade flow (Import +Export) and cross-section data for
the distance which we assume as a proxy of transportation costs and time in the trade. We
also assume that transport cost to be an increasing function of distance keeping in concern
the importance of geographical distance in bilateral trade flows.
The model takes the functional form:
logTVrpt= β0+ β1 (logGDPrt×logGDPpt)+ β2logdistance + β3logpcGDPrt+
+ β4logpcGDPpt+ β5Tariffrpt+ εrpt (1)
TFrpt (trade flows) represents total trade values (imports + exports) between country
r(Host country) and p(partner country). The GDP and per capita GDP variables are stated
in thousands of current US dollars. The variable GDPrt represents the GDP of the reporter
country and the variable GDPpt represents the GDP of the partner country, which also
implies the economic size which facilitates the trade as per the gravity model. Per capita
Income also has a positive link with the trade. With the specification of Per capita GDP
as independent variable it can explore the link between a country’s trade and its stage of
economic development. The GDP and per capita GDP coefficients of the model are
expected to have a positive sign as trade, economic size, and income has direct relation
among them. The coefficient beta on economic size (β1) i.e. β1 is the coefficient of
combined GDP i.e. the product of GDPrt and GDPpt. The distancerp variable captures the
trade distance in Km between the trading country pairs. The distance variable has a
negative relation with the trade between the two countries as per the gravity model. The
coefficient β2 represent the coefficient of distance variables which likely to be negative
due to the negative relationship between the trade and distance as detailed mention in the
above gravity model section. The variable pcGDPrt represents the per capita income of
the reporter country and the variable pcGDPpt represents the per capita income of the
partner country. Where, tariffrpt represents the tariff between the reporter country and the
partner country.
The coefficient of per capita income levels (β3and β4) i.e. β3 is the coefficient of the per
capita income of reporting country whereas β4 is the coefficient of the per capita income
of the partner country. The tariff also includes an explanatory variable due to the reason
that trade depends on the rate of tariff. Export of the reporting country depends on the
tariff applied by the partner country and export of the partner country depends on the
tariff applied by the reporting country. The tariff variable can be proxied by dummy
variables indicating the presence of preferential trading arrangements as in the basic
gravity equation. The coefficient β5 represents the effect of the tariff on the trade volume.
All the explanatory and the dependent variables are expressed in log form and hence their
coefficient interpretation is one of constant elasticity. Whereas at the end the εrpt is the
stochastic term which is log-normally distributed error term in our model. We take
distance as a proxy variable which will represent the cost of trade i.e. transport cost
between the two countries. There are several reasons to include of distance as an
explanatory variable as it can be use as proxy for transport costs. Synchronization costs
and Cultural differences also results with the increase in the distance as it can impede
trade (Batra, 2006).
The traditional approach of estimation of gravity model has the problem of significant
biases as it estimate the log-linearized equation by Ordinary Least Squares (OLS) and
the other problem of unable to use observations with zero (Silva and Tenreyro, 2006). So
the maximum likelihood method of estimation is usually used for the estimation of the
result. Maximum Likelihood (ML) is the method of estimation of the paper. Preliminary
data analysis is employed to ensure that the Multi-level mixed-effect model with linear
regression and ML method of estimation is appropriate for estimating the model. This
estimation is done for the trade blocs which include ASEAN, EU, and NAFTA. First, we
apply this regression method to estimate the results regarding the gravity model variables
between India and ASEAN. Data which we used for the estimation of the results between
India and ASEAN is based on the assumption that Singapore, Malaysia, Indonesia, and
Thailand the main members of ASEAN represent the ASEAN. The selection of these
countries is based on their size of the economy i.e. GDP as we know Indonesia,
Singapore, Malaysia, Thailand, Vietnam, Philippines and Myanmar are the major
economy of the ASEAN and also due to the unavailability of India's data with some
ASEAN members. Similar is the case with the other trade bloc which is EU whose
estimation result acquire with this similar method of estimation as used for ASEAN. So,
data which we used for the estimation of the results between India and EU are based on
the assumption that Germany, Italy, UK, France, Netherlands, Belgium and Spain which
are the main members of EU represent the EU. As we know Germany, France, UK, Italy,
Netherlands, Belgium and Spain is the major economy of the EU in terms of GDP is a
concern. In this study, we consider UK still as a member of the EU. As far as NAFTA is a
concern we regress upon the data on the variables mention in the model and estimate the
result of the gravity model.
So we first study the trading relationship between India and ASEAN i.e. the whole
scenario of the trade, Free trade agreements between the two and other important issues
and then interpret the actual result of the gravity model. The analysis has been performed
for a few major countries within the trade blocs like only seven major countries of the EU
and ASEAN and all three countries of NAFTA. The results of only these countries
namely, France, Germany, UK, Italy, Netherlands, Belgium and Spain from the EU bloc
and Indonesia, Thailand, Malaysia, Singapore, Vietnam, Philippines and Myanmar from
the ASEAN bloc and the USA, Mexico, and Canada from NAFTA are reported. So,
countries which we have taken into the account for various trade blocs their result will
represent the result with the blocs. However, countries which are not included because
data for sufficient periods is not available for some member countries and others have not
much trade with India. There are many types of gravity equations were given by various
economists to evaluate the effects of different variables like GDP, distance, etc. Gravity
models have been augmented with various quantitative and qualitative variables which
either facilitate or restrict trade. This study uses a gravity model to analyze the trade
flows between India and different trade blocs which is EU, ASEAN, and NAFTA. We
use Multi-level mixed effect technique in STATA package to obtain results of the
mentioned model in equation (1).
We collected the panel data for bilateral trade flow (exports and imports) and FDI
inflow for the selected group of countries from the period of 1991 to 2017. Given our
focus on the trade blocs, we centered our country selection around the seven major
economies of ASEAN, seven major economies of EU and all NAFTA member
countries. Data on GDP for India and the EU, ASEAN and NAFTA economies were
obtained from the World Bank Development Database through the WDI data query
(https://dvdata.worldbank.org/data-query). Data on export and import has been taken
from the Direction of Trade Statistics, IMF. The data for distance is taken from Centre
D'etudes Prospectives et D'informations Internationales. The Data on the tariff (AHS
weighted average) i.e. effectively Applied is taken from the WITS. Since the models
estimate trade flows from 1991 to 2017. Data for per capita GDP has also taken from the
World Bank database. All trade volume figures which has been used for the estimation
are in millions of US dollars. While both GDP and per capita GDP figures are stated in
PPP (constant 2011 US $) terms. Missing value of the variable is treated as zero values.
The data on trade flow (Import and Export), Distance, GDP and per capita GDP are
converted to natural log form. The data for this analysis is unbalanced and the period for
this analysis is from 1991 to 2017.
In 2017, India total value of exports is US$ 294,364 million and the total value of imports
is US$ 444,052 million. United States, United Arab Emirates, Hong Kong, China,
Singapore, Germany and UK are among the major export partners of India. US with a
total export of US$ 46,018 million (15.6%) is the top exporter of India. Whereas China,
United States, United Arab Emirates, Saudi Arabia, Switzerland, Indonesia are among the
major import partners of India. Indonesia, Thailand, Singapore, Malaysia, Philippines,
Myanmar and Vietnam are the major economies of the ASEAN in terms of economic
size. These countries also have a significant trade with India in 2017. Germany, France,
United Kingdom, Italy, Netherlands, Belgium and Spain are the major economies of the
EU in terms of GDP and these economies also have significant trade with India in 2017.
As per the Direction of trade statistics IMF (2018) India -ASEAN share of total trade
grew from 1.94 billion US$ in 1991 to 80.74 billion US$ in 2017 and India-EU total trade
grew from 11.18billion US$ in 1991 to 96.51 billion US$ in 2017.As per NAFTA is
concern India total trade value grew from 5.35 billion US$ in 1991 to 84.51 billion US$
in 2017 (Direction of trade statistics IMF, 2018). Free Trade Agreement in Goods between
ASEAN and India was signed in 2009 and enacted in 2010. ASEAN-India Agreements on
Trade in Service and Investments has came into force in 1 July 2015 which mark the
completion of the ASEAN-India Free Trade Area (AIFTA). India and the EU have enjoyed
healthy economic relations especially after the 1993 cooperation agreement signed between
the two which took their bilateral relations beyond merely trade and economic cooperation.
The first India-EU Summit in June 2000 marked has took the relationship of India and EU
at a new peak. According to the direction of trade statistics (IMF 2018), the EU accounted
for 13.3% of India's total trade in 2017. In terms of exports, the EU is the largest (16.9%)
and in terms of imports, the EU is the second-largest (9.7%) trading partner of India. EU is
India's largest regional trading partner while India is also a largest trading partner of the EU.
NAFTA is one of largest trading block in the world. India trading relation with NAFTA as
trade bloc mainly depends on its relation with the USA. Within NAFTA, the USA
continued to remain the most prominent importer of Indian products.
We estimate the model with annual data for 7 ASEAN member, 7 EU member and
NAFTA member countries for the period 1991 to 2017. They include Indonesia,
Malaysia, Singapore, Thailand, Myanmar, Philippines and Vietnam for the ASEAN and
France, Germany, UK, Italy, Belgium, Spain and Netherlands for EU while for NAFTA
all members are taken into account i.e. USA, Canada and Mexico. We estimated three
sets of regression models for each trade bloc to measure the significance of gravity model
in trade flow between India and trade blocs during the period of 1991 to 2017. The
present study reports the results of the panel data estimations of the gravity equation for
trade flows in between 1991 and 2017 in Tables 1, 2 and 3 when a trade is a dependent
variable, and Combined GDP product, Per capita Income, Distance and tariff as the
independent variable. This section provides the analytical perspective on ASEAN-India,
EU- India and NAFTA- India trade with the help of the gravity model.
4.1. India and ASEAN trade
Empirical results of the gravity model for ASEAN and India are given step by step with
their interpretations. The significant coefficient of the gravity model will strengthen the
study that trade between India and Trade blocs will reside on the gravity model. Table 1
provide the results of the gravity models estimated for India as a reporting country and
ASEAN as a partner based on equation (1). The unbalanced panel data regression model
has been estimated for the ith combination of reporting (r) country (i.e., India) with trade
bloc over the study period; i.e. for trade bloc (ASEAN). The model has been estimated
using the multilevel mixed-effect regression estimation technique. Table 1 reports the
results of the gravity model for India and ASEAN. Table 2 reports the results of the
gravity model for India and EU and lastly, Table 3 reports the results of the gravity model
for India and NAFTA.GDP product of India and ASEAN, per capita Income of reporting
country (pcGDPr), and per capita income of partner bloc (pcGDPp) are significant at 1%
level of significance. Distance and Tariff impact positively on the trade flow between
India and ASEAN, however, the effect is negligible and the value is not significant. The
per capita GDP of India and ASEAN impact positively on the trade flow between them.
Table 1. Gravity model estimates for India and ASEAN

Mixed-effects ML regression Number of obs. = 162
Wald chi2(5) = 1678.27
Log likelihood = 26.982391 Prob > chi2 = 0.0000
LogTVrp Coef. Std. Err. z P>|z| [95% Conf. Interval]
Log(GDPr*GDPp) .333*** .055 6.03 0.000 .225138 .4421212
Logdistance .003 .121 0.03 0.979 -.234487 .2408084
LogpcGDPr 1.942*** .177 10.97 0.000 1.595757 2.289759
LogpcGDPp .617*** .043 14.24 0.000 .5324432 .7023651
Tariffrp .002 .001 1.32 0.186 -.000907 .0046727
_cons -14.054 .623 -22.54 0.000 -15.276 -12.8323
Random-effects Parameters Estimate Std. Err. [95% Conf. Interval]
Var(Residual) .0419619 .0046624 .0337503 .0521715
Notes: ***Significant at 1%, **Significant at 5%.
Source: Author’s calculation.
The variable GDP has a positive impact on the trade flow between the India and ASEAN.
The variables distance and tariff have unexpected signs while the Combined GDP
product, per capita GDP of the reporting country and per capita income of partner bloc
have the expected theoretical signs during the time period. India should need to increase
its trade with ASEAN with which India has a geographical advantage in the terms of
distance. Distance variable is not the regressive factor in the case of ASEAN which
implies that the cost of transportation is not high. The combined GDP of India and
ASEAN has encouraged the trade between the two. India should need to increase its trade
with the ASEAN because the per capita income of both India and ASEAN also encourage
the trade between the two partners. Hence, given the fact, India should encourage the
trade with ASEAN keeping in mind the cost of a trade. Gravity model variables also
support the argument that India should maintain the trade with ASEAN.
4.2. India and EU trade
Table 2 provide the results of the gravity models estimated for India as a reporting
country and EU as a partner based on equation 1. Distance, per capita income of the
reporting country (India) (pcGDPr) and per capita GDP of EU are significant at 1% level
of significance. The variable tariff is significant at 5% level of significance while the
Combined GDP variable is insignificant.
Table 2. Gravity model estimates for India and EU
Wald chi2(5) = 543.87
Log likelihood = 18.635846 Prob > chi2 = 0.0000
LogTVrp Coef. Std. Err. Z P>|z| [95% Conf. Interval]
Log(GDPr*GDPp .10827 .0838139 1.29 0.196 -.0560006 .2725439
LogpcGDPr 2.0154*** .1592666 12.65 0.000 1.703316 2.327629
Logdistance -.64435*** .1314878 -4.90 0.000 -.9020642 -.3866416
LogpcGDPp -.34384*** .1126421 -3.05 0.002 -.5646178 -.1230689
Tariffrp .00297** .0014808 2.01 0.044 .0000752 .0058799
_cons -2.056006 1.744049 -1.18 0.238 -5.47428 1.362268
Var(Residual) .0480707 .004945 .0392933 .0588089
Notes: *** Significant at 1%, ** Significant at 5%.
The variable Distance and per capita GDP of the EU bloc has a negative impact on the
trade flow between the India and EU. The combined GDP product, per capita GDP of
India and tariff impact positively on the trade flow between them in which only the
Combined GDP variable is insignificant. Per capita income of the EU have the
unexpected theoretical signs while the other variables like per capita GDP of India,
distance, and are significant and have the expected theoretical signs during the period.
India need to do less trade with the EU with which India has a geographical disadvantage
in the terms of distance. Distance increases the cost of merchandise trade between India
and EU. The trade with the EU is also found to be discouraged if we consider the per
capita GDP of EU which has an adverse impact on the trade flow between India and the
EU whereas per capita income of India has encourage the trade between the two partners.
Hence trade flow between India and EU need not to be much encouraged as per the
gravity model which we apply on its selected member countries.
4.3. India and NAFTA trade
Empirical results of the gravity model for NAFTA and India are given step by step with
their interpretations. The significant coefficient of the gravity model will strengthen the
study that trade between India and Trade blocs will reside on the gravity model.
Table 3 provide the results of the gravity models estimated for India as a reporting
country and NAFTA as a partner based on equation 1. Distance, per capita income of the
reporting country (India) is significant at 1% level of significance. The variable GDP has
a negative impact on the trade flow between the India and NAFTA; however the value is
highly insignificant. The variable distance has a negative impact on trade flow between
India and i.e. with 1% increase in the distance leads to on an average 16.15% decrease in
the trade flow between India and NAFTA. Tariff variable impact negatively on the trade
flow between India and NAFTA, however, the effect is negligible and the value is also
not significant. The per capita GDP of the NAFTA and India has a positive impact on the
trade flow between India. GDP variable has unexpected signs while the other variables
like per capita GDP, distance, and tariff have the expected theoretical signs during the
period of 1991 to 2017.
Table 3.Gravity model estimates for India and NAFTA
Log likelihood = 33.110524 Wald chi2(5) = 1928.75
Prob > chi2 = 0.0000
LogTVrp Coef. Std. Err. Z P>|z| [95% Conf. Interval]
Log (GDPr*GDPp) -.0461 .343 -0.13 0.893 -.719 .626
Logdistance -16.1523*** 5.658 -2.86 0.004 -27.229 -5.074
LogpcGDPr 2.7504*** .728 3.78 0.000 1.322 4.178
LogpcGDPp .0561 .452 0.12 0.901 -.830 .942
Tariff -.0005 .001 -0.38 0.704 -.003 .002
_cons 58.27434** 30.208 1.93 0.054 -.933 117.482
Var(Residual) .0258506 .004062 .0189984 .0351742
Notes: *** Significant at 1%, **Significant at 5%.
India needs to do less trade with NAFTA because India has a geographical disadvantage
with it in terms of distance. It increases the cost of merchandise trade between India and
NAFTA. So, it will not be beneficial for India to do much trade with NAFTA keeping in
mind the cost of a trade. India trade with NAFTA is found to be discouraged if we
consider the GDP which has an adverse impact on the trade; however, the GDP has a
highly insignificant impact on the merchandise trade flow between India with NAFTA.
The per capita income has a positive impact on the trade between India and NAFTA. So
India should need to trade with the NAFTA because the per capita income of both India
and NAFTA encourage the trade between the two partners. Tariff discourages the trade
flow between the two; however, the tariff has no significant impact on the trade flow.
Hence, given the fact, India maintains the trade with NAFTA which higher GDP,
however, the gravity model does not fully support the argument.
4.5. Comparisons of gravity coefficients

Combined GDP is positively and significantly affecting the trade flow between India and
ASEAN whereas in case of EU and NAFTA it is not significant. It is evident from the
coefficients of combined GDP in the case of ASEAN that a 1% percent increase in
combined GDP of India with the ASEAN partners will increase the trade flow between
India and ASEAN by 0.33%. In the case of NAFTA and EU it has been observed with the
coefficient of 0.108 and -0.046 respectively, but the coefficients are insignificant in each
bloc. By comparing these coefficients with each other it may be concluded that the GDP
coefficient of ASEAN has largest effect on its trade flow and the coefficient of NAFTA
comes in the second rank while the EU combined GDP coefficient has the lowest effect
on the trade flow. The distance coefficient of EU and NAFTA are negatively affecting the
trade as per theoretical expectations while the coefficient of ASEAN is positively
affecting the trade not as per expectations. By observing the p-values, the coefficients of
distance for NAFTA and EU are statistically significant at 1% level of significance while
in case of ASEAN it is not significant. The distance coefficient in the case of NAFTA is
much stronger than the EU and ASEAN i.e. distance plays a major role in the case of
NAFTA.
Table 4. Comparison of gravity model coefficients for ASEAN, EU, and NAFTA
ASEAN EU NAFTA
Variable Coefficient p >ІzІ Coefficient p >ІzІ Coefficient p >ІzІ
Log(GDPr*GDPp) 0.3336*** 0.000 0.108 0.196 -0.046 0.893
Logdistance 0.00316 0.979 -0.644*** 0.000 -16.15*** 0.004
LogpcGDPr 1.9427*** 0.000 2.015*** 0.000 2.75*** 0.000
LogpcGDPp 0.617*** 0.001 -0.343*** 0.002 0.056 0.901
Tariff 0.002 0.186 0.003** 0.044 -0.0005 0.704
Notes:** Significant at 5%, *** Significant at 1%.
The variable per capita GDP of India, i.e., reporting country per capita income positively
affected the trade flow of India with each partner blocs. By comparing per capita GDP of
India coefficients with each other it may be concluded that the coefficient in case of
NAFTA has the largest effect on trade flow and the coefficient in the case of EU comes in
second rank while for ASEAN the coefficient has the lowest effect on the trade flow. It
can be said that as the per capita income of India increase this leads to the increase in the
trade flow with these trade blocs. The variable per capita GDP of partner bloc also
positively affected the trade flow of India with partner blocs for ASEAN and NAFTA
whereas in case of EU it impact negatively. By comparing these coefficients with each
other it may be concluded that the coefficient of ASEAN has the largest effect on trade
flow and the coefficient of EU comes in the second rank while NAFTA coefficient has
the lowest effect on the trade flow and it is not significant for NAFTA. The tariff imposed
by India on these trade blocs is positively affecting the trade flow between India and these
trade blocs except NAFTA with which tariff variable shows a negative impact on the
trade flow. However the coefficient is not significant for the ASEAN and NAFTA.
6. Conclusion and policy implications

The paper aims to analyze the trade bloc which has significant trade with India as per the
gravity model is a concern. In this context, we tested for the effectiveness of the gravity
model variables for the trade flow between the India and trade blocs which includes
ASEAN, EU and NAFTA. This paper indicates that the pattern of international trade of
India has changed from the EU members to the ASEAN member's countries i.e. India is
shifting more towards the ASEAN economies. It shows the effects of different economic
variables on the trade flow of India with the trade blocs. These variables are GDP of
reporting country with the partner member countries as a whole, a distance of reporting
country with a partner, per capita income of the reporting country, per capita income of
the partner bloc member countries as a whole and tariff (effectively applied tariff) by
reporting country. Applying the gravity model of bilateral trade flows between India and
trade blocs, we found that these variables have a significant impact on the trade flow.
Theoretically, GDP, per capita Income must positively affect the trade flow, while tariff
and distance are supposed to affect the trade flow negatively. Nevertheless, variables such
as combined GDP product, per capita income of reporting and trading partner have the
expected signs and are statistically significant for ASEAN except the variable Distance
which shows an unexpected sign for the ASEAN and is also not significant. Distance, per
capita income of the reporting country variables shows an expected signs and is
statistically significant for the EU and NAFTA except the variable per capita income of
the partner bloc which shows an unexpected sign for EU bloc. Tariff variable shows an
unexpected sign and significant for the EU blocs. So, the coefficients of the distance
variable in the case of EU and NAFTA support the gravity model but not in the case of
ASEAN, however in the case of ASEAN it is not significant. So distance has significant
impact on the Trade flow between India and Trade blocs.
We can draw some policy recommendations on the basis of our analysis regards the trade
relationship. In this scenario, as our empirical results shows the size and distance as main
determinants of trade. Policymakers need to understand the inter-relationship between
India and trade blocs which include ASEAN, EU, and NAFTA so that jointly
strengthening policies can be developed. India should need to increase the depth of trade
with ASEAN and also need to increase its per capita income so that it further leads to
increase in trade through various agreements and polices.
References
Anderson, J.E., 1979. A Theoretical Foundation for the Gravity Equation. American Economic
Review, Vol. 69, pp. 106-116.
ASEAN Statistical Year book, ASEANSTAT, 2017. Retrieved from <https://www.
aseanstats.org/publication/asyb-2017>
Batra, A., 2006. India's Global Trade Potential: The Gravity Model Approach. Global Economic
Review: Perspectives on East Asian Economies and Industries, Vol. 35(3), pp. 327-361.
DOI: 10.1080/12265080600888090.
Bergstrand, H.J., 1985. The Gravity Equation in International Trade: Some Microeconomic
Foundations and Empirical Evidence. The Review of Economics and Statistics, Vol. 67,
No. 3, pp. 474-481.
Bergstrand, J.H., 1989. The Generalised Gravity Equation, Monopolistic Competition, and the
Factor-Proportions Theory in International Trade. Review of Economics and Statistics, 71(1),
pp. 143-53.
Bhattacharya, K.S., 2005. India and the European Union: Trade and Non-Tariff Barriers, Delhi.
Bhattacharyya, R. and Banerjee, T., 2006. Does the Gravity Model Explain India’s Direction of
Trade? A Panel Data Approach, W.P. No.2006-09-01,Indian Institute of Management,
Ahmedabad.
Bhattacharyya, R. and Mandal, A., 2014. The India – ASEAN Free Trade Agreement: How Will
Indian Industries Be Affected? SSRN Electronic Journal. DOI: 10.2139/ssrn.2446414.
Deardorff, A.V., 1995. Determinants of Bilateral Trade: Does Gravity Work in a Neoclassical
World? Working Paper Series, No. 5377, National Bureau of Economic Research.
Deardorff, A.V., 1998. Determinants of Bilateral Trade: Does Gravity Work in a Neoclassical
World? Review of Agricultural Economics, 28(3), pp.408-415.
Department of Commerce and Industry, GoI, 2016. Retrieved from <https://commerce.gov.in/>
Department of Promotion and Industry, Government of India, 2016. Retrieved from
<https://dipp.gov.in/>
Direction of Trade Statistics, IMF, 2018. Retrieved from <https://data.imf.org/>
Doss, N. and Cabalu, H., 2000. When east meets south: economic gains from India-APEC trade.
Applied Economics, 32(11), 1405-1418. DOI: 10.1080/00036840050151485.
Eaton, J. and Kortum, S., 1997. Technology and Bilateral Trade, Working Paper, No. 6253,
Cambridge, MA: National Bureau of Economic Research.
European Commission, 2016. Retrieved from <https://ec.europa.eu/trade/policy/countries-and-
regions/countries/india/>
Eurostat, EU, 2018. Retrieved from <https://ec.europa.eu/eurostat>
Frankel, J., Stein, E. and Wei, S., 1995. Trading Blocs and the Americas: The natural, the
unnatural, and the super- natural. Journal of Development Economics, 47(1), pp. 61-95.
Fukase, E. and Martin, W., 2016. The Economic Potential of an India-US Free Trade Agreement.
Journal of Economic Integration, Vol. 31, No. 4, pp. 774-816.
Helpman, E., 1987. Imperfect Competition and International Trade: Evidence from Fourteen
Industrial Countries. Journal of Japanese and International Economics, Vol. 1, pp. 62-81.
Helpman, E. and Krugman, P., 1985. Market Structure and Foreign Trade: Increasing Returns,
Imperfect Competition and the International Economy, Cambridge, MIT Press.
Krugman, P., 1980. Scale Economies, Product Differentiation, and the Pattern of Trade. American
Economic Review, Vol. 70, pp. 950-59.
Linnemann, H., 1966. An Econometric Study of International Trade Flows, North Holland,
Amsterdam.
Ministry of External Affairs of India, GOI, 2018. Retrieved from <https://www.mea.gov.in/>
Ministry of External Finance, GOI, 2018. Retrieved from <https://mea.gov.in/Portal/Foreign
Relation/Singapore_24_12_2018.pdf>
Nachane D.M. and Lakshmi R., 1997. Regional Trading Blocs and Their Implications for the
Indian Economy. In: Gupta S.D., Choudhry N.K. (eds.) Globalization, Growth and
Sustainability. Recent Economic Thought Series, Vol. 58. Springer, Boston, MA.
Poyhonen, P., 1963. A tentative model for the volume of trade between countries. Weltwirt Arch,
Vol. 90, pp. 93-99.
Sickert, A., 1995. The Emergence of International Trading Blocs: Trends and Consequences,
Paradigms, Vol. 9, No. l.
Silva, S.J. and Tenreyro, S., 2006. The Log of Gravity, Review of Economics and Statistics,
Vol. 88(4), pp. 641-658.
Tinbergen, J., 1962. Shaping the World Economy. The Twentieth Century Fund, New York.
UNCTAD, 2017. Handbook of Statistics. Retrieved from <https://unctad.org/en/
PublicationsLibrary/tdstat42_en.pdf>
Zarzoso, M.I., 2003. Gravity Model: An Application to Trade between Regional Blocs. Atlantic
Econ. J., Vol. 31 (2), pp. 174-187.
Predicting the volatility in stock return of emerging economy:

An empirical approach
Aastha KHERA
Kurukshetra University, Kurukshetra, India
aasthakhera12@gmail.com
Dr. Miklesh Prasad YADAV
Amity University, Noida, India
miklesh12@gmail.com
Abstract. Investors become jittery when they do not earn return on their hard earned money. In the
same time, they want to make their investment in safe place rather than losing it. For better return,
they also want to estimate the volatility in stock market. The basic purpose of the present study is to
forecast the volatility in stock return of emerging economies. For the same, the adjusted daily
closing price of eleven countries is considered for five years. Generalized Autoregressive
Conditional Heteroscedasticity (GARCH) has been applied to predict the stock return of these
countries. The different orders of GARCH have been applied in predicting the volatility. It is found
that the volatility of every stock return can be forecasted.
Keywords: emerging countries, stock return, GARCH.
JEL Classification: C22, G12, L11.

234 Aastha Khera, Dr. Miklesh Prasad Yadav
Introduction
Financial markets are being characterized by their stochastic nature that involves the
precarious movement of prices of securities in both directions. Financial market volatility
is vital barometer of dynamic inconstancy of the stock markets (Raja and Selvam, 2011).
Variability in the returns that is caused due to changes in prices of the securities is specified
as volatility which is measured by the standard deviation or variance. It is defined as extent
of fluctuations in the prices of variable i.e. stock prices that could move up and down. Ups
and down are the normal phenomenon in the markets but if their price oscillation is sharp
and steep, then it becomes complicated for the investors to forecast and plan their strategies.
Investors invest in the market for the gain of risk premium. High fluctuations in the market
augments the ambiguity in the market regarding the future returns and hence increases the
risk. Inconsistent performance of the markets creates further dubious condition for returns
creating difficulty for investors to predict the market. Excessive fluctuations makes the
financial planning a challenging task. Researchers and academicians are delving deep into
this topic to gain useful insights for modeling and forecasting patterns of volatility.
Emerging markets are witnessing high volatility as compared to developed countries,
making it an interesting areas for the researchers (Santis and Imrohoroglu, 1997). This
could be due to information asymmetry in these types of the markets. Emerging markets
are becoming hot new option for investors to enhance their wealth. Emerging markets are
being characterized by high volatility (Cohen, 2001). High volatility could lead to huge
gains as well as huge losses. These financial markets in emerging countries are harboring
high attention from economists, practitioners and researchers from all over the world.
Emerging markets are on the path to growth and development. Economic growth is on rise
and is driven by uncertain upswings and downswings.
In such highly volatile markets, it becomes hard for the companies to raise funds in the
capital markets. Investors lose their confidence in these types of situations. Excessive
fluctuations cause losses to risk averse investors creating unfavorable impact (Premaratne
and Balasubramanyam, 2011). These are accumulated to further aggravate the situation that
could lead to crashes in the market. Accurate estimation of volatility is crucial for risk
management practices. Investors are interested in studying the volatility so that they can
apply this knowledge in maximizing returns and minimizing risk. They can analyze the
impacts of volatility on their investment assets. Fluctuations in the market could be
attributed to various macroeconomic factors like inflation, exchange rate, interest rate and
certain institutional factors like market performance (Islam, 2013). Investment philosophy
of investors also plays an important role in maintaining the stability in the market. Retail
investors usually gets involved in rat race creating disturbances in the market. High
variations in the market are caused by lack of proper skills and knowledge of investors.
Estimating volatility is the key factor to be analyzed in taking the financial decisions.
Financial strategies are framed after due investigation of financial market volatility.
Appropriate investment and leverage decision are taken after considering the volatility
variable. Normal volatility is favorable for active and healthy stock market (Lin, 2018).
Generally securities markets face serious issues regarding variability. Modeling volatility
is also required in derivative pricing, risk management and portfolio management. So
Predicting the volatility in stock return of emerging economy: An empirical approach 235
appropriate model has to be selected that can estimate and forecast volatility of financial time
series accurately as financial markets are stochastic in nature. Stock prices are not constant
and same is the case with volatility. Some periods show calmness in the markets and some
are characterized by high movements. In statistical terms, it is defined as heteroskedastic
property. It implies that volatility changes with the time horizon. Various time series models
have been developed to capture this characteristic in modeling volatility of the series.
Volatility could be measured in two ways one is historical volatility that is calculated on
the past data and the other is implied volatility that is derived from market price of traded
security. There are different volatility models i.e. traditional estimators, extreme value
estimators and conditional models. Traditional econometric models assumes volatility to
be constant in describing the volatility of stock market return. These are the oldest ones
that measures unconditional volatility through simple standard deviation. It is calculated as
close to close estimated volatility. With the knowledge and awareness that in the financial
markets volatility is time varying and showing serial correlations, these models lost their
importance and conditional models came into picture. Extensive studies have been
conducted to show that these traditional estimators are incapable of explaining the financial
data precisely. Other one is extreme value estimator which calculates volatility on the basis
of extreme values (high and low prices) in the stock market. Then came the conditional
models that take into consideration time-varying nature of the volatility.
Usual characteristics being found in financial time series are volatility clustering and
leptokurtosis (Mandelbrot, 1963). There has been exceptional progress in advancement of the
econometric models which have been able to apprehend distinctive characteristics in the
financial time series. One of the model that has the ability to model conditional volatility or
variance after incorporating these characteristics is Generalized Autoregressive Conditional
Heteroskedasticity (GARCH) which was proposed by Bollerslev (1986). It has ability to
annotate the common characteristics being present in financial series like leptokurtosis,
volatility clustering and asymmetric or leverage effects. It is an extension of Autoregressive
conditional heteroscasdicity (ARCH) that was given by Engle in 1982. It provides a more
flexible approach to capture dynamic structures of conditional variance (Chou, 1988). ARCH
model was initially developed to study the temporal nature of stock returns volatility. It
proved to be effective model however in order to capture dynamic order of conditional
variance higher order ARCH is required. GARCH cured this problem and completed the
requirement as it make its base on infinite ARCH specification by which parameters are
reduced from infinity to two. GARCH models explained today's volatility as the function of
the constant term plus alpha times the yesterday squared residuals and beta times the
yesterday variance. There are two parameters that is one ARCH term and one GARCH term.
Further extensions of ARCH and GARCH models have also been developed to grab the
other distinctive effects like leverage and asymmetric effect. EGARCH was propounded
by Nelson in 1991, TGARCH was developed by Zakoinin in 1994, GJR-GARCH by
Glosten, Jganathan and Runkle in 1993. GARCH models are appropriate for high
frequency financial data like stock returns that witness conditional variance. GARCH is
used frequently to model the volatility/variance which depends upon past residual squared
observation and past variance of the series.
In this paper, focus is on forecasting and modeling volatility of emerging countries using
GARCH(1,1) technique. This would enable the investors to make sound decisions. They
would be familiarized with movement of stock indices among the countries. The paper is
organized as follows Section 2 discusses the related works section 3 explains GARCH
modeling. Section 4 presents analysis and discussion. Section 5 concludes the paper with
summary and future directions.
Review of literature
GARCH models have been frequently used by many researchers to model the volatility.
Various empirical studies have been conducted for estimating and forecasting financial
volatility.
Chou (1988) investigated the persistence of volatility and risk premium scenario in the US
Market from 1962 to 1985 with the help of GARCH model. US markets have witnessed
changing equity premiums with a point of risk estimate aversion of 4:5. Uncertainty was
there in 1974, that causes the markets to plunge causing increase in volatility by 26%.
Lamourex and Lastrapes (1990) in their study contributed to the empirical evidence in
usage of ARCH to capture the heteroskedasticity in stock returns. Conclusions found to be
valid in the cases of analyzed stocks and this study further paved the way for the
employment of ARCH and GACH models in studying behavior of asset prices.
Pandey (2005) analyzed the estimation and forecasting ability of three different traditional
estimators, four extreme value estimators, and two conditional volatility models. Realized
volatility estimates were used as benchmarks for the purpose of comparison. High
frequency data was analyzed from the Jan 1999 to Dec 2001. Three time periods were used
i.e. 1 day, 5 day and 1 month. Indexes have shown the characteristics of high frequency
financial time series. GARCH(1,1), GARCH(1,1)-Rolling and EGARCH models were
applied under conditional models If bias terms are talked about conditional volatility
models have performed better than extreme value estimators. In terms of predictive power
of models extreme value estimators have performed better than conditional models in
providing 5 day and 1 month ahead volatility forecasts. Frimpng and Abaiye (2006)
employed random walk, GARCH(1,1), EGARCH(1,1) and TAGRCH(1,1) to forecast and
model conditional variance in Ghana stock exchange. Index studied was Databank Stock
Index (DSI) for 10 years. AIC and LL criteria was used to measure and compare the
performance. DSI index rejected the hypothesis of random walk hypothesis. One can
predict and forecast the market. Assuming the normal distribution of the series,
GARCH(1,1) model has outperformed the other models.
Floros (2008) selected two Middle East indices Egypt (CMA General Index) and Israeli
(TASE-100 index) and studied their patterns of volatility and market risk. Daily data for
the time period 1997-2007 was taken and symmetric GARCH models were applied and
they were successful in capturing dynamics of the series. Insignificant results came out
regarding relationship between expected returns and risk. GARCH-M model concluded
that higher expected risk does not necessarily leads to high returns. In the series of
estimating volatilities using symmetric models, Islam (2013) in their paper applied two
symmetric models GARCH(1,1) and GARCH-M(1,1). Sample was of three Asian markets
namely Kuala Lumpur Composite Index (KLCI) of Malaysia, Jakarta Stock Exchange
Composite Index (JKSE) of Indonesia and Straits Times Index (STI) of Singapore. Daily
observations were analyzed for the time period of five years. GARCH-M model tested the
hypothesis between expected returns and risk. Results stated that GARCH models have
superior ability in capturing volatility clustering and leptokurtosis. Coefficient of risk
premium was significant in case of Indonesian markets as it tends to be more volatile.
Alberg, Shalit and Yosef (2008) employed asymmetric models to forecast two major Tel-
Aviv Stock Exchange (TASE) indices: TA100 and TA25. Different density functions are
used to compare the forecasting performance of both symmetric as well non-symmetric
models namely GARCH, EGARCH, GJR and APARCH models together. E GARCH models
have outperformed the other three models while using students t distribution. EGARCH has
better risk management ability for this exchange returns as per the study. Eminike (2010)
analysed the behavior of stock returns of Nigerian stock exchange using GARCH(1,1) and
GJR-GARCH(1,1). General error distribution method was used in calculating the volatility.
Monthly data of indices were taken from 1999 to 2008. Evidences of volatility clustering, fat
tailed distribution and leverage effects were present in the series for the study. GARCH
models were successful in capturing the conditional volatility of the series.
In modeling and forecasting of Malaysian stock market proxy KLCI was done by Angabini
and Wasiuzzaman (2012). In their study, an attempt was made to study the impact of
financial crisis of 2007-2008 on volatility of Kuala Lampur Stock Exchange (KLCI) in
Malaysian stock market. Two study period were taken one including the crisis other one
excluding the crisis. AR(4) was finally selected as fit model for conditional mean and for
conditional variance GARCH(1,1) EGARCH(1,1) and GJR-GARCH was finalized. Impact
of global financial crisis could be stated with significant increase in volatility and presence
of leverage effect is detected with a small drop in the persistency. Another study regarding
Malaysian stock market was by Shamiri and Zaidi in 2009 applied GARCH, EGARCH and
NGARCH models. These Models were applied and their performance was compared using
different error distributions. Indexes depicted the characteristics of volatility clustering and
leptokurtosis. Choice of error distribution has played an important role that the choice of
models during the measurement of performance.
Lim and Sek (2013) in their paper employed both symmetric and asymmetric to analyze
the Malaysian stock market volatility. Three time periods are used for analysis -pre-crisis
1997, during crisis and post-crisis 1997. Mean squared error, Root mean squared error and
the mean absolute percentage error are the error statistical tools to compare the performance
of GARCH models. Crisis period favored asymmetric models and symmetric models were
preferred during pre and post crisis period. Impact of exogenous variables named exchange
rate and crude oil prices was significant only in pre and post crisis period.
Oberholzer and Venter (2015) investigated the impact of financial crisis of 2007-09 on the
behavior of volatility. The study period was divided into three sub periods. Daily volatility
of 5 indices of Johennserberg Stock exchange were studied and their changes were
analyzed using various extensions. GARCH(1,1), GJR-GARCH(1,1) and EGARCH(1,1)
models were the selected ones and their performance were analysed with the help of various
error statistical tools. GJR-GARCH(1,1) came out to be the best model for all indices
except FTSE top 40 index. With the inclusion of impact of the financial crisis, E-GARCH
became the best model.
Lin (2018) analysed forecasting performance of various extensions of GARCH models.
SSE Composite index was the subject of the study. Significant time-varying and volatility
clustering features were prevalent in the volatility of Shanghai composite index. Good
estimation was observed in both symmetric as well as asymmetric models. Best performing
model was found to be EGARCH model. Awalludin, Ulfa and Soro (2018) modeled the
stock price return volatility in Indonesian stock market. GARCH(1,1) was employed to
estimate the volatility and it showed the evidences of volatility clustering in few stocks.
GARCH(1,1) a linear model captures volatility clustering successfully. Maximum
Likelihood estimation method was used to estimate the parameters. Volatility series was
fitted using natural cubic spline function.
Data and research methodology

Data consist of daily 5 year observations of 11 emerging countries. Their major stock
indexes are taken and values are obtained from yahoo.finance.com from 1st January 2014
to 31st December 2018. R studio software is employed to estimate and forecast conditional
stock returns.
Studies of Mandebrot (1963), Fama (1965) and Black (1976) focused on the distinctive
features of stock returns that is volatility clustering, leptokurtosis and asymmetric or
leverage effects. Conditional heteroskedastic models incorporate these time-varying
characteristics of second moments of distribution notably. They are not like traditional and
extreme value estimators whose assumption is regarding unconditional variance. In 1982
Engle proposed Autoregressive conditional heteroscedasticity model that processes lagged
disturbances if forecasted variance is on the basis of variance of the error terms known as
ARCH effect. ARCH(1) model states Time varying variance as the function of b0 (constant
term) and b1e2t-1 (squared error of previous term one lagged period.).Under this Volatility
is modeled by relating the conditional variance of error terms to the linear combination of
the squared disturbances in the recent past.
Further research concluded that in order to apportion the potent nature of conditional
variance, higher order ARCH is needed. So this condition was fulfilled and Generalized
ARCH (GARCH) was introduced by Bollerslev in 1986. Both ARCH and GARCH models
became prominent and were employed intensively in various empirical studies. If
forecasted variance is on the basis of variance of the error terms and past value of the terms
it is known as GARCH(1,1).It helps to model the volatility or variance which is dependent
upon past residual squared observations and past variance of the series. There are both
symmetric as well as non-symmetric GARCH models. Linear and non linear extensions
have been developed by another academicians. In this paper GARCH(1,1) model is used
to estimate and forecast the volatility of stock returns of emerging countries. Conditional
mean equation for the stock return is defined as constant term plus residuals.
rt = θ + ut
The basic and widespread GARCH(1,1) model can be presented as:
σ2t= α0 + α1u2t−1+ α2σ2t−1
σ2t – current day variance or volatility;
α0 – omega or constant;
u2t−1 – previous squared error residuals, known as ARCH term;
σ2t−1 – yesterday variance or volatility, known as GARCH term.
All parameters α0, α1 and α2 are non-negative. In order to ensure weakly stationarity of the
GARCH process, α1 + α2 < 1 this condition should be fulfilled. α1 depicts the short run
persistency of the shocks and α2 indicates the long run persistency of the shocks.
GARCH(1,1) is the linear model and models the symmetric in the series.
Two basic conditions for applying GARCH model:
1. Volatility Clustering: Large changes are succeeded by large changes, small changes are
being followed by small changes. It is the behaviour of the assets in which prices tends
to cluster in the group.
2. ARCH effect: The presence of the autocorrelation in squared residuals of the series is
known as Arch effect. It is known as heteroskedasticity also known as non-constant
variance.
Framework for GARCH(1,1) model:
After the fulfillment of above two conditions, GARCH(1,1) is applied in the stock returns.
Volatility clustering is checked by the time series plot of the returns series. Arch LM test
is applied to detect the presence of autocorrelation. Then GARCH(1,1) is applied and
coefficients are generated along with their p-values.
Data analysis
Descriptive statistics
On the basis of availability of data, 11 emerging countries are identified. Closing prices of
these indexes are taken into consideration for 5 years. Table 1 provides descriptive statistics
depicting mean, minimum, maximum, standard deviation, skewness and kurtosis of indices
of these selected emerging countries. Figure 1 depicts the change in log of daily closing
prices of indices of selected emerging eleven countries. Figure 2 represents change in
returns of daily closing prices that is volatility clustering.
Table 1. Descriptive statistics of selected indices of these emerging countries
Country Index N Min Max Mean Std dev Skew Kurt
symbol
India Sensex 1223 -0.061197 0.033236 0.000437 0.008399 -0.491976 6.014365
Brazil iBovespa 1236 -0.092107 0.063887 0.000451 0.014734 -0.076022 4.854293
(Bvsp)
Korea KS11 1221 -0.045411 0.034728 3.02E-05 0.007504 -0.489692 5.582291
Indonesia JKSE 1211 -0.040884 0.044514 0.000296 0.008963 -0.434803 5.682779
Argentina MERV 1210 -0.106400 0.090651 0.001437 0.021956 -0.271899 5.266557
Mexico MXX 1251 -0.059884 0.035251 -1.05E-05 0.008546 -0.529666 7.233409
Country Index N Min Max Mean Std dev Skew Kurt

symbol
China 000001.SS 1219 -0.088732 0.056036 0.000137 0.015044 -1.224012 10.00030
Russia IMOEX.ME 1221 -0.114189 0.093670 0.000391 0.011725 -0.593853 16.31261
Philippines PSEi 1214 -0.069391 0.035762 0.000182 0.009678 -0.345535 5.776407
Turkey XU100 791 -0.102597 0.088424 0.000668 0.013700 -0.254488 10.64279
Chile Ipsa 1237 -0.060343 0.066691 0.000262 0.007480 0.228051 11.92841
Figure 1. Time plot of change in log of closing price indices of emerging countries
INDIA SENSEX BRAZIL BVSP
KOREA KS11 INDONESIA JKSE
ARGENTINA MERV MEXICO MXX
RUSSIA IMOEX.ME
CHINA 000001.SS
PHIILLIPHINES PSEi TURKEY XU100
CHILE IPSA
Figure 2. Time plot of change in stock returns of closing price indices of emerging countries (volatility
clustering)
INDIA SENSEX BRAZIL BVSP
KOREA KS11 INDONESIA JKSE

ARGENTINA MERV MEXICO MXX
CHINA 000001.SS RUSSIA IMOEX.ME
PHIILLIPHINES PSEi TURKEY XU100
CHILE IPSA
Application of GARCH model

Every return series is found stationary. Augment dicker Fuller test was applied for detecting
the presence of unit root. Graphs of return series are showing volatility clustering as shown
in Figure 2. All financial time series are highly leptokurtic having fat tails. Table 2 presents
results of Arch Test. p vale is less than 5% thus rejecting null hypothesis of having ARCH
effect. Every indices is showing presence of Arch effect which is prerequisite for applying
GARCH models. Table 3 is showing output of GARCH estimation.
Table 2. ARCH LM Test

Country Returns Data Chi-squared df p-value
India Sensex 43.755 12 1.681e-05
Brazil iBovespa 42.831 12 2.411e-05
(Bvsp)
Korea KS11 63.545 12 5.058e-09
Indonesia JKSE 86.164 12 2.718e-13
Argentina MERV 122.96 12 2.2e-16
Mexico MXX 172.35 12 2.2e-16
China 000001.SS 203.8 12 2.2e-16
Russia IMOEX.ME 34.641 12 0.000534
Philippines PSEi 39.3 12 9.392e-05
Turkey XU100 102.6 12 2.2e-16
Chile Ipsa 40.214 12 6.626e-05
Table 3. GARCH(1,1) output

Country Returns data w A1 A2 A1+ A2
India Sensex 0.000001 0.052015 0.932767 0.984782
(0.002035) (0.000000)
Brazil iBovespa 0.000007 0.053707 0.914147 0.967854
(Bvsp) (0.000000) (0.000000)
Korea KS11 0.000004 0.077164 0.858457 0.935621
(0.000000) (0.000000)
Indonesia JKSE 0.000001 0.047484 0.938231 0.985715
(0.062221) (0.000000)
Argentina MERV 0.000027 0.179525 0.777671 0.957516
(0.000000) (0.000000)
Mexico MXX 0.000004 0.112321 0.839955 0.952276
(0.000345) (0.000000)
China 000001.SS 0.000001 0.069240 0.929758 0.998998
(0.000026) (0.000000)
Russia IMOEX.ME 0.000018 0.156461 0.727802 0.884263
(0.000313) (0.000000)
Philippines PSEi 0.000004 0.070236 0.889333 0.959569
(0.000000) (0.000000)
Turkey XU100 0.000001 0.023788 0.972067 0.995855
(0.000000) (0.000000)
Chile Ipsa 0.000010 0.221234 0.615812 0.837046
(0.000000) (0.000000)
Conclusions
The basic objective of the present study is to examine and forecast the volatility of the stock
exchanges of emerging countries. It is found that the volatility of every stock return can be
forecasted. Both ARCH and GARCH terms are significant in all the cases. Their sum of
the coefficients are large enough to denote the persistence of the volatility. The overall
persistency of shock is largest in China’s stock return and lowest in case of Chile’s stock
exchange as their parameters sum is highest and lowest respectively. The sum of α1 & α2 is
less than one (α1 + α2 < 1) implies the mean reverting GARCH model. Comparing the result
of short run and long run shock persistency, it is found that long run shock is more persistent
than short run as their α2 is larger than α1.
References
Alberg, D., Shalit, H., and Yosef, R., 2008. Estimating stock market volatility using asymmetric
GARCH models. Applied Financial Economics, 18(15), pp. 1201-1208.
Angabini, A., and Wasiuzzaman, S., 1997. GARCH models and the financial crisis-A study of the
Malaysian stock market. Studies, 2007 (2008).
Awalludin, S.A., Ulfah, S., and Soro, S., 2018. Modeling the stock price returns volatility using
GARCH(1, 1) in some Indonesia stock prices. Journal of Physics: Conference Series, Vol.
948, No. 1, January, p. 012068. IOP Publishing.
Bollerslev, T., 1986. Generalized autoregressive conditional heteroskedasticity, Journal of
Econometrics, 31, pp. 307-27.
Chou, R.Y., 1988. Volatility persistence and stock valuations: Some empirical evidence using
GARCH. Journal of Applied Econometrics, 3(4), pp. 279-294.
Cohen, S.I., 2001. Stock performance of emerging markets. The Developing Economies, 39(2),
pp. 168-188.
Emenike, K.O., 2010. Modelling stock returns volatility in Nigeria using GARCH models.
Floros, C., 2008. Modeling Volatility using GARCH Models. Evidence from Egypt and Israel.
Middle Eastern Finance and Economics, 2, pp. 30-41.
Frimpong, J.M., and Oteng-Abayie, E.F., 2006. Modelling and forecasting volatility of returns on
the Ghana stock exchange using GARCH models.
Islam, M.A., and Mahkota, B.I., 2013. Estimating volatility of stock index returns by using
symmetric GARCH models. Middle-East Journal of Scientific Research, 18(7), pp. 991-999.
Lamoureux, C.G., and Lastrapes, W.D., 1990. Heteroskedasticity in stock return data: Volume
versus GARCH effects. The journal of finance, 45(1), pp. 221-229.
Lim, C.M., and Sek, S.K., 2013. Comparing the performances of GARCH-type models in capturing
the stock market volatility in Malaysia. Procedia Economics and Finance, 5, pp. 478-487.
Lin, Z., 2018. Modelling and forecasting the stock market volatility of SSE Composite Index using
GARCH models. Future Generation Computer Systems, 79, pp. 960-972.
Mandelbrot, B., 1963. The variation of certain speculative prices, Journal of Business, 36,
pp. 394-419.
Oberholzer, N., and Venter, P., 2015. Univariate GARCH models applied to the JSE/FTSE stock
indices. Procedia Economics and Finance, 24, pp. 491-500.
Pandey, A., 2005. Volatility models and their performance in Indian capital markets. Vikalpa, 30(2),
pp. 27-46.
Premaratne, G., and Balasubramanyan, L., 2003. Stock Market Volatility: Examining North
America, Europe and Asia. National University of Singapore, Economics Working Paper.
Raja, M., and Selvam, M., 2011. Measuring the time varying volatility of futures and options. The
International Journal of Applied Economics and Finance, 5(1), pp. 18-29.
De Santis, G., and Imrohoroglu, S., 1997. Stock Returns and Volatility in Emerging Financial
Markets, Journal of International Money and Finance, 16, pp. 561-579.
Shin, J., 2005. Stock returns and volatility in emerging stock markets. International Journal of
Business and economics, 4(1), p. 31.
Portfolio optimization with VaR approach:

A comparative analysis for Japan, London, New York and India
Parul BHATIA
Apeejay School of Management, Dwarka, New Delhi
parul.bhatia84@gmail.com
Priya GUPTA
Lal Bahadur Shastri Institute of Management, Dwarka, New Delhi
Priyagupta.1704@gmail.com
Abstract. Risk managers use various types of techniques to estimate different kinds of risk and ways
to minimize its impact. VaR which stands for Value at Risk is one of those techniques. Various new
methods for calculation of VaR have been developed. In this study, four techniques of VaR estimations
have been employed: i) Historical Simulation; ii) Variance Covariance Approach; iii) Monte Carlo
Simulation, and iv) AR-GARCH method. The purpose of this study is to compare the different VaR
estimation methods and draw conclusions based on the Back- Testing methods.
As per the analysis, historical method proved to be the best method for estimating value at risk. This
method is widely preferred by risk managers and practitioners in the banking sector. Though the
portfolios used in the study was diversified and contained stocks from different sectors, still the
historical simulation method came out to be on the top as it was accepted for all the four portfolios.
This method does have some limitations as the patterns generated from the past data may not hold
true all the time.
Keywords: Value-at-Risk (VaR); simulation model; variance-covariance matrix; Monte Carlo simulation;
GARCH approach.
JEL Classification: D53, E17, E37, F65, G14, G17.

246 Parul Bhatia, Priya Gupta
Introduction
Uncertainties in the world of finance are inevitable. If an investor is willing to invest in
various assets, then he can never avoid facing risk. Magnitude of risk varies across different
domains of assets class. The primary role of risk managers is to hedge risk to avoid
maximum loss and gain as much profit as possible. Risk managers use various types of
techniques to estimate different kinds of risk and ways to minimize its impact. VaR which
stands for value at risk is one of those techniques. It is one of the popular methods for
estimating the market risk. “VaR is a tool which gives a possibility to estimate the total risk
of the portfolio”, defined by John C. Hull. Total risk of a portfolio is categorized into two
components: i) specific risk and ii) systematic risk. Specific risk refers to a kind of risk
which can be diversified and thus be reduced or removed whereas Systematic risk is a type
of risk which cannot be diversified and it depends on various factors like economic
conditions of the country, inflation, political structure and their influence and on the stock
market.
Figure 1. Portfolio risk proportion
In the past decade, VaR has become one of the most common techniques for risk
management. In 1994, JP Morgan introduced Risk Metrics, a newly designed risk measured
system which included various measures for risk over 300 financial instruments across 14
countries with proper defined methodologies of value at risk. The BCBS which stands for
Basle Committee on Banking Supervision amended the Basel Capital Accord and later on
obliged the banks which aligned with the Basel Committed to reserve capital based on the
measurement of their market risk using the Value at Risk (Basle Committee, 1996).
According to these norms, banks now must maintain three times of their VaR
measurements in order to cover itself from potential risks. Other than that, banks can
introduce their own internal methods and should be subjective to approval by Bank for
International Settlements. While VaR is a standard tool for measuring risk, it has undergone
various refinements. Obtaining accurate value of VaR has become the most important
challenge for researchers and risk managers. Various new methods for calculation of VaR
have been developed. In this project we will be covering the following techniques of VaR
estimations: i) historical simulation; ii) Variance Covariance Approach; iii) Monte Carlo
Simulation, and iv) AR-GARCH method.
Portfolio optimization with VaR approach: A comparative analysis for Japan, London, New York and India 247
Value at Risk
“Value at Risk is such a loss in market value of a portfolio that probability it will occur or
even will be exceeded in a given period of time is equal to the predefined tolerance level”
(Jajuga, 2007). For example, a VaR value at 95% confidence interval of 10% risk depicts
an expected loss of at least 10% one of every 20 days on average.
Mathematically, it can be represented as:
Where:
V – value of the portfolio at the end;
Vo – value of the portfolio at the beginning;
VaR – Value at Risk;
α – significance level.
Thus, VaR can be interpreted as the worst-case scenario loss over a specified time period.
Another way of representing value at risk mathematically would be:
Where:
rt – rate of return on the portfolio;
F-1 r, t (α) – quantile of loss distribution related to the probability of 1–𝛼.
There are various approaches for measuring value at risk. This study has included four
different methods: i) Historical Simulation; ii) Variance Covariance Approach; iii) Monte
Carlo Simulation, and iv) AR-GARCH method.
Historical simulation
Historical Simulation is a VaR technique which uses the historical data of returns in order
to obtain the value at risk. A Nth percentile is chosen as a target value which eventually
determines the value at risk. For example, we want to determine the VaR by using the
historical worst 1% of daily return as the targeted value. All the daily returns of the past
data of the specified time period will be ranked from the lowest to highest and then the 1st
percentile will be taken as the value at risk value. This value will then be multiplied by the
current value of the asset to get the 1-day VaR estimate.
In case of a portfolio, daily returns for each stock is calculated and then multiplied by the
respective weights assigned to it. Total return is calculated by adding the returns of all the
stock. Further the daily returns thus calculated are ranked from lowest to highest and then
the nth percentile is calculated and the required VaR is calculated. Also, it is again multiplied
by the current value of the portfolio to get the 1-day VaR estimate.
The historical simulation technique is the most used VaR technique practiced by the risk
managers. One of the best advantages of this technique is that it is easy to use and
implement. It does not require any complex mathematical calculations or any software or
platform to calculate. Also, the data need not be normally distributed. It can be of any
distribution. It only uses the past data of returns and the past patterns irrespective of its
distribution and provide estimates based on that.
However, the assumption of constant pattern of distribution may not be true or it may be
true for short time horizon that the newer data explains the behavior of returns better than
older data does. On the other hand, the method requires significant level of sample size
because the Nth percentile being selected must come from large discrete data to make sense.
Hendricks figured out that greater the sample size, lesser will be the variance in VaR
estimation. He also found imprecise VaR estimate when the sampling period is too short.
The major drawback of this method is the difficulty in selecting the proper sample size for
its calculations.
Variance covariance approach
Variance Covariance method also known as the Linear VaR or Delta Normal VaR is a
simple and most used method for estimating value at risk. It uses some application of the
Markowitz model given by Harry Markowitz, which uses the correlation factor among
different assets. With the help of the historical data, daily returns can be calculated which
further help in calculating the different parameters like mean, standard deviation and the
correlation among them. It assumes that the data follows any specific distribution which is
usually normally distributed. This helps in defining volatility in terms of the standard
deviation of the portfolio. Since the data is normally distributed, this implies that it is
symmetrical thus, it has a skewness 0 and a kurtosis 3. If we want to find position of a
random variable (X) in a normal distribution, we use standard value of variable Z (Z-score).
Every variable can be transformed to standard variable with formula:
Where:
Z = (X-µ)/σ;
µ – mean;
σ – standard deviation.
Thus, VaR is calculated as the measure of the standard deviation.
Figure 2. Normal distribution depiction
The above figure shows normal distribution with value of x -2,326 σ. Probability for losses
greater than x is shown as an area under the normal curve, left of x. This area is 1% of
the total area under the curve, so there is a confidence of 99% that losses will not exceed
-2,326 σ. During VaR calculation, the downward price changes or the price changes that
exceeds some multiple of standard deviation.
VaR for the entire portfolio can be calculated by using the formula:
Where:
Z – standard value;
V – volatility or the standard deviation of the portfolio;
P – portfolio position value.
Usually it is calculated by using the matrix formula:
Where:
V – row vector of VaRs for each position;
C – matrix of correlations.
Monte Carlo simulation
The Monte Carlo simulation is a process which helps in generating random price paths to
determine the approximate movements of future prices of the financial assets. In the
application of VaR estimation, by Monte Carlo simulation, we can generate the probability
distribution for rates of return of portfolios. In order to calculate a 1-day VaR for a portfolio,
Monte Carlo simulation follows below procedures:
 Determine the current market value of the portfolio.
 Generate random samples of rates of return for all underlying assets of the portfolio
from multivariate normal distribution.
 Apply the samples of returns for each asset and obtain their sampling values from current
market values.
 Compose the portfolio value based on the sampling values of assets and find the rate of
change from current market value of the portfolio.
 Repeat above steps many times to generate the probability distribution of rates of return
of the portfolio.
Now the VaR of the portfolio is directly gathered from the distribution of simulated
portfolio values. One way to implement step 2 is to estimate ST based on below equation:
Where:
ℇ – random sample from a standard normal distribution.
A common way to generate the ℇ‘s is from following function:
Where:
Ri – independent random numbers between zero and one.
In the world of the financial modelling, Monte Carlo simulation has proven to be a powerful
tool due to its flexible approach. This approach can be used to generate random scenarios
or scenarios by using historical data. One of the approach follows the normal distribution
pattern for estimating the VaR using Monte Carlo simulation. This approach has undergone
criticism as well as appreciation. Some practitioners believed that this approach is better
than the historical simulation and produced better results while incorporating fat tails.
The biggest drawback of this method is that it generates iterations randomly which may
rupture the results or could lead to biasedness in the results. The values obtained may not
always align with the market. Also, compared to other approaches, the marginal cost for
this method is higher than its benefits.
AR-GARCH method
VaR estimation using GARCH models is one of the common methods for estimating daily
volatility. It doesn't mull over the normality of returns obtained. “GARCH model is a model
that estimates the conditional variance of an asset’s returns as a function of past returns
and conditional variances.”
Parameters of a GARCH(m, n) models for a portfolio can be calculated using the following
equations:
The second equation has the following constraints associated with it:
Where:
ω, αi, βj – parameters of the model;
ht – conditional variance of the returns of the portfolio;
ℇt – white noise.
In accordance with the above model, the conditional variance is given by the following
equations:
Error term in the conditional variance forecast is given by:
And thus, the final equation for estimating the value at risk is:
Literature review
VaR is the most used tool for estimating market risk by the risk managers. It is a method
which is used by the managers and practitioners daily to determine the possible worst-case
scenario of the loss that they can incur (Stefaniak, 2018). Investors need to choose the
method in accordance with the properties and characteristics exhibited by their portfolio
(Corkalo, 2011) VAR methods are being commonly used by different banks, pension funds
and mutual funds. Their weakness was examined by using five backtesting techniques
which are test based on the time until first failure, test based on failure rate, test based on
expected value, test based on autocorrelation and test based on mean absolute percentage
error. Although GARCH approach was the best alternative for the VaR estimating
technique. Results for different price indices were different using all the methods (Sinha
and Chamu, 2000).
Back testing of the independence of first and higher orders were implemented for the 4
mentioned models above, for a forecast period of 1 and 10 days. The commitment depends
on a more thorough criteria than those utilized in the writing for approving VaR models, as
we performed backtesting for infringement autonomy of higher requests on conjecture time
of 10 days. The final products indicated that just GARCH models were giving the
satisfactory outcomes (Maluf and Asano, 2019). A new approach was used in this study
which is the extreme value theory (EVT) and the kernel estimator technique. The data was
taken for 13 years and the end results stated that the VaR estimates that were generated
through the extreme value theory and kernel estimator techniques were far better than the
traditional approaches of the value at risk (YiHou, 2000).
The mean VAR efficient frontier has been tested for its predictive abilities and has been
found to be a good tool for determining value at risk for investors with which portfolio
decisions may be made (Gaivoronski and Pflug, 2005). VAR may be helpful in predicting
the behaviour of portfolios on the basis of their past observations. However, it has been
suggested that to improve the ability of the model minimum VAR constraints may be
adopted (Charpentier and Oulidi, 2009). VAR adjusted Sharpe ratio was proposed in the
study in comparison to the traditional Sharpe ratio. It was found that worse-case scenario
adjustment provided better results from VAR-Sharpe ratio than the later (Deng et al., 2013).
Optimization techniques have been used in three phases with simple solver to theoretical
convergence. It has been found that the phase wise comparison of portfolios improvised
the analytics for portfolio development and revision (Lim et al., 2010). A movement from
traditional Markowitz model and Sharpe ratio requires introducing constraints on the
various portfolios. The study tried to assimilate the results from various traditional models
and linear programming models for portfolio optimization (Ogryczak et al., 2017). The
employment of column generating algorithm allows for better results from Copula based
conditional value at risk models. A deeper analysis may be done with the help of these
models for study of pessimistic scenarios (Krzemienowski and Szymczyk, 2016).
Normal distribution and historical simulations have been used to analyse the bond
portfolios. Findings have suggested that these distributions provide reliable results to
predict VAR. However, this technique has been found to be weak for showing results with
constraints imposed on portfolios and in real life hidden risk may always be present
(Winker and Maringer, 2007). Diversification strategies have been tested with various
approaches like Markowitz, conditional VAR and back testing during the global financial
crisis. It has been discovered that the simple diversification strategies have been generating
the best estimates among others (Allen et al., 2016).
Linear programming models have started gaining popularity in estimating value at risk for
various investment portfolios. The capability to introduce various constraints while
appraising the risk profile of the portfolios allows for more comparisons (Mansini et al.,
2007). The theory of Markowitz model has been studied with few violations and it has been
found that a simple optimization technique may help individual investors to earn superior
returns as compared to value-weighting approach (Dye and Groth, 2000).
Methodology
Sample size
The sample size considered for this study are four portfolios containing four stocks each
from different sectors. They are selected based on their market capitalization. Stocks with
the highest market capitalization are chosen for the study. Portfolios are diversified with
reference to their geographical locations in order to increase the independence of the results
and the influence of systematic risk on the results.
Data
The data taken is 10 years’ time series data from 2010-19 with varied observations
representing the stock markets of Japan- 2447 (Nikkei225), London- 2522 (FTSE100),
New York- 2519 (DJIA) and India- 2479 (Nifty50).
Components of the portfolio
 Nifty50 (Reliance Industries, Tata Consultancy Services (TCS), Bharti Airtel, Sun
Pharmaceuticals Ltd.)
 Nikkei225 (Toyota Motor Corporation, Mitsubishi UFJ Financial Group, Docomo; Soft
Bank Group)
 DJIA (Microsoft, Exxon, Apple, Johnson and Johnson)
 FTSE100 (Royal Bank of Scotland, British American Tobacco, BP, Unilever)
All the stocks in each portfolio is assigned a weightage of 25%. VaR is estimated at a 95%
confidence interval.
Techniques used
After the data collection comparison of different VaR techniques is done. The techniques
used are as follows:
 Historical Simulation.
 Variance-Covariance Approach.
 Monte Carlo Simulation.
 AR-GARCH approach.
For calculating the returns of the portfolio, log returns is used for calculating the returns
generated by each stock in a portfolio and then they are multiplied by their weights assigned
to determine the total returns of the portfolio.
Log returns is calculated by using the formula:
Rt = ln(Pt /Pt-1)
Where:
Rt – log return of the stock;
Pt – price of the stock at present day;
Pt-1 – price of the stock the previous day.
Lastly, these tests were backed by the backtesting methods in order to validate the results.
Back testing
After a VaR model is developed, it is important to check how reliable the results are and
what is the accuracy of the results obtained. The statistical method for estimating VaR
accurately is called backtesting. The aim of backtesting is to determine whether the amount
of losses calculated by VaR is correct. For this purpose, various tests are implemented.
These tests check whether the frequency of exceptions, during the prescribed time interval,
are in accordance with the confidence level chosen. The commonly used tests are:
 POF Test: It is one of the most common used backtesting methods used by the
practitioners. It measures how many exceptions occur in the model i.e. when the actual
return is less than the estimated VaR value. It is very simple to calculate as it does not
require many variables. The only information required for this are i) number of
observations, ii) number of exceptions and iii) the confidence level. The null hypothesis
for the test is:
The rationale behind this is figure out whether the observed failure rate is in sync with the
actual failure rate. It works based on the likelihood ratio test.
Null hypothesis for this is that the model is correct. If the value of the LRPOF is greater than
the critical value, then the null hypothesis is rejected, and the model is inaccurate.
 Loss function: This method of backtesting does not take into consideration any
hypothesis, instead, it is based on determining the magnitude of loss that can occur if any
violation occurs. The loss function could be stated as:
This score increases with the magnitude of loss. For conducting the backtest method, a
benchmark loss is taken and when the sample loss is greater than the benchmark loss then
the model is rejected.
Data analysis
After the data collection, their properties and characteristics are examined using descriptive
statistics which is shown in Table 1 below:
Table 1. Comparative Analysis of the returns of the portfolios
Indices Nikkei225 FTSE100 DJIA Nifty50
Mean 0.0002 0.0003 0.0005 0.0004
Minimum -0.1750 -0.0529 -0.0471 -0.0595
Maximum 0.0727 0.5983 0.0530 0.0378
Count 2447 2522 2515 2479
Figure 3. Returns distribution of Nikkei225 portfolio
The returns generated by the portfolio shows that initially the frequency is close to zero
and then it increases with the high returns generated and then again it closes to zero
following a bell shape curve.
Figure 4. Returns distribution of FTSE100 portfolio
and then it increases with the high returns generated and then again it closes to zero for
high number of returns following a bell shape curve.
Figure 5. Returns distribution of DJIA portfolio
Figure 6. Returns distribution of Nifty50 portfolio
Empirical results and findings

After the preliminary analysis is done, following results are generated out of testing the
VaR Models:
Table 2. VaR calculations for Nikkei225
Nikkei225
Model Variance Covariance Historical Simulation Monte Carlo AR-GARCH
Confidence Interval 95% 95% 95% 95%
VaR -2.15% -1.85% -0.13% -1.89%
Note: Own compilation done on MS-Excel.
The results from the above table indicates the calculations of the value at risk using
different methods. From Table 2, for the Nikkei225 index the variance covariance
approach gives the highest VaR estimate of -2.15% whereas the results from historical
and AR-GARCH methods are almost identical. Monte Carlo simulation has produced the
lowest VaR estimate for this portfolio.
Table 3. VaR calculations for DJIA
DJIA
VaR -1.42% -1.29% -0.09% -1.27%
From Table 3, for the DJIA index the variance covariance approach gives the highest VaR
estimate of -1.42% whereas the results from historical and AR-GARCH methods are almost
similar. Monte Carlo simulation has produced the lowest VaR estimate for this portfolio.
Table 4. VaR calculations for FTSE100

FTSE100
VaR -2.77% -1.74% -0.13% -2.00%
From Table 4, for FTSE100 index the variance covariance approach gives the highest VaR
estimate of -2.77% whereas the results from historical simulation is -1.74% and AR-
GARCH methods is -2.00%. Still the results produced by these two methods almost lie
near to each other. Monte Carlo simulation has produced the lowest VaR estimate for this
portfolio.
Table 5. VaR calculations for Nifty50
Nifty50
VaR -1.78% -1.69% -0.11% -1.65%
From Table 5, for Nifty50 index all the approaches except the Monte Carlo simulation is
similar. Monte Carlo gives the lowest VaR estimate.
POF-Test results
The abbreviations used in this tests are as follows:
T – total number of observations;
x – number of exceptions occurred;
CI – confidence interval, in this case, 95% confidence interval is used;
LR – Likelihood ratio following the chi-square distribution for n degree of freedoms.
As mentioned earlier if the exceptions occurrence percentage is less than or equal to the
significance level, i.e. 1-CI, then the model is estimating the results accurately. Also, the
LR ratio should be less than the critical value, which is 3.84 in our case, for the model to
be accurate and accepted.
Type I error refers to rejecting a correct model and Type 2 refers to accepting an incorrect
model. Suppose if the number of exceptions is selected as 110 for rejecting a model, then by
using the binomial distribution, the probability of committing a type 1 error would be 2%.
Table 6. POF backtest results for Nikkei225
Nikkei225
T x T-x x/T CI LR Inference
Historical 2447 124 2323 5% 95% 0.0233 Accept
Variance-Covariance 2447 85 2362 3% 95% 3.3374 Accept
AR-GARCH 2447 2340 107 96% 95% 12.6588 Reject
Monte Carlo 2447 1044 1403 43% 95% 5.2479 Reject
Figure 7. Type I error of Nikkei225
Nikkei225 Type I
4%
3%
2%
1%
0%
110 112 114 116 118 120 122 124 126 128 130 132 134 136 138 140 142 144 146 148 150
From Table 6, only historical and variance-covariance approach models were accepted for
Nikkei225 portfolio. Assuming the exception for rejecting the model to be at 125, there is
a 4% probability of committing a type I error for this portfolio.
Table 7. POF backtest results for DJIA
DJIA
Variance-Covariance 2515 1130 1385 45% 95% 4.1987 Reject
AR-GARCH 2515 104 2411 4% 95% 3.1972 Accept
Figure 8. Type I error of DJIA
DJIA Type I
4%
3%
2%
1%
0%
110 112 114 116 118 120 122 124 126 128 130 132 134 136 138 140 142 144 146 148 150
From Table 7, only historical and AR-GARCH approach models were accepted for DJIA
portfolio. Assuming the exception for rejecting the model to be at 125, there is a 4%
probability of committing a type I error for this portfolio.
Table 8. POF back test results for FTSE100
FTSE100
AR-GARCH 2522 67 2455 3% 95% 3.4907 Accept
Figure 9. Type I error of FTSE100
FTSE100 Type I
4%
3%
2%
1%
0%
110 112 114 116 118 120 122 124 126 128 130 132 134 136 138 140 142 144 146 148 150
From Table 8, only historical and AR-GARCH approach models were accepted for
FTSE100 portfolio. Assuming the exception for rejecting the model to be at 125, there is a
4% probability of committing a type I error for this portfolio.
Table 9. POF back test results for Nifty50
Nifty50
T X T-x x/T CI LR Inference
AR-GARCH 2479 112 2367 5% 95% 1.2516 Accept
Figure 10. Type I error of Nifty50
Nifty50 Type I
4%
3%
2%
1%
0%
110 112 114 116 118 120 122 124 126 128 130 132 134 136 138 140 142 144 146 148 150
From Table 9, only historical and AR-GARCH approach models were accepted for Nifty50
portfolio. Assuming the exception for rejecting the model to be at 125, there is a 4%
probability of committing a type I error for this portfolio.
Loss function results
As mentioned, loss function considers the magnitude. For this the average sample loss is
calculated and a benchmark loss figure is considered for its calculation. If the average
sample loss is greater than the benchmark loss, then the model is rejected.
In this case the benchmark was taken as the maximum return produced on a single day for
the period of 10 years. Reason being that if the investor has made a big return in a period
of 10 years say, then he will not be willing to lose that amount. So, if the VaR estimate is
giving a loss greater than his maximum profit, the investor will not consider the
implications of such a model and will reject it.
Table 10. Loss Function backtest results for Nikkei225
Nikkei225
Sample Loss Benchmark Loss Inference
Historical 5.03% 7.27% Accept
Variance-Covariance 3.49% 7.27% Accept
AR-GARCH 4.37% 7.27% Accept
Monte Carlo 44.37% 7.27% Reject
Table 11. Loss Function backtest results for DJIA
DJIA
Table 12. Loss Function backtest results for FTSE100
FTSE100
Sample loss Benchmark Loss Inference
Table 13. Loss Function backtest results for Nifty50
Nifty50
Historical 5% 3.78% Reject
Variance-Covariance 4.20% 3.78% Reject
AR-GARCH 4.52% 3.78% Reject
From Tables 11, 12 and 13, except for the Monte Carlo simulation method, all the other
models are accepted for Nikkei225, DJIA and FTSE100 portfolio. From Table 13 all the
methods are rejected by the loss function backtesting technique.
Conclusion
The purpose of this study was to compare the different VaR estimation methods and draw
conclusions based on the backtesting methods. As per the analysis, historical method
proved to be the best method for estimating value at risk. This method is widely preferred
by risk managers and practitioners in the banking sector. Though the portfolios used in the
study was diversified and contained stocks from different sectors, still the historical
simulation method came out to be on the top as it was accepted for all the 4 portfolios. This
method does have some limitations as the patterns generated from the past data may not
hold true all the time.
Another alternative that prompted out of this study was the AR-GARCH approach of
estimating the volatility. This method was accepted for 3 portfolios as well and could be
the second alternative for the VaR estimation as per our study. Variance Covariance
approach was also accepted in a few cases, but the overall implication of this method does
not provide accurate results. Monte Carlo proved to be the most ineffective method in our
study and was rejected for all the portfolios.
To conclude, one cannot determine a standard method for VaR estimation. It depends upon
the properties and the characteristics possessed by the portfolio. Also, for different size of
portfolios different results can be obtained using different methods. Risk managers should
choose their method wisely after attaining proper information about the portfolio and the
market.
Limitations and scope for further study

VAR does not measure worst case loss. 95% percent VaR means that in 5% of cases the
loss is expected to be greater than the VAR amount. Value at Risk does not say anything
about the size of losses within this 5% of cases and by no means does it say anything about
the maximum possible loss. There are many alternative techniques available now like
expected shortfall and CoVaR, which provides the magnitude of the losses and thus
preferred more by the investors and managers to determine the risk. Different VaR models
are appropriate for different portfolios depending upon their properties and characteristics
of distribution of returns. This study was limited to only four portfolios and the results are
limited to those only. The study only restricted to high-capitalization stocks only, scenarios
for mid and small capitalization are not considered in this study. Only traditional
approaches are used in this study for calculating VaR, new emerging methodologies like
transformation-based approach and extreme value theory are not included in this study.
These considerations may be used for studying the VAR strategies in portfolio
diversification further with more indices and models.
References
Allen, D.E., McAleer, M., Powell, R.J. and Singh, A.K., 2016. Down-side risk metrics as portfolio
diversification strategies across the global financial crisis. Journal of Risk and Financial
Management, 9(2) doi:http://dx.doi.org/10.3390/jrfm9020006
Charpentier, A. and Oulidi, A., 2009. Estimating allocations for value-at-risk portfolio optimization.
Mathematical Methods of Operations Research, 69(3), pp. 395-410.
doi:http://dx.doi.org/10.1007/s00186-008-0244-7
Corkalo, S., 2011. Comparison of Value at Risk Approaches on a Stock Portfolio. Croatian
Operational Research Review, 2.
Deng, G., Dulaney, T., Mccann, C. and Wang, O., 2013. Robust portfolio optimization with value-
at-risk-adjusted sharpe ratios. Journal of Asset Management, 14(5), pp. 293-305.
doi:http://dx.doi.org/10.1057/jam.2013.21
Dye, R.T. and Groth, J.C., 2000. Value weighting and simple optimization of portfolios: An
empirical examination. Managerial Finance, 26(6), pp. 23-35. doi:http://dx.doi.org/10.1108/
03074350010766729
Gaivoronski, A.A. and Pflug, G., 2005. Value-at-risk in portfolio optimization: Properties and
computational approach. The Journal of Risk, 7(2), pp. 1-31. Retrieved from
<https://search.proquest.com/docview/197274394?accountid=187781>>
Glosten, L., 1993. On the Relationship between the Expected Value and the Volatility for the normal
Excess Return on Stocks. Journal of Finance, Vol. 48, pp. 1779-1801.
Hendricks, D., 2019. Evaluation of Value-at-Risk Models Using Historical Data. [online]
Newyorkfed.org.
Krzemienowski, A. and Szymczyk, S., 2016. Portfolio optimization with a copula-based extension
of conditional value-at-risk. Annals of Operations Research, 237(1-2), pp. 219-236.
doi:http://dx.doi.org/10.1007/s10479-014-1625-3
Lim, C., Sherali, H.D. and Uryasev, S., 2010. Portfolio optimization by minimizing conditional
value-at-risk via nondifferentiable optimization. Computational Optimization and
Applications, 46(3), pp. 391-415. doi:http://dx.doi.org/10.1007/s10589-008-9196-3
Maluf, L. and Asano, J., 2019. Comparison of VaR Models to the Brazilian Stock Market under the
Hypothesis of Serial Independence in Higher Orders: Are GARCH Models Really
Indispensable? Brazilian Business Review, 16(6), pp. 626-645.
Mansini, R., Ogryczak, W. and Speranza, M.G., 2007. Conditional value at risk and related linear
programming models for portfolio optimization. Annals of Operations Research, 152(1),
pp. 227-256. doi:http://dx.doi.org/10.1007/s10479-006-0142-4
Ogryczak, W., Przyłuski, M. and Śliwiński, T., 2017. Efficient optimization of the reward-risk ratio
with polyhedral risk measures. Mathematical Methods of Operations Research, 86(3),
pp. 625-653. doi:http://dx.doi.org/10.1007/s00186-017-0613-1
Sinha, T. and Chamu, F., 2000. Comparing Different Methods of Calculating Value at Risk. SSRN
Electronic Journal.
Stefaniak, R., 2018. Review of Value at Risk Estimation Methods. Prace Naukowe Uniwersytetu
Ekonomicznego We Wrocławiu, 519, pp. 173-183.
Winker, P. and Maringer, D., 2007. The hidden risks of optimizing bond portfolios under VaR. The
Journal of Risk, 9(4), pp. 1-19. Retrieved from <https://search.proquest.com/docview/
197276791?accountid=187781>
Yi Hou Huang, A., 2000. A Comparison of Value at Risk Approaches and a New Method with
Extreme Value Theory and Kernel Estimator.
Volume XXVII (2020), No. 4(625), Winter, pp. 263- 274
Income-dependent impacts of financial development

and human capital on economic growth.
A non-stationary panel analysis
Mohsen MOHAGHEGH
Ohio State University, USA
mohaghegh.1@osu.edu
A.S. VALIPOUR
Ohio State University, USA
setayeshvalipour.1@osu.edu
Abstract. We investigate the interdependency of financial development and human capital on

economic development in a cross-country setting. We show that the impacts exhibit interesting
variations across different income levels. In fact, human capital is the most important element in
fostering economic growth in high-income countries, but its importance falls with the country’s
income level. Financial development, however, is relatively more important for mid-income
countries but does not seem to have a significant impact on high-income nations. We find that low-
income countries do not benefit as much from either one of these factors.
Keywords: economic growth, economic development, human capital, financial development, panel data,
non-stationary panels.
JEL Classification: O11, I25, G28, C33.

264 Mohsen Mohaghegh, A.S. Valipour
Introduction
This paper investigates how the impacts of financial development and human capital on
economic growth vary with the national income. Although the effects of both factors are
extensively examined in the literature, whether they have a disproportionate impact on
countries that are in different stages of development is not well understood. This issue the
main question we tackle in our paper. To answer this question, we set a panel of 194
countries between 1965 and 2017. We divide these countries to three groups of high-
income, mid-income, and low-income based on their national income level data. We use
non-stationary panel data techniques to estimate the relative importance of these factors in
each group. In particular, we use FMOLS and DOLS estimators to compare how financial
development and human capital, in the long run, affect economic growth in different
countries.
Romer (1990) is among the first to emphasize the importance of human capital by
introducing R&D activities in his model. Since R&D activities are mostly carried out by
educated workers, a rise in human capital increases R&D activities, thereby directly
improving economic growth. Since then, several theories of endogenous growth have
explored the role of human capital in enhancing output growth (e.g. Aghion and Howitt,
1992, 1998; Acemoglu, 1996). To further characterize its importance, Barro (1991) shows
that a country’s GDP growth rate is positively related to its initial human capital level; a
finding that is similar to what Benhabib and Spiegel (1994) report. Also, Mankiw et al.
(1992) augment Solow’s growth model by incorporating human capital in the production
function and show that this change helps explain a significant amount of cross-country
differences in income per capita.
Building on these theories, we empirically examine whether the positive effects of human
capital vary across countries with different income levels. Our results show that human
capital is the most important element in fostering economic growth in high-income
countries. We find that the relative importance of human capital for economic growth
noticeably falls with countries’ national income level. Low-income countries gain the least
from an increase in their human capital level. This finding is robust across both FMOLS
and DOLS estimations.
The observed decline in the effectiveness of human capital can be related to complexities
of the relationship between human capital and output growth. For instance, human capital
is believed to enhance economic growth through innovation and technological
improvement. However, empirical investigations suggest that this impact depends on the
technology frontier in the country (See, among others, Vandenbussche et al., 2006; and
Ang et al., 2011). An important finding of our paper is that an increase in the level of human
capital creates a larger boost in high-income countries than it does in low-income countries.
This is perhaps because low-income countries suffer from institutional barriers such as lack
of effective governance, skill mismatch, or corruption which diminish the effectiveness of
human capital in fostering growth.
In addition to human capital, the importance of financial development has been central to
the findings of many theoretical and empirical studies. Financial institutions improve the
Income-dependent impacts of financial development and human capital on economic growth 265
allocation of capital in the economy by mitigating asymmetric information issue between

borrowers and lenders. Fama (1985) argues bank loans have especially high return
suggesting banks are superior in monitoring borrowers and in screening out bad ones, a
function that cannot be achieved without these financial institutions. James (1987) presents
further evidence on bank loan’s uniqueness by looking at firm’s stock market reaction to
announcement of receiving a bank loan and shows that unlike other methods of borrowing,
the announcement has a positive impact on firm’s value. Financial intermediaries also
reduce the risk a lender is prone to by disconnecting the direct lender-borrower relationship,
sitting between the two, and undertaking risk sharing practices.
Our estimates show that financial development is most effective in mid-income countries.
That financial development is not effective in developed countries pinpoints the friction-
reducing nature of financial development and suggests that high-income countries have
already undertaken a well-functioning financial market that mitigates the possible frictions
in the real economy so that there in not much room to provide further financial services.
On the other hand, the significance of financial development in mid-income countries
implies that in these countries there are still non-negligible financial frictions that prevent
the economy from performing at its full capacity and removing them would accelerate
economic growth.
Low-income countries do not benefit from financial development as much as mid-income
countries. This could be related to the distribution of capital in the economy. How
efficiently the financial system allocates capital in the economy has enormous effects on
technological innovation and growth. Financial development in low-income countries does
not lead to economic growth as it does in mid-income countries. However, perhaps lower
efficiency and accessibility of financial services can be blamed for its mild effect in these
countries. It is worth mentioning that, as it is customary in the literature, our financial proxy
measures the depth of the financial system.
A well-functioning financial system improves the economy’s performance by directing the
capital to the hands of the most productive firms and individuals (Levine and Zervos, 1998).
Our findings suggest that if the financial system is already well-developed- as it is the case
in high-income countries- or if institutional barriers are so high that prevent an efficient
allocation of resources among those seeking funds- as it is the case in low-income
countries- then economic policies that focus on financial development may not have a
strong impact on output growth.
It should be noted that several studies have investigated the interactions of financial
development and human capital as it relates to economic growth. For example, it has been
argued that a developed financial system leads to an increase in investment in human capital
(Sehrawat and Giri, 2017; Kilic and Ozcan, 2018). This augments the effects of both factors
on economic growth. Such arguments led to a number of empirical studies both at the
national and international level. Kargbo et al. (2016) show that financial development and
human capital amplify the effect of each other on economic growth in Sierra Leone. Also,
in a cross-country analysis, Das et al. (2014) show that the reliance of economic growth on
financial development lessens as human capital improves.
The empirical results on how financial development and human capital interact with one
another, and how this interaction affects economic growth have been mixed. Some
conclude that these factors are complements (Hakeem, 2010) while other argue that they
are substitutes (Kendall, 2012). We contribute to this literature by examining the relative
importance of financial development and human capital conditional on national income
level. We show that the importance of financial development on economic growth is not
monotonic with respect to human capital. In fact, its impact peaks for mid-income countries
with a coefficient higher than that of both high-income and low-income countries. The
dependence of the interaction between financial development and human capital on
national income level could explain previous mixed findings in the literature. For instance,
Hassan et al. (2011) study the role of financial development on economic growth in
different geographic regions. Consistent with our results and contradicting Das et al.
(2014), they argue that financial development is more effective in mid-income countries
than low-income countries because of lack of other important determinants of growth in
low-income countries. Moreover, Acikgoz and Ali (2019) consider MENA countries and
argue their economic growth is mostly due to physical capital accumulation rather than
human capital accumulation. As MENA countries are mostly categorized as mid-income
and low-income countries, their results are consistent with our results that human capital is
less effective for less developed countries.
The remainder of this paper is organized as follows; section two describes our data. Section
three introduces our econometric model, and reports panel unit root tests and panel co-
integration tests. Section four discusses our estimation results, and section five concludes.
2. Data
To measure economic growth, we use GDP per capita (constant 2005 USD) of each country
as reported by the World Development Indicators (WDI) provided by World Bank and
available online. The level of financial development is proxied by the ratio of total liquid
liabilities to the aggregate output. Liquid liabilities (Liquidity), also known as broad
money, are the sum of currency and deposits in the central bank (M0), plus transferable
deposits and electronic currency (M1), plus time and savings deposits, foreign currency
transferable deposits, certificates of deposit, and securities repurchase agreements (M2),
plus travelers checks, foreign currency time deposits, commercial paper, and shares of
mutual funds or market funds held by residents. International Financial Statistics (IFS) and
International Monetary Fund (IMF) report this ratio.
Human Capital Index (HCI) is a hybrid national level index, which is calculated by Penn
World Tables (PWT). This index is constructed based on the average years of schooling
and an assumed rate of return to education, based on Mincer equation estimates around the
world. Also, control variables in our statistical model are the size of government, and
inflation. The ratio of government expenditures to the GDP, and the consumer price index
(2010 = 100) (CPI) measure these economic factors, respectively.
Table 1. Summary statistics

Output Liquidity HCI Gov. Size CPI
mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d.
High-income 10.01 0.73 4.09 0.68 2.81 0.50 0.16 0.05 3.36 2.95
Mid-income 7.81 0.84 3.44 0.65 1.94 0.55 0.21 0.10 2.66 3.34
Low-income 6.25 0.51 2.83 0.99 1.36 0.37 0.19 0.11 2.10 5.31
Table 1 reports mean and standard deviation of all six variables in our econometric model.
Our panel data covers 164 countries between 1965 and 2017. These countries are divided
to three different categories: “High-income” countries, “Mid-income” countries, and
“Low-income” countries. For this categorization, we follow the world bank’s classification
based on GNI threshold. Our sample consists of 41 high-income, 102 mid-income, and 21
low-income countries.
There are some notable trends in the data. The stock of both liquid liabilities and human
capital steadily rise with the national income. Advanced economies, on average, have
smaller governments while the largest governments are in mid-income countries. In
addition, though price levels seem to fall in national income, their standard deviation
significantly rises, which is a sign of higher volatility in nominal variables in lower income
countries suggesting less-effective stabilizing policies or less-effective governance.
3. Model and empirical analysis

The main objective of this paper is to compare the relative importance of financial
development and human capital accumulation for economic growth. The general effect of
these two factors has been already established in the literature. However, in this paper we
investigate whether the magnitude of such effects varies with the national income level. In
particular, as far as GDP growth is concerned, we are interested in examining if human
capital is as important in low-income economies as it is in high-income ones. To answer
this question, we set three different panels of 164 countries combined, which run from 1965
to 2017.
In each panel, we estimate the following augmented growth model:
𝑌 𝛼 𝛽𝑡 𝛽 𝐿𝑖𝑞𝑢𝑖𝑑𝑖𝑡𝑦 𝛽 𝐻𝐶𝐼 𝛽 𝐺𝑜𝑣 𝛽 𝐶𝑃𝐼 𝜀
Where 𝑌 is per capita GDP of country i in year t. 𝐿𝑖𝑞𝑢𝑖𝑑𝑖𝑡𝑦 is the financial indicator
while 𝐻𝐶𝐼 represents the human capital index. The size of government, 𝐺𝑜𝑣 , is
measured by the ratio of governments’ total expenditure to GDP. Finally, 𝐶𝑃𝐼 is the
consumer price index. Given the nature of our data, we include country-specific intercepts
and a time trend in our co-integrating equation.
An important topic in cross-country panels is the issue of non-stationarity. Since the data
series of our model are found to have unit roots at their level, we follow methods and
procedures to estimate non-stationary panels. We first report the results of panel unit root
tests for both the level and the first difference of each data series and investigate their co-
integration orders. After establishing that all series are integrated of order one, we run panel
co-integration tests to verify the existence of a long-run relationship between variables.
Finally, we estimate the coefficients of long-run models, which will illustrate whether the
impact of human capital (or financial development) on economic growth depends on
income level.
3.1. Panel unit root tests
Most common panel unit root tests are extensions of residual based tests that are frequently
used in non-stationary time series analyses. These tests, in particular, extend the
Augmented Dickey–Fuller test (ADF) to apply to panel data series. Broadly speaking, such
panel unit root tests can be divided into two categories: (1) tests that assume homogeneity
in the data generating processes across all cross section units of the panel, e.g. Levin, Lin
and Chu (2002); (2) tests that allow for some form of heterogeneity across units, especially
in the autoregressive parameters, e.g. Im, Pesaran and Shin (2003). These tests offer a more
flexible specification of the data generating process and capture a wider range of non-
stationarity in the data as they only require one or some of the panel series to have unit root.
The null hypothesis of these tests implies the presence of a unit root. Therefore, when the
null hypothesis is rejected, we conclude that the series is stationary.
In this study, for each series we conduct five different unit root tests. Table 2 shows the
results of these tests performed first at the level, and then after taking the first difference.
With the exception of some irregularities in human capital index, the results strongly
confirm that all series have a unit root at the level. However, after taking the first difference,
the series become stationary. This establishes that all the variables are integrated of order
one, i.e. I(1).
Table 2. Panel Unit Root Tests
Output Liquidity HCI Gov. Size CPI
Log ∆log log ∆log Log ∆log log ∆log log ∆log
High-income Countries
Fisher-ADF 18.32 -16.23*** 9.67 -26.10*** 2.59 -5.48*** 3.41 -33.30*** 5.05 -7.11***
Fisher-PP 24.72 -19.34*** 0.88 -18.23*** 9.39 -1.42* 3.60 -36.09*** 17.52 -8.03***
IPS 0.77 -20.14*** 0.26 -17.11*** 1.40 -1.43* -0.71 -14.93*** 16.35 -6.86***
LLC -1.10 -21.03*** 0.31 -12.96*** 4.75 -2.36*** 1.09 -29.17*** 3.39 -8.40***
Breitung 4.89 -16.69*** -0.95 -9.43*** 3.38 -1.98** 0.57 -10.8*** 5.69 -8.83***
Mid-income Countries
Fisher-ADF -4.98 -23.66*** 1.07 -30.11*** 9.60 -2.99*** 0.95 -50.04*** 12.81 -14.07***
Fisher-PP 0.85 -30.75*** -1.14 -34.34*** 9.36 1.39 1.15 -59.51*** 23.49 -18.02***
IPS 3.09 -29.29*** -0.42 -33.65*** 3.31 -4.25*** -2.11 -6.51*** 2.49 -19.28***
LLC 18.95 -27.51*** -1.26 -32.70*** 9.25 -1.21 0.75 -38.91*** 14.98 -26.59***
Breitung 7.52 -16.45*** 1.06 -24.82*** 0.54 0.72 0.53 -34.39*** 2.64 -10.25***
Low-income Countries
Fisher-ADF 1.83 -14.02*** 0.48 -14.86*** 4.21 -2.07*** -1.24 -15.98*** -0.65 -8.54***
Fisher-PP 3.23 -19.37*** 0.32 -19.26*** 8.54 1.32 0.72 -22.58*** 2.53 -10.78***
IPS 2.28 -18.26*** 1.15 -14.70*** 3.56 -2.78*** 1.31 -14.39*** -1.23 -9.45***
LLC 1.97 -18.12*** -0.31 -13.14*** 0.38 0.95 0.17 -14.91*** -1.11 -10.54***
Breitung 3.28 -9.17*** -1.21 -9.64*** 7.12 -5.57*** 1.11 -9.17*** 1.06 -8.55***
Note: *, **, and *** denote significance at 10, 5, and 1 percent, respectively.
3.2. Co-integration tests

To avoid spurious regression analysis, we conduct panel co-integration tests that establish
the existence of a long run relationship between our panel variables. Establishing the
existence of such a relationship is crucial for our study because the focus of our question is
on how certain channels affect economic growth. Investigating these effects requires a long
run relationship between variables as economic growth, in essence, is a long-term concept.
In this paper, we conduct Pedroni co-integration tests that are the most widely used tests in
non-stationary panel data analysis. It should be noted that our panels are macro panels
consisting of country level data that are not systematically linked to each other and cross-
sectional dependence is less of a concern.
Pedroni (1999 and 2004) extends his previous residual-based tests and proposes several
tests to examine whether panel variables are co-integrated. These tests are based on
applying standard unit root tests on the residuals of a regression equation. The null
hypothesis of all of these tests is that variables are not co-integrated. Therefore, rejecting
the null hypothesis implies a co-integration relationship. The main difference between
these seven tests is the degree of heterogeneity that they assume among cross-sectional
units. Based on this assumption, Pedroni tests are divided to two major groups: within-
dimension and between-dimension test statistics. The within-dimension statistics allow for
heterogenous autoregressive parameters across panel whereas the between-dimension
statistics are constructed with the average of individual coefficients.
Table 3. Panel co-integration tests
High-income Mid-income Low-income
Panel Specific Parameter:
Modified PP-t 2.17** 2.77*** -2.35***
PP-t 3.88*** 4.84*** -2.36***
ADF-t 3.96*** 3.91*** -3.88***
Common Parameter:
Modified v- ratio -0.19 2.26** 3.88***
Modified PP-t 4.09*** 4.61*** -0.98
PP-t 4.09*** 5.03*** -1.69**
ADF-t 5.78*** 4.43*** -0.27
Note: *, **, and *** denote significance at 10, 5, and 1 percent, respectively.
The results of Pedroni co-integration tests are reported in Table 3. Given the nature of our
variables, in the specification of the equation, a panel-specific time trend is included. Also,
the optimal number of lags are determined using the Akaike Information Criterion (AIC).
The overall results strongly confirm that variables in all three panels are co-integrated.
Once the presence of a co-integration is approved, we proceed to estimate the co-integrating
vectors. The coefficients of these vectors allow us to answer the main question of this paper.
3.3. Long-run relations
We estimate the co-integrating vector between variables using two widely used methods in
non-stationary panels; namely Dynamic Ordinary Least Squares (DOLS) and Fully
Modified Ordinary Least Squares (FMOLS). FMOLS estimator was first developed by
Pedroni (2000 and 2001) as a consistent method to estimate co-integrating vectors in
dynamic panels. FMOLS follows a non-parametric approach to correct for biases caused
by endogeneity. In contrast, the Dynamic OLS estimator proposed by Kao and Chiang
(2000) includes leads and lags of variables to resolve this issue.
Table 4. Dynamic ordinary least square estimators

Liability HCI Government CPI
-0.055** 1.09*** -0.973*** -0.014***
High-income Countries
(0.028) (0.045) (0.234) (0.003)
0.207*** 0.737*** 0.301** -0.016***
Mid-income Countries
(0.031) (0.038) (0.141) (0.003)
0.109*** 0.326** 0.060 -0.076***
Low-income Countries
(0.021) (0.128) (0.163) (0.013)
Note: numbers in parenthesis are standard errors. *, **, and *** denote significance at 10, 5, and 1
percent, respectively.
Table 4 and Table 5 report the FMOLS and DOLS estimators for each panel of countries
separately. We include a deterministic trend in our co-integrating vectors, and in both
methods, use pooled estimators. Also, leads and lags in all Dynamic OLS estimations were
determined using the AIC. Given the nature of our economic question, we only analyze the
long run coefficients of the co-integrating vectors.
Table 5. Fully modified ordinary least square estimators
Liquidity HCI Government CPI
High-income Countries 0.026 1.103*** -1.473*** -0.014***
(0.020) (0.033) (0.205) (0.003)
Mid-income Countries 0.252*** 0.668*** -0.179*** -0.017***
(0.025) (0.037) (0.121) (0.003)
Low-income Countries 0.027 0.532*** 0.299* -0.049***
(0.021) (0.115) (0.168) (0.011)
Note: numbers in parenthesis are standard errors. *, **, and *** denote significance at 10, 5, and 1 percent,
respectively.
4. Discussion
As our main contribution, we add a new dimension to the existing analysis of how financial
development and human capital accumulation affect output growth. We separate countries
based on their income levels to investigate whether the relationship between economic
growth, on the one hand, and financial development or human capital, on the other, varies
with national income level.
Our findings reported in Table 4 and Table 5 suggest that when we consider all countries
together (1) human capital accumulation has the strongest positive impact on economic
growth; (2) financial development measured by the size of total liquid liabilities has a
positive impact on growth, though its effect varies throughout the entire sample; (3) size of
government almost always negatively affects economic growth. And (5) a rise in nominal
prices moderately lowers production.
These findings are in line with the existing literature. In our analysis, human capital
emerges as the dominant economic factor in determining the national income. This is
consistent with a long tradition of endogenous growth theory models that emphasize on the
role of human capital in the aggregate production of the economy (e.g. Romer, 1990; Lucas,
1988). Our findings do not suggest a robust and undisputable positive role for financial
development. This could be related to two factors: (1) the friction-reducing nature of
financial development. High-income with already well-functioning financial markets do
not seem to benefit from financial development as much as mid-income countries do.
Also, (2) this finding could be because of the proxy we use for measuring financial
development. Previous studies have shown that the effect of financial development on
growth varies with the financial indicator (e.g. Khan and Senhadji, 2000). In particular,
when it comes to financial liberalization, it’s been noted that different countries respond
differently to measures of volume or efficiency. This depends on the regulatory
environment in the country, and the channel through which financial development
transmits to growth (De Gregorio and Guidotti, 1995).
When we turn our attention to an income-based analysis of the size of these effects,
however, we can identify an important trend. The relative importance of human capital,
measured by its long run coefficient, steadily falls with the national income. Human capital
is highly effective in high-income countries, but its importance drastically falls in low-
income countries. This is robust across estimation methods, though best seen in DOLS
methods. On the other hand, financial development seems to have no effect in high-income
countries while it strongly contributes to growth among mid-income countries.
The sizeable change in the impacts of human capital that this paper documents has
important implications for policy makers. Our findings suggest that investing in human
capital is a far more effective policy in advanced economies than it is in mid-income or
low-income nations. This could be because of the skill complementarity of technology.
More developed nations have access to better technologies that require highly skilled
workers. In other words, when a country is using a technology that is closer to the frontier,
an increase in human capital is much more effective to achieve sustained economic growth.
On the other hand, in lower-income countries, though still effective, human capital
accumulation is not as important. We believe this could be attributed to institutional barriers
such as skill mismatch or bad governance that cause problems in the transformation of
human capital to innovation and technological improvement.
5. Conclusion
The relationship between economic growth, financial development, and human capital has
been extensively studied in the literature. It is well accepted that both financial development
and human capital enhance economic growth. However, perhaps due to the complex nature
of their interaction, not every aspect of these relations is fully understood. This study
explores whether the relative importance of these factors for economic growth vary with
country’ national income level. Our findings could have valuable policy implications for
designing growth-inducing economic policies for countries in different development
stages.
We collect data for 164 countries between 1965 and 2017, and based on their national
income, divide them to three groups of high-, mid-, and Low-income countries. For each
group, we set up a panel that includes output per capita, liquid liability, and human capital
index, as well as some control variables. We first establish the existence of a long-run
relationship between variables in all panels. Then, we estimate the parameters of the co-
integrating vector using DOLS and FMOLS techniques.
Our findings confirm that human capital has a strong positive impact on output in all panels.
However, an important finding of our paper is that high-income nations gain the most from
accumulating human capital. As the national income falls, the impact of human capital on
output per capita steadily shrinks such that low-income countries benefit the least from this
economic factor. Based on the existing literature, one could argue that various institutional
barriers such as legal system and corruption may impede skilled workers ability to innovate
and improve the level of technology in the economy, which translates into lower aggregate
output. Another explanation for this pattern is that further accumulation of human capital
in high-income countries that are closer to the technology frontier is more prolific
compared to lower-income nations that rely on technologies that are far from the frontier.
Anyhow, more studies are necessary to shed light on the underlying reasons behind this
pattern.
As for the role of financial markets, our findings suggest that mid-income countries seem
to gain the most from developing their financial system. This is perhaps due to the fact
high-income countries are already enjoying a well-functioning system that allocates
financial resources efficiently. Therefore, further de-regulation may not be as impactful as
long as economic growth is the main concern.
Our results have strong policy implications regarding international organizations and
national authorities who design a development path for a country. As we show in this paper,
countries in different stages of economic development have varying responsiveness to
changes in financial development and human capital. If the policy target is to maximize
growth rates, the optimal policy solution should take into account these differential impacts
and accordingly adjust the weights attributed to different growth determinants so to
successfully lead the country to higher economic growth rates.
References
Acemoglu, D., 1996. A Microfoundation for Social Increasing Returns in Human Capital
Accumulation. Quarterly Journal of Economics, 111(3), pp. 779-804.
Acikgoz, S. and Ali, M.S., 2019. Where does economic growth in the Middle Eastern and North
African countries come from? The Quarterly Review of Economics and Finance, pp. 172-183.
Aghion, P. and Howitt, P., 1992. A Model of Growth through Creative Destruction. Econometrica,
60(2), pp. 323-351.
Ang, J.B., Madsen, J.B. and Isla, R., 2011. The effects of human capital composition on
technological convergence. Journal of Macroeconomics, 33(3), pp. 465-476.
Benhabib, J. and Spiegel, M.M., 1994. The role of human capital in economic development evidence
from aggregate cross-country data. Journal of Monetary economics, 34(2), pp. 143-173.
Das, K., Harper, J. and Arora, R., 2014. Financial development, economic growth and human capital
accumulation: what is the link? Working Paper.
Fama, E.F., 1985. What's different about banks? Journal of monetary economics, 15(1), pp. 29-39.
De Gregorio, J. and Guidotti, P.E., 1995. Financial development and economic growth. World
Development, 23(3), pp. 433-448.
Hakeem, M., 2010. Banking development, human capital and economic growth in Sub-Saharan
Africa (SSA). Journal of Economic Studies, 37(5), pp. 557-577.
Hassan, M.K., Sanchez, B. and Yu, J.-S., 2011. Financial development and economic growth: New
evidence from panel data. The Quarterly Review of Economics and Finance, pp. 88-104.
Howitt, P. and Aghion, P., 1998. Capital Accumulation and Innovation as Complementary Factors
in Long-Run Growth. Journal of Economic Growth, 3, pp. 111-130.
Im, K.S., Pesaran, M. and Shin, Y., 2003. Testing for unit roots in heterogeneous panels. Journal of
Econometrics, 115(1), pp. 53-74.
James, C., 1987. Some evidence on the uniqueness of bank loans. Journal of financial economics,
19(2), pp. 217-235.
Kao, C. and Chiang, M., 2000. On the estimation and inference of a co-integrated regression in panel
data. Advances in Econometrics, 15, pp. 179-222.
Kargbo, A.A., Ding, Y. and Kargbo, M., 2016. Financial development, human capital and economic
growth: New evidence from Sierra Leone. Journal of Finance and Bank Management, 4(1),
pp. 49-67.
Kendall, J., 2012. Local financial development and growth. Journal of Banking and Finance, 36,
pp. 1548-1562.
Khan, M.S. and Senhadji, A., 2000. Financial Development and Economic Growth: An Overview.
IMF Working Paper, pp. 1-24.
Kilic, C. and Burcu, O., 2018. The Impact of Financial Development on Human Capital: Evidence
from Emerging Market Economies. International Journal of Economics and Financial Issues,
8(1), pp. 258-267.
Levin, A., Lin, C.F. and Chu, J.C., 2002. Unit root tests in panel data: asymptotic and finite-sample
properties. Journal of Econometrics, 108(1), pp. 1-24.
Levine, R. and Zervos, S., 1998. Stock markets, banks, and economic growth. American economic
review, pp. 537-558.
Lucas, R.E., 1988. On the mechanics of economic development. Journal of Monetary Economics,
22(1), pp. 3-42.
Mankiw, N.G., Romer, D. and Weil, D.N., 1992. A contribution to the empirics of economic growth.
Quarterly journal of economics, 107(2), pp. 407-437.
Pedroni, P., 2000. Fully Modified OLS for Heterogeneous Co-integrated Panels. Working Papers
No 2000-03.
Rajan, R. and Zingales, L., 1998. Financial development and growth. American Economic Review,
88(3), pp. 559-586.
Romer, P.M., 1990. Endogenous technological change. Journal of political Economy, 98(5),
pp. 71-102.
Sehrawat, M. and Giri, A.K., 2017. An empirical relationship between financial development
indicators and human capital in some selected Asian countries. International Journal of Social
Economics, 44(3), pp. 337-349.
Vandenbussche, J., Aghion, P. and Meghir, C., 2006. Growth, distance to frontier and composition
of human capital. Journal of Economic Growth, 11, pp. 97-127.
Appendix
We provide country names that we have classified as high-, mid-, and low-income in Table 6.
The classification is based on that of World Bank. We consider both lower middle income
and upper middle income classes as mid-income class. Our dataset includes 41 high-income,
102 mid-income, and 21 low-income countries.
Table 6. List of countries in each income level
High-income Mid-income Low-income
Argentina Angola Kyrgyz Republic Tonga Burundi
Australia Albania Cambodia Tunisia Benin
Austria Armenia Kiribati Turkey Burkina Faso
Belgium Azerbaijan Lao PDR Tuvalu Central African Republic
Bahrain Bangladesh Lebanon Ukraine Congo, Dem. Rep.
Barbados Bulgaria Libya Uzbekistan Ethiopia
Canada Belarus St. Lucia St. Vincent and the Gambia, The
Switzerland Bosnia and Sri Lanka Grenadines Haiti
Chile Herzegovina Lesotho Venezuela, RB Liberia
Cyprus Belize Morocco Vietnam Madagascar
Germany Bolivia Moldova Vanuatu Mali
Denmark Brazil Maldives Samoa Mozambique
Spain Bhutan Mexico Kosovo Malawi
Finland Botswana Marshall Islands South Africa Niger
France China North Macedonia Zambia Nepal
United Kingdom Côte d'Ivoire Myanmar Rwanda
Greece Cameroon Montenegro Senegal
Hong Kong SAR, China Congo, Rep. Mongolia Sierra Leone
Hungary Colombia Mauritania Togo
Ireland Cabo Verde Mauritius Tajikistan
Iceland Costa Rica Malaysia Tanzania
Israel Cuba Namibia Uganda
Italy Djibouti Nigeria Yemen, Rep.
Japan Dominica Nicaragua Zimbabwe
Korea, Rep. Dominican Republic Nauru
Luxembourg Algeria Pakistan
Macao SAR, China Ecuador Peru
Malta Egypt, Arab Rep. Philippines
Netherlands Fiji Papua New Guinea
Norway Micronesia, Fed. Sts. Paraguay
New Zealand Gabon West Bank and
Panama Georgia Gaza
Poland Ghana Romania
Portugal Equatorial Guinea Russian Federation
Saudi Arabia Grenada Sudan
Singapore Guatemala Solomon Islands
Slovenia Guyana El Salvador
Sweden Honduras Serbia
Trinidad and Tobago Indonesia São Tomé and
Uruguay India Principe
United States Iran, Islamic Rep. Suriname
Iraq Eswatini
Jamaica Thailand
Jordan Turkmenistan
Kazakhstan Timor-Leste
Kenya
Monetary policy and exchange rate pass-through in India
Bhavish SHARMA
Narsee Monjee Institute of Management Studies (NMIMS), Bengaluru, India
sharma.bhavish30@gmail.com
Abstract. The paper tests the causality between monetary policy and exchange rate based on data
from 1991 onwards as that was when India opened up the economy. According to theory, interest
rates are inversely related to exchange rates with the main linkage between them being capital flows.
External debt also has considerable effects. The paper uses trend-line and regression analysis to
test the impact of various regressors on the dependent variable exchange rate. The paper finds that
the monetary policy is not as effective as theory would suggest in its ability to control the exchange
rate and that the pass-through from repo rate to exchange rate is not complete. The paper provides
suggestions for improving the linkage or pass-through as well as suggest usage of other policies in
tandem with monetary policy of RBI for India to achieve the economic objectives for long-term
growth as well as tackling potential short-term crises.
Keywords: monetary policy, exchange rate, capital, interest rate.
JEL Classification: E52, E43, F31.

276 Bhavish Sharma
1. Introduction
Monetary Policy, as we know, is central bank’s directives to manage the money supply
through its various communications and policies which in turn influence the shape of the
economy. There are various tools at the disposal of the Central Bank to influence the
Monetary Policy such as interest rates, bank reserve requirements, and the amount of
government bonds that banks must hold and such more. The objective of the monetary
policy can be set too many things, such as inflation targeting or reducing unemployment
and so on. Given the particular objectives, a country’s central bank can either pursue a
contractionary or an expansionary monetary policy.
However, the monetary policy transmission may not always be effective and lead to the
policy objective as desired by the central bank of a country. The process through which
general prices, output and the economy as a whole is impacted by a change in the stance of
monetary policy by a central bank is referred to as monetary policy transmission. An
example of the monetary transmission in India is, let us suppose that the RBI wants to
conduct an expansionary monetary policy, in that case, it would lower the interest rates
which would boost investment in the economy as the RBI will be able to lend more to
commercial banks as interest rates are less and so the commercial banks will be able to pass
these benefits on to their consumers which reduces the cost of borrowing which stimulates
credit in the economy and thereby, aggregate demand and increase the output. If the public
is able to realize the benefits of RBI cutting the interest rate and are able to borrow cheaply
then monetary transmission is considered to be effective. The decision of the monetary
policy is supposed to effect things such as aggregate demand, price level, the exchange rate
and so on. However, this may not always be the case as sometimes the pass-through levels
are not up-to the mark.
A lot of economists have in fact argued that monetary policy is incapable of fulfilling the
objectives it sets out to achieve. There is a huge outcry amongst economists over the
conduct of monetary policy and many new theories are coming out which suggest that
fiscal policy is more capable of having the desirable impact on the economy. These views
imply the need for studies to be conducted which test the effectiveness of the monetary
policy so that other aspects such as effectiveness of fiscal policy can be given a look at.
A lot of economists in India as well are suggesting that fiscal policy needs to be more
proactive than it is and that the government should not set itself arbitrary fiscal deficit
numbers and instead should focus on achieving the policy objectives which the monetary
policy is unable to.
Keeping the above context in mind, it is also of common knowledge that emerging
economies suffer from incomplete pass through (Bhattacharya, 2011) whether it be interest
rate pass through or exchange rate pass through. India is no different when it comes to this
as it is one of the leading emerging economies of the world and is growing at a faster rate
than most of the other countries. However, it encounters several problems such as
unemployment and increasing levels of debt which needs tackling and sorting out.
We have seen over the past few years itself that the Reserve Bank of India, which is vested
with the powers to control the monetary policy in India, has continuously been dropping
Monetary policy and exchange rate pass-through in India 277
rates in an effort to boost India’s economy which in itself plants the seeds of doubt over the
effectiveness of these particular changes in the interest rate on the economy.
Now the objectives of the Reserve Bank of India are obviously one, to keep a check on
inflation levels whilst stimulating growth and also to control the exchange rate of the
economy. Especially in a developing country like India, it is imperative to have a check on
the exchange rate levels. An exchange rate too high can result in high devaluation of the
currency which may raise inflation and imports might take a hit or an exchange rate too
low can cause a large fall in export which stagnates the economy. Thus, it is crucial for the
RBI to control the exchange rate of the country via its monetary policy. Higher interest
rates ideally increase the value of the currency and decreases the exchange rate.
Conversely, lower interest rate decreases the currency’s relative value as it is unattractive
for foreign investment which increases the exchange rate. As we are aware, India follows
a managed floating exchange rate system which means that the exchange rate is decided by
market factors but only up-to a certain range, beyond which the RBI tries to intervene. The
effectiveness, timeliness of the RBI policy is of extreme importance especially given the
amount of people who rely on the exchange rates as India is a developing country vastly
reliant on trading. Thus, it is important to test for the causality and effectiveness of the
interest rate on the exchange rate especially in the Indian context.
There is a need for a study focusing on the effectiveness of the monetary policy to control
the exchange rate especially in an emerging economy like India. Over the past 20 years, an
effective link may be established over interest rate and exchange rate and the impact of
various other variables on this channel may also be looked at. The effectiveness of
monetary policy has always been in question and through this study that particular query
can be answered in context of its relationship with exchange rate. Finding out the causality
between monetary policy and exchange rate can help provide insights into some important
questions like how can the RBI effectively use the interest rate to control the exchange rate.
If the impact of interest rate on exchange rate is weak, what are the underlying factors
causing the above phenomenon? What can be done to improve the link between interest
rate and exchange rate so as to give the RBI a more effective control?
In this context, the present study challenges to analyse the linkage between monetary policy
and exchange rate, especially in the context of RBI’s policy effectiveness. Also the study
tries to assess possible policy measures and suggestions to improve the linkage between
RBI’s policy rate and exchange rate in India.
Literature in this field more often than not ends up coming to the conclusion that the pass-
through is incomplete in India. However, the way they arrive at the conclusion is vastly
different and provides one with the knowledge of the existing theory whilst with the
understanding that there are numerous different avenues that one can take to test the
monetary transmission in India. Pattanaik et al. (2008) examined the effectiveness of an
increase in interest rate to impact and defend the exchange rate. In India, in times of extreme
pressure on the external value of the rupee, the Reserve Bank of India has traditionally had
278 Bhavish Sharma
a propensity to opt for high interest rates. In its analysis, the paper found that the exchange
rate could not be impacted well enough in the long run showing a degree of ineffectiveness
of the interest rate’s impact on the exchange rate. Capital flows also seemed to have a
considerable impact on the channel between interest rate and exchange rate (Chakraborty,
2006). For the years 1993 to 2003, the researcher used quarterly data showing that net
capital inflows moved up and down over the years and did not show a pattern as such. The
paper also analysed how capital inflows have been trending in India since 1993 to
adjustments in the multiple variables especially the real exchange rate. Capital inflows,
exchange rate and interest rate all seemed to have a connected relationship. Since it was
real exchange rate, inflation too was accounted for. The paper argued that the Reserve Bank
of India's intervention had caused co-movement in these variables which reduced the
dependency of the exchange rate on the amount of flows of capital which was an interesting
insight. The influence of the aforementioned variables showcases the need for the inclusion
of these variables when testing monetary policy and exchange rate pass-through in India.
A lot of Indian studies regarding this topic take inspiration from the following foreign
literatures.
Firstly, Ito et al. (2008) assessed the exchange rate magnitude effect through a sample size
consisting of emerging economies and countries in East Asia. It is said that there has been
endogeneity of the exchange rate and monetary policy instruments. There is a greater
change in import prices than in domestic prices in the countries of East Asia. In contrast to
many literatures, Zorzi et al. (2005) conducted a study in which impact of interest rate
shocks were taken into account on exchange rate and it was found that a larger pass through
existed for emerging markets than in developed countries. This contrast again provides a
justification for the need to test for the pass-through in the setting of an emerging economy
that this paper aims to do.
In the current scenario existing in the world where we are in a global recession it is also of
significance to view the impact of monetary policy after a crisis. Goldfajn et al. (2003)
talked about the ability to which monetary policy could stabilize exchange rate especially
in the immediate aftermath of a currency crisis. In this context, it focuses on each country
which has experienced such a crisis and inferred that potent monetary policy increased the
probability of reversing a postcrisis currency undervaluation that an economy might be
experiencing and that a tight monetary policy is effective as it reverses the process through
normal and graduation appreciation of the home currency.
Since, in this paper we are looking at the Indian context, it is imperative to glance upon
studies revolving around south Asian countries. Yunus (2001) investigated the impact or
influence of monetary policy on the determination of exchange rate in the South Asian
countries. It employs the Johannsen cointegration model to test for cointegration. He found
that there was a long run relationship between the monetary policy and exchange rate,
however monetary policy by itself did not completely determine the exchange rate process
and used the cointegrating vectors from the models coupled with ADF tests to reach upon
this conclusion in his analysis. Along the same lines, Khan et al. (2016) examined the
monetary policy conduct of various south East Asian economies. The paper found that the
main objectives for monetary policy in these countries was to control for inflationary
expectations. Many countries especially India with the exclusion of Bangladesh were found
to have used the monetary policy stance to also affect the real exchange rate. Foreign
interest rates also played an active role in dictating the monetary policy stance of the
aforementioned countries.
Coming back to India based literature reviews, D'Souza (2003) investigated the current
objectives being pursued by monetary policy which included stabilizing the foreign capital
flows by using open market operations. There was also the intention to directly impact the
foreign exchange market by attacking the exchange rates so as to allow better integration
of India into the world economy. Money supply targets also needed to be revised as import
tariffs were reduced and downward price rigidity was present. Samantaraya (2009)
developed a unique indicator to gather the exact stance of India’s monetary policy in the
post-reform era. Monetary aggregates or policy rates were deemed not to be enough to
determine a particular stance. The index created gives the impact of monetary policy on
many variables of macroeconomics such as interest rate, bank’s interest, inflation and
output growth and so on. There was instant impact on interest rates with lags on output and
inflation which would be expected normally as transmission takes a certain time as well as
the degree of pass-through matters as well. Thus, this paper provides a good overview of
various economic variables to consider when dealing with monetary policy and exchange
rate.
It is also of importance to understand literature which dealt with the impact of monetary
policy on exchange rate in the context of India getting liberalized in 1991 such as
Hutchinson et al. (2010) that discussed the impact of RBI’s monetary policy conduct pre
and post liberalization. The paper aimed at ascertaining whether a discretionary monetary
policy by the RBI was plausible to be estimated by the normative monetary policy
equations like the Taylor’s rule and found that output of the economy seemed to be more
of a policy objective for the RBI rather than inflation or exchange rates and so on. Ray et
al. (1998) analysed the role of the two variables most integral to the study namely the
interest rate and exchange rate. The methodology used by the paper is an OLS regression
but the fully modified cointegration version of it. A long run relationship between these
two variables coupled with price and output is executed and it is deemed that the two
variables were endogenously determined by the central bank in the pre-liberalization era
and that with the advent of new reforms of 1991, a change in the transmission of monetary
policy was to be expected
Now, Bhattacharya et al. (2011) investigated the significance of monetary policy through
interest rate pass through into various channels. As has been documented throughout,
developing economies have weak financial structures and in the case of India, a large
informal sector is present along with it which results in reducing the effects of monetary
policy transmission. How so ever, a relatively more complete pass-through of exchange
rate was seen to be prevalent. The paper examined the interest rate pass through as well as
the exchange rate pass through in an attempt to find the most effective monetary policy
transmission. The paper finds that in India, the most influential way of conducting
monetary policy is by controlling exchange rate which controls inflation and other
economic variables showing definite linkage of the two factors.
280 Bhavish Sharma
Thus, the literatures for the combined study of interest rate and exchange rate largely show
a relatively incomplete linkage for emerging economies. There is also considerable impact
of capital flows along with inflation in relation with these two variables after review of
various pieces of literature.
The importance of a research between interest rate and exchange rate is undoubtedly high
especially in a country like India as such studies can help the RBI effectively control a
massive parameter such as the exchange rate through the monetary tools at its disposal.
Theoretically speaking, higher interest rates should lead to lower exchange rates and vice
versa. The model can follow a route such as the RBI increasing interest rate which would
increase the capital (portfolio and other kind) which would then increase the value of
India’s currency which would then decrease exchange rate.
Now the question of whether such an outcome can actually occur in a country like India
will be seen later in the paper. The review of literature tells us that India has had incomplete
pass through whether it be of interest rate or exchange rate. Though, interest rate pass
through is considerably weaker in India (Ansari, 2014). Thus, one could say that the
expected outcome of the model would be that there is a weak linkage between interest rate
and exchange rate. The addition of variables such as inflation and debt could show us their
degree of impact in the process of monetary policy transmission to the exchange rate.
3. Research methodology
This paper uses secondary data and attempts to complete a comprehensive quantitative
analysis. Along with the regression analysis, a trend-line analysis between the exchange
rate and repo rate will be conducted to establish a concrete relation between the two most
prominent variables of the study the trend-line analysis, using the same model, is carried
out through the software of Tableau. The regression analysis is carried out through the
usage of the software E-views.
The variables used for the analysis are:
 Exchange rate – It is the dependent variable as we need to test the effectiveness of
monetary policy for it.
The following variables are the independent variables in this analysis:
 Repo rate – It is naturally taken as this is the tool through which the RBI controls the
interest rate and attempts to achieve its objectives via this rate. The data for repo rate is
only from the year 2000 onwards. Therefore, for the years 1991-2000 bank rate numbers
are used as proxy for the repo rate.
 Inflation – The level of correlation as we are aware between exchange rate and inflation
is very significant. The proxy used for inflation is WPI i.e. the wholesale price index
and averages are calculated to obtain data for an annual basis.
 External Debt – The degree of impact external debt has will be a significant factor as
the rate of exchange rate always gets impacted by the degree of government debt and
repayments over the years.
 Capital Flows – As mentioned an increase in interest rate by RBI should increase capital
inflows which in turn influences the exchange rate. The proxy for capital flows used is
the foreign capital inflow for each year as that gives an idea of the trend of capital
inflows for the sample size taken.
The model used for the analysis of the dependent and the independent variable is the
Dynamic Ordinary Least Square model as this model is more effective than the normal
OLS and FMOLS as it takes into account the leads and lags in the model. It is less biased
than the other two models.
The model equation of the DOLS is given below
𝑦 𝑋𝛽 𝐷 𝛾 𝛥𝑋 𝛿 𝑣
Yt is the dependent variable and Xt in the above is the explanatory variable or the
independent variable. In our study, as discussed above we have four explanatory variables.
The lags and leads, q and r, are used to nullify the long-term correlation and also make the
stochastic error term independent of the regressors in the model and hence the term
summation of change in Xt augments the cointegrating regression using the leads and lags.
Based on the literature review, the cointegration test used will be the Johannsen- Julius
Cointegration model. Along with the cointegration test, a unit root test will also be
performed which is the Augmented Dickey-Fuller test which is one of the most popular
unit root tests and allows for a higher order autoregressive model and tells us whether the
equation is a ‘spurious regression’. Using the dynamic OLS model is a new take on this
very widely studied topic.
3.1. Period of study and sources of data
The data sample is for the years 1991 to 2019 as taking this sample provides a period
lengthy enough to run an authentic regression analysis. The data is collected from The
Reserve Bank of India’s database for the given period of study and changes are made to
the data at hand wherever deemed necessary in accordance to the regression equation.
The significance of the year 1991 when it comes to India can be easily contextualized. This
was the year when India officially opened up its economy. Though India had shown
tendencies of an open economy or a pro-business economy in the 80’s as well, the official
turnaround was in 1991 only under the guidance of Dr. Narsimha Rao and Dr. Manmohan
Singh. Since this paper aims to test the relation and causation of monetary policy and
exchange rate, it makes sense to take data from 1991 onwards only as that was when
exchange rate came into prominence naturally. The coefficient observed from running the
regression model will let us know the degree of effectiveness of changes in the monetary
policy on the exchange rate and provide us with imperative insights on the effectiveness of
282 Bhavish Sharma
monetary policy and whether some policy changes are required or not especially
considering the recent slowdown prevalent in the economy of India.
After performing the Augmented Dickey Fuller test using the AIC (Alkaine Information
Criteria) on all the variables in the model, it was found that all of them were stationary
Table 1. Results from ADF test tabulated
Variable T-stat Probability
Exchange_Rate -4.76142 0.0007
Repo_Rate -4.028365 0.0049
External_debt -6.262509 0
WPI -6.557076 0
Capital -8.111367 0
Source: Computed.
The null hypothesis for each of these variables was the presence of a unit root and as can
be ascertained from the t statistic scores as well as probability (all below 0.05) that the null
hypothesis can be rejected for each of the above cases. The t statistic scores are less than
the t critical scores at the 1%, 5% and 10% level which rejects the null hypothesis.
Therefore, all the variables used to conduct the analysis are found to have stationarity.
Table 2. Results from cointegration test
Unrestricted Cointegration Rank Test (Trace)
Hypothesized Trace 0.05
No. of CE(s) Eigenvalue Statistic Critical Value Prob.**
None* 0.697602 70.37438 69.81889 0.0451
At most 1 0.503128 38.08203 47.85613 0.2984
At most 2 0.329539 19.19763 29.79707 0.4788
At most 3 0.258175 8.403302 15.49471 0.4233
At most 4 0.012513 0.339981 3.841465 0.5598
*Denotes rejection of the hypothesis at the 0.05 level.
**MacKinnon-Haug-Michelis (1999) p-values.
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level.
Source: Computed.
Table 3. Results from cointegration test

Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized Max-Eigen 0.05
No. of CE(s) Eigenvalue Statistic Critical Value Prob.**
None 0.697602 32.29234 33.87687 0.0763
At most 1 0.503128 18.88441 27.58434 0.4235
At most 2 0.329539 10.79432 21.13162 0.6678
At most 3 0.258175 8.063321 14.26460 0.3723
At most 4 0.012513 0.339981 3.841465 0.5598
* Denotes rejection of the hypothesis at the 0.05 level.
** MacKinnon-Haug-Michelis (1999) p-values.
Max-eigenvalue test indicates no cointegration at the 0.05 level.
Source: Computed.
The Johannsen- Julius cointegration test was also performed using all the variables in the
model, though the model this paper enlists corrects for any cointegration or endogeny
present in the variables. The cointegration results have shown that in the maximum
eigenvalue test no traces of cointegration was found. However, in the trace test it is seen
that one co-integrating equation had been found at the 0.05 level which is in a way a
positive aspect as the variables are of the same field but as has been mentioned already, the
prevalent cointegration will be corrected by the model in application.
The trend line analysis performed between the two most pertinent variables in the paper
i.e. interest rate and exchange rate provided interesting insights as can be seen from the
illustration below
Figure 1. Trend line between exchange rate and repo rate
The trend-line analysis shows that there is an inverse relationship present between interest
rate and exchange rate which exemplifies the basic theory that an increase in interest should
decrease exchange rate and vice-versa. A linear trendline was not used as that would not
have given the realistic picture. Using an exponential trendline signifies the relationship
but at the same time shows that the pass-through is not as complete as it should be.
However, a trendline is not a significant enough measure to conclude the relationship
between the given variables which brings us to the results of the regression analysis which
are as follows:
284 Bhavish Sharma
Table 4. Results from running DOLS regression

Variable Coefficient Std. Error t-Statistic Prob.
REPO_RATE -0.455331 0.401708 -1.133488 0.2863
EXTERNAL_DEBT 0.000280 4.90E-05 5.723528 0.0003*
WPI 0.012203 0.020147 0.605694 0.5597
CAPITAL -9.01E-05 3.06E-05 -2.943440 0.0164*
C 15.20599 7.326829 2.075385 0.0678
R-squared 0.991723 Mean dependent var. 46.43623
Adjusted R-squared 0.977007 S.D. dependent var. 10.49333
S.E. of regression 1.591140 Sum squared resid. 22.78553
Long-run variance 2.991012
* Significant at 5%.
Source: Computed.
The above results from the regression analysis show a very high R-square value which
shows the goodness of fit for a regression model and the high value proves the capability
and the appropriateness of the model in the sense of the ability of regressors to explain the
changes in the dependent variable (exchange rate) collectively as a whole. The coefficients
tell us the degree as to which exchange rate changes with a change in the independent
variable. The signs associated with the variables are as expected according to the theory.
The negative coefficient between repo rate and interest rate are as expected. Similarly, with
capital (capital inflows) the negative sign is in accordance to the theory that higher capital
inflows would lead to decrease in exchange rate due to appreciation in the domestic
currency. As far as external debt and inflation is concerned, increase in WPI and external
debt would cause a rise in exchange rate as they are factors that tend to induce depreciation
in the home currency. However, the repo rate and WPI values are not significant even
though the signs are in accordance. External debt and capital have significant coefficients
at the 5% level implying that the p value obtained was less than 0.05 thereby making them
significant.
The coefficients value obtained also paint an interesting picture. The -0.45-coefficient
value implies that the pass-through effect from RBI’s monetary policy to the exchange rate
is incomplete and that there is scope for improvement. The really low value of capital
implies that maybe the influx of foreign direct investment in influencing the exchange rate
is not as significant as initially thought out to be or maybe the usage of foreign direct
investment as proxy does not paint the reality of capital flows in India. The inflation rate
influences the exchange rate more than India’s external debt though both factors cannot be
ignored in their role especially external debt which had a significant value. External debt
directly impacts a country’s forex reserves and thereby influences a country’s ability to
borrow and import which impacts the exchange rate. Thus, these are the observations and
inferences drawn from the analysis conducted in the paper.
Conclusion and policy suggestions

The extensive development of India’s financial markets especially the domestic financial
markets have resulted in the RBI relying more on the interest rate in comparison to any
other policy tool to dictate India’s monetary policy stance (Mohanty, 2012). Thus, it
becomes imperative to study the effectiveness of the monetary policy transmission and in
this study, the transmission to the exchange rate. After the discussion and analysis done in
the paper, it is sufficing to say that there exists a linkage between the interest rate and
exchange rate. The direct trend line analysis and regression analysis shows there is a
definite correlation between the two variables and as theory suggests, causation to go along
with it as well.
However, as seen in the paper the degree of pass-through of the monetary policy conducted
by the RBI to the exchange rate is not as complete as one would imagine. There seems to
be a certain amount of inefficiency when it comes to the ability of RBI to impact the
exchange rate. In fact, a less than half change is seen in the exchange rate with a change in
interest rate. Now, there could be several underlying factors that could cause this
discrepancy. One of them could be a result of the amount of forex reserves in the country.
In recent years, the RBI has taken large amount of short-term foreign debt which as of now
amounts to 57% of forex reserves (Source: RBI) with the RBI which proves that the foreign
reserves at hand are depleting continuously with the result being a bleak future when we
have to pay off the debt. This along with the external debt impacts the exchange rate
separately and may counter the effect a change in interest rates will have. As we all know
that in recent times, RBI has been continuously been cutting interest rates, now more so
than ever, but its impact on exchange rate is not as well pronounced. Now a reduction in
interest rates is considered to be expansionary monetary policy as it would be considered
to boost investment in the economy and output but at the same time the amount of capital
flows is impacted negatively which hampers the economy. Our current account deficit
(CAD) is largely plugged by foreign international inflow and FDI with them constituting
as much as 1% of our CAD (Source: CMIE) so a flight of capital means an increase in
current account deficit which weakens India’s position as an economy. Now a possibility
of reduction in capital flight should decrease the value of currency and that in turn should
increase exchange rate to help exporters in the economy but as analysis suggests that this
pass through from interest rate to the exchange is not seen so it could be said that the
monetary policy employed in the Indian context is quite ineffective much to the dismay of
RBI. Review of literature suggests that the lag on effects of monetary policy is lesser than
that of fiscal policy so in a crisis especially one we are experiencing now, monetary policy
would have been quite effective to resolve the economic issues that are being experienced
such as low output and GDP growth.
However, monetary policy is not very effective and that means maybe as a country, India
should give a stronger look towards expansionary fiscal policy especially in the form of
government spending and not be bogged down by a specific target number to maintain the
fiscal deficit especially at a time when inflation is not at worrying levels. At the same time
some measures that could be taken by the RBI to improve the effectiveness of monetary
policy to the exchange rate would be to decrease the amount of debt burdened by the forex
reserves as that acts as a deterrent to the capability of India to control the exchange rate by
buying or selling the rupee and allows foreign circumstances to dictate the value of the
exchange rate especially the US as the Indian rupee is held against the dollar. A high level
of debt coupled with low reserves makes the exchange rate vulnerable to outside pressure
and their demands. These pressures would only increase the current account deficit which
is harmful for the economy.
286 Bhavish Sharma
The reason to ensure effective transmission of monetary policy to the exchange rate can be
to use the interest rate to impact the exchange rate which in turn can control inflation in the
economy. Hence on one side, monetary policy can be used to control inflation via the
exchange rate as exchange rate pass-through to inflation is high according to a lot of studies
and on the other side expansionary fiscal policy can be implemented to boost output and
demand. Therefore, for both to work in tandem, it is imperative of RBI to improve its ability
to control the movement of the exchange rate and not allow it to get influenced by external
factors.
Limitations and scope for further research
There were limitations that were encountered during the conduct of the study. The most
important being the fact that repo rate began only in the 2000 so bank rate had to be used
as proxy for the years before which may have led to fluctuations in the results. Another
inhibiting factor according to the results may have been the usage of foreign direct
investment as proxy for capital inflows.
There is scope for further study as various other models can be implemented to prove the
causation between the repo rate and exchange rate. A shorter period of study could be taken
with focus on the fluctuations year by year which may give a better insight into the reasons
behind the effectiveness/ineffectiveness of monetary policy. The current scenario where a
pandemic has pushed the world into global recession opens up new avenues of studies to
be conducted to see and test the effectiveness of the monetary policy amid this crisis and
in the aftermath as well.
References
Ansari, J. and Ashima, G., 2014. Bank Competition, Managerial Efficiency and the Interest Rate
Pass-Through in India. Risk Management Post Financial Crisis: A Period of Monetary Easing.
Emerald Group Publishing Limited, pp. 317-339.
Bhattacharya, R., Ila, P. and Ajay, S., 2008. Exchange rate pass-through in India. Macro/Finance
Group at National Institute of Public Finance and Policy, [Online]. Available at:
<http://macrofinance.nipfp.org.in/PDF/BPS2008_erpt.pdf>
Bhattacharya, R., Ila, P., and Ajay, S., 2011. Monetary policy transmission in an emerging market
setting. IMF Working Papers, pp. 1-25.
Ca' Zorzi, M., Hahn, E. and Sanchez, M., 2007. Exchange Rate Pass-Through in Emerging Markets
ECB Working Paper No. 739, Available at SSRN: <https://ssrn.com/abstract=970654>
Das, S., 2015. IMF Working Paper: Monetary Policy in India – Transmission to Bank Interest Rates.
USA: International Monetary Fund. doi: <https://doi.org/10.5089/9781513598796.001>
Deepak, M. and Mitra, A.K., 1999. Experience with Monetary Targeting in India. Economic and
Political Weekly, Vol. 34, No. 3/4, pp. 123-132. JSTOR, <www.jstor.org/stable/4407572>
Dhal, S.C. and Ansari, J., 2013. Interest Rate Pass-through and Determinants of Commercial Banks'
Loan Pricing Decisions in India: Empirical Evidence from Dynamic Panel Data
Model. Banking & Finance Review, 5(1).
Engel, C., 2014. Exchange Rate Stabilization and Welfare. Annual Review of Economics, 6,
pp. 155-177. Retrieved April 6, 2020, from <www.jstor.org/stable/42940286>
Errol D'Souza, 2003. What Is Monetary Policy Doing? Economic and Political Weekly, 38(8),
pp. 821-823. Retrieved April 6, 2020, from <www.jstor.org/stable/4413253>
Goldfajn, I. and Gupta, P., 2003. Does Monetary Policy Stabilize the Exchange Rate Following a
Currency Crisis? IMF Staff Papers, 50(1), pp. 90-114. Retrieved March 30, 2020, from
<www.jstor.org/stable/4149949>
Gregori, T., 2009. Currency crisis duration and interest defence. International Journal of Finance
& Economics 14.3 pp. 256-267.
Hofmann, B., and Mizen, P., 2004. Interest Rate Pass-Through and Monetary Transmission:
Evidence from Individual Financial Institutions' Retail Rates. Economica, Vol. 71, No. 281,
pp. 99-123. JSTOR, <www.jstor.org/stable/3549170>
Hutchison, M.M., Sengupta, R. and Singh, N. (2010). Estimating a monetary policy rule for
India. Economic and Political Weekly, pp. 67-69.
Indrani, C., 2006. Capital Inflows during the Post-Liberalisation Period. Economic and Political
Weekly, Vol. 41, No. 2, pp. 143-150. JSTOR, <www.jstor.org/stable/4417673>
Khan, M., Ahmed, A., 2016. Conducting Monetary Policy in South Asian Economies: An
Investigation. The Pakistan Development Review, 55(3), pp. 161-190. Retrieved April 6, 2020,
from <www.jstor.org/stable/44986055>
Khundrakpam, J.K., 2008. Have economic reforms affected exchange rate pass-through to prices in
India? Economic and Political Weekly, pp. 71-79.
Monacelli, T., 2005. Monetary policy in a low pass-through environment. Journal of Money, Credit
and Banking, pp. 1047-1066.
Partha, R., Himanshu, J., and Mridul, S., 1998. New Monetary Transmission Channels: Role of
Interest Rates and Exchange Rate in Conduct of Indian Monetary Policy. Economic and
Political Weekly, 33(44), pp. 2787-2794. Retrieved March 30, 2020, from
Pattnaik, R.K., Kapur, M. and Dhal, S.C. (2003). Exchange rate policy and management: The Indian
experience. Economic and Political Weekly, pp. 2139-2153.
Rajwade, A., 2014. India's 'Dutch Disease' and the Exchange Rate. Economic and Political Weekly,
49(18), pp. 17-19. Retrieved April 6, 2020, from <www.jstor.org/stable/24480215>
Ramachandran, M. and Naveen, S., 2007. Asymmetric exchange rate intervention and international
reserve accumulation in India. Economics Letters 94(2), pp. 259-265.
Ratnu, S. (2019). Interest rate pass through in India: Bank Lending Channel. NMIMS Journal of
Economics and Public Policy, 4, 1, pp. 74-81.
Renu, K., 2001. Will Interest Rate Cut Work? Economic and Political Weekly, 36(10), pp. 809-810.
Retrieved April 6, 2020, from <www.jstor.org/stable/4410363>
Johansen, S., and Juselius, K., 1990. Maximum Likelihood Estimation and Inference on Co-
integration-with Application to the Demand for Money, Oxford Bulletin of Economics and
Statistics, Vol. 52, pp. 169-210.
288 Bhavish Sharma
Samantaraya, A., 2009. An Index to Assess the Stance of Monetary Policy in India in the Post-
Reform Period. Economic and Political Weekly, 44(20), pp. 46-50. Retrieved April 6, 2020,
from <www.jstor.org/stable/40279012>
Sivramkrishna, S. (2016). The ornithology of macroeconomic policy: India's new monetary policy
framework. NMIMS Journal of Economics and Public Policy, 1(1), 45-55.
Singh, B., and Sitikantha, P., 2012. Monetary Policy and Asset Price Interactions in India: Should
Financial Stability Concerns from Asset Prices Be Addressed Through Monetary Policy?
Journal of Economic Integration, Vol. 27, No. 1, pp. 167-194. JSTOR,
Smets, F., and Raf, W., 2002. Openness, imperfect exchange rate pass-through and monetary policy.
Journal of monetary Economics, 49(5), pp. 947-981.
Ito, T, Sato, K., 2008. Exchange Rate Changes and Inflation in Post-Crisis Asian Economies: VAR
Analysis of the Exchange Rate Pass-Through, Journal of Money Credit and banking, Vol. 40,
pp. 1407-1438.
Yunus, M., 2001. Monetary Interpretation of Exchange Rates in the South Asian Countries.
The Bangladesh Development Studies, 27(1), pp. 73-103. Retrieved March 30, 2020, from
Impact of medium of instruction in education

on economic growth and development
Durbhita V VAISHNAV
The Maharaja Sayajirao University of Baroda, Gujarat, India
durbhitavaishnav1994@gmail.com
Abstract. The main factors behind modern education system are, subjects taught from primary
schools to universities. Information, theories, concepts, ideas, life stories – everything is western.
Today it is known by English medium education. Someone rightly said that “If we want to destroy
any people or nation without any blood transfusion and without zipping a bomb then it is necessary
to separate the person from his religion, culture and language”.
Hence these three parameters (religion, culture and language) play a significant role in nation
transformation but we will consider only one parameter in this paper i.e. language. To find out the
effect of medium of instruction in education on social and economic aspects, the researcher will
select all the countries having developed, developing and miserable economy of the world.
Further, this paper had discussed about the effect of English medium education in higher education
on economic development. “Till date, as far as I know no study in economic growth has investigated
the effect of English on growth either empirically or theoretically. A small number of studies have
considered the importance of language for spillovers”. So, researcher is curious to know the
spillover effect of English language on the economic development.
Keywords: education and human development, economic thought, human growth and development,
India.
JEL Classification: F6, O4, O5.

290 Durbhita V Vaishnav
Introduction
“English is the dominant international language of the 21st century. It is spoken at a useful
level by some 1.75 billion people – a quarter of the world’s population. As the language of
communications, science, information, technology, business, entertainment and
diplomacy, it has increasingly become the operating system for the global conversation.”
Due to colonization, most of countries use the colonial languages as their official
languages. The colonial languages such as English, French, Portuguese and Spanish were
still holding a major part of the world. “The official language is defined as “The language
in which the primary affairs of the community – the government, the media, the courts and
the school – are conducted.”
The western domination over the globe was gradually increased because of the trading by
Europeans. It was widely accepted as a result of its impact over economic growth.
They destroyed countries, not by the physical attack but by the mental attack. It is worldly
known that how Macaulay frolicked in the case of India. He wrote to his father that “If we
want to destroy India, we have to attack its education system” and they exactly did the same
thing in India. Now the question is how?
1. First, they established an English medium school in India.
2. Second, they introduced English as one of the compulsory subjects in every grade
irrespective of medium of instruction. Without this, we will not get admission in college
and university.
3. Thirdly, in every sense create an inferiority complex in the mind of a non-English
medium student at school and university level.
4. They destroyed the nation as a whole by establishing western philosophy through
education.
These were indirect attacks by the British on Indian mindset. They destroyed Indian
philosophical value system, their norms and rituals etc. Indians still feel proud to say hello
and good morning in everyday usage instead of greeting “નમસ્તે” (Namaste).
Every developing country may have same story like India. Every colonial country was
prosperous in terms of knowledge, wealth and their value system and had unique status in
the world. But as a result of colonization, the colonial countries lost their status, wealth and
value system. Though these countries became independent from the colonial ruler, but all
were either under or less developed. What is the reason behind this? “By birth, they are the
citizen of their respective countries but their mindset is still colonized” as they follow
western culture and consider their culture as inferior one.
In less developed, underdeveloped and developing country, more than 50% population
belongs to the middle-class family. These all middle men are engaged in farming or small-
scale business activity. They put huge efforts into giving the best education to their child
in a good institution, even at the cost of their own basic needs.
1. The elite class of society in these countries send their kids to English medium school
and its spillover effect on middle man, so due to the lack of social and economic
overheads in rural area, many of them also migrates to urban area in the search of good
English medium institution for their kids.
Impact of medium of instruction in education on economic growth and development 291
2. Due to this agriculture-related business fell, but it doesn’t mean that dependency on
agriculture is reducing and contribution of the manufacturing sector is increased.
Actually, all those who migrated from rural to an urban area for the education purpose
were largely belonging to the unskilled workforce. They did not get a job in the industry
and thus they started a small-scale business activity to earn minimum daily wages.
So, it is directly or indirectly affecting the Agro-economy. For instance, Latin et al. (2014)
using a cross-country and micro-level analysis have found that the average distant to
official language, which captures the ability to speak an official language, has a negative
and statistical impact on growth and individual outcome (literacy, income, etc.).
In the above two-paragraph, we discussed social and economic effect due to the trend of
English medium of instruction in higher education in the world. To check whether the
English as a medium of instruction in higher education in various countries of the world
affects the HDI and GNI ranks of the country.
What do we mean by “development”?
In earlier days economic development was confined to GDP per-capita. It has been argued
by development economists and even by Nobel Prize laureate Samuelson in a famous paper
from 1951, income is clearly an imperfect metric as a measure of development. As it is
obvious from the UNDP's Human Development Index, or the focus of a number of world-
renowned research institutions on very particular indicators (the International Food Policy
Research Institute's use of child anthropometrics being a typical example), development is
a multidimensional process that can be measured along a plethora of dimensions. Maternal
and infant mortality, access to clean drinking water, the prevalence of infectious diseases,
the empowerment of women, literacy or school enrolment rates, measures of the incidence
of poverty or access to sufficient calorie and protein intake, life expectancy at birth.
As said above. this paper had analysis of the effect of English medium education in higher
education and economic growth of countries, researcher selected 7 continents of the world,
51 countries of Asia, 44 countries of Europe, 51 countries of Africa, 37 countries of North
America, 14 countries of South America and 14 countries of Australia.
Statement of problem
Impact of medium of instruction in education on economic growth and development.
Need and significance
Human resource is a fundamental unit of economic development. It is developed through
education which is received in different medium of instruction. This also defines the state
of development. Thus, the medium of instruction is prime factor to be analyzed for
economic development.
Scope and coverage of research
There are several aspects for measuring economic development such as availability of human
resource, physical capital, natural resources and technology. But here, the human resource is
taken into consideration with confinement to medium of instruction in education.
Research objective
1. To explore the consequences of a language that is chosen as an official language and is
very “distant” from the languages spoken by the majority of the population.
2. To analyses, whether the English as a medium of instruction in education in various
countries affects the HDI and GNI ranks of the respective country.
3. To check how vernacular languages as a medium of instruction in higher education in
various countries of the world affects the ultimate economic development of the country.
4. To determined effects of medium of instruction on economic development by
comparing GNI, HDI and medium of instruction in higher education through flow
diagram.
Materials and methods
Research methodology
Conduct a study based on secondary data which are collected from online sources for the
medium of instruction. GNI and HDI rank are referred from World Bank report 2018.
Data analysis and interpretation
1. Asian Continent
Figure 1. Medium of instruction in various country in Asia
Source: languages of the world – Nation online project.

Table A-1. Description of various categories of class used for study of data of medium of instruction in
higher education for Table 1.1
Colour Medium of instruction HDI and GNI rank No. of countries
Yellow English Low Scale 14
Red English and Vernacular language Low Scale 06
Blue Vernacular language Medium Scale 8
Pink English High Scale 8
Green Vernacular language Low Scale 3
Grey English and Vernacular language High Scale 3
No colour Data not available Data not available 9
Table 1.1. Asian countries GNI and HDI rank compare with its medium of instruction 1
Sr. Country Moi HDI Rank GNI (per capita)
No. 2018 2018
1. Afghanistan English 168 183
2. Armenia English 83 111
3. Azerbaijan English 80 118
4. Sri Lanka English 76 116
5. Bangladesh English 136 151
6. Bhutan English 134 132
7. Turkey English 64 68
8. Cambodia English 146 157
9. Maldives English 101 73
10. Pakistan English 150 155
11. Vietnam English 116 139
12. Georgia English 70 115
13. India English 130 144
14. Indonesia English 11 120
15. Oman English 48 56
16. Singapore English 9 11
17. Saudi Arabia English 39 40
18. Qatar English 37 08
19. Bahrain English 43 38
20. Brunei (Darussalam) English 39 29
21. Brunei (Darussalam) English 39 29
22. Cyprus English 32 34
23. Jordan Arabic, English 95 113
24. Myanmar-Burma Burmese, English 148 163
25. Iraq Arabic, English 120 102
26. Philippines Filipinos And English 113 121
27. Tajikistan Russian And Tajik 127 167
28. Lebanon English And French 80 81
29. Iran Persian, English 60 98
30. Israel Hebrew, Arabic, English 22 25
31. Japan Japanese 19 21
32. Malaysia Malay 57 66
33. Kuwait Kuwait 56 27
34. Kazakhstan Kazakh Or Russian 58 79
35. Turkmenistan Turkmen 108 86
36. United Arab Emirates Uzbek, Tajik, Turkish, 34 24
37. Thailand Thai 83 87
38. China Chinese 86 71
39. Uzbekistan Uzbek, Russian 105 144
40. Yemen Arabic 179 168
41. Nepal Nepali 149 168
42. Mongolia Chinese and Tonglia 92 124
43. Kyrgyzstan Uzbek,Tazik, Turkish 122 ------
44. Korea- North Macau, Cantonese 22 ------
Sr. Country Moi HDI Rank GNI (per capita)

No. 2018 2018
45. Laos Lao 139 -----
46. Korea- South Macau, Cantonese 22 -----
47. Macau Portuguese ---- -----
48. Palestine Arabic 119 -----
49. Syria Arabic 155 -----
50. Taiwan Chinese ------ -------
51. Timor-Leste English ----- 149
Source: (Languages of the world – Nation online project/Human Development Report Office, 2018/Gross
national income per capita 2018, Atlas method and PPP).
Countries having a medium of instruction in the vernacular language have good HDI and
GNI ranks but countries having English as a medium of instruction in the higher education
possess lower GNI and HDI rank. According to Betty Mkwinda-Nyasulu “Communication
is an important prerequisite of development and this is manifested through language.
Language must be seen to be communicating the intended meaning as a vehicle to
achieving set goals, and not an end in itself. These set goals in this case, would be education,
national unity or identity, and socio-economic development.”
According to Christopher McCormick “Billions of people around the globe are desperately
trying to learn English-not simply for self-improvement, but as an economic necessity. It’s
easy to take for granted being born in a country here people speak the lingua franca of
global business, but for people in emerging economies such as china, Russia and India,
where English is not the official language, good English is a criteria tool, which people
rightly believe will help them tap into new opportunities at home and abroad”. Thus, it is
discerned from the above analysis that the English as a medium of instruction in higher
education, directly or indirectly affects the Human development and National Economy in
Asian continent.
Study of “lee Che Ging Januar” totally opposed this study. According to him “The level of
English proficiency can be viewed as a part of human capital that the ability to absorb
knowledge is positively related to the level of English proficiency. Therefore, countries
with having higher level of English proficiency among other, the fraction of its population
are likely to grow faster. This empirical result provides evidence of positive correlation
between initial English proficiency and economic growth only for the countries in the Asia
and Europe. The spectacular growth of Asian countries can be attributed to the heavy
investment in the creation of human capital that fosters English speaking culture and
promotes a climate of the use of English. According to him increase in English proficiency
will directly accelerate the knowledge, absorptive capabilities of worker.”
This is not true in the case of developing nations. Because in majority, colonial countries
of Asia use English as medium of instruction in higher education. Due to English as
medium of instruction in higher education, less student enrolls for higher education. Also,
in developing countries major people are residing in rural area, so student migrates from
rural to urban area for education. Out of these majority students are from vernacular
medium. They attend lecture regularly in college but they did not understand any concept
during class. So, it’s like west of time to attend lecture because of not being able to
understand anything whatever taught in the class and without understanding the concept,
how one can remember the concept and failing to remember the concept led to mug up the
concept and appear with the same in the exam. Some of them crack the exam and other
may fails continuously. So, next year they dropout from college and start looking to earn
money by their own business or go on with their father’s occupation. Now those who passes
the exam by mug-up method without understanding the concept face the difficulties at the
time of employment. When they appear in interview, some of them may know the concept
but could not express themselves in English language. So ultimately economic develop-
ment of country directly and indirectly depends upon social overheads, in Asia social
infrastructure is week due to use of colonial language in higher education. “Mustafa said
English cannot solve our ills. There are not enough teachers who know English and can
teach in English. Children cannot comprehend what they are taught”, that’s why we are
generating inferior quality of scientist, doctors, engineers, teachers, technologist. Due to
poor social infrastructure of education, developing countries are facing many challenges
like unemployment, poverty, corruption, etc. The main reason of this evil is that the wisdom
of particular branch does not reach to student. So, they do not understand values system,
ethos, morals value, norm and logic behind the particular branch of knowledge.
2. Europe continent
Figure 2. Medium of instruction in various country in Europe

Table A-2. Description of various categories of class used for study of data of medium of instruction in
Colour Medium of instruction HDI and GNI ranks No. Of countries
Red English Good 4
Brown Vernacular language Good Scale 40
Table 1.2. European countries GNI and HDI rank compare with its medium of instruction
Sr. HDI
Country Moi GNI Rank
No. Rank
1. Germany English 5 18
2. United Kingdom English 13 21
3. Netherlands English 10
4. Sweden English 7 12
5. Russia Russian 49 69
6. France French 24 23
7. Italy Italian 28 28
8. Spain Spanish 26 32
9. Ukraine Ukrainian 88 136
10. Poland Polish
11. Romania Romanian 52 65
12. Belgium Dutch And French 17 19
13. Greece Greek 31 44
14. Czechia English
15. Portugal Portuguese 41 39
16. Hungary Hungarian, English 45 58
17. Belarus Russian 53 97
18. Austria German 20
19. Serbia Serbian 67 89
20. Switzerland German 2 01
21. Bulgaria Bulgarian 51 76
22. Denmark Danish 11 09
23. Finland Finnish And Swedish 15 17
24. Slovakia Slovak 38
25. Norway Norwegian 1 02
26. Ireland Irish 4 10
27. Croatia Croatia 46 61
28. Moldova Romanian And Russian 112 133
29. Bosnia And Herzegovina Croatia 77 95
30. Albania Albanian 68 105
31. Lithuania Lithuanian 35 48
32. North Macedonia Macedonian 80 99
33. Slovenia Sloven 25 36
34. Latvia Russian And Latvian, Ukrainian 41 50
35. Estonia Estonian 30 42
36. Montenegro Montenegrin 50 77
37. Luxembourg English French And German 21 5
38. Malta English 29 35
39. Iceland Icelandic 6 6
40. Andorra Spanish 35
41. Monaco Monaco
42. Liechtenstein Melty Language 17 ----
43. San Marino Italian, English Na
44. Holy see Latin Na
Countries having a medium of instruction in the vernacular languages have good HDI and
GNI rank in the world but those having English as medium of instruction in the higher
education have lower GNI and HDI rank in the world. Thus, it is perceived from the above
analysis that the medium of instruction in higher education directly or indirectly affects the
positive on Human development and National Economy.
The countries from this continent ruled the world till 19th century and due to industrialization,
this country gained economic power in the world. So, they achieved a privilege of high
development in early stage. Thus, these countries come under the category of developed
countries. Though the UK’s political and military power was crucial in the 19th and early
20th centuries, this established English as dominant over French in the United States; and
then – as the UK’s empire shrank in the 20th century – rapidly growing American global
influence gave the language a momentum perhaps unique in modern history.
1. The countries which left uncolonized has no spillover effect on them. It is easy for those
countries to develop on the basis of their norms, value system, rituals, moral and
knowledge. They select their native language as official language in all place. So, their
success is because of local language as people can understand, communicate, write and
read easily and hence it led to increase growth of the country.
2. Due to industrialization people migrated from rural area to urban area, so they have easy
availability of social and economic overheads.
3. They believe that the UK must continue to invest in English for the benefits,
opportunities and value it brings to our trade, our culture and our people.
3. Africa continent
Figure 3. Medium of instruction in various country in Africa

Table A-3. Description of various categories of class used for study of data of Medium of instruction in
Colour Medium of instruction HDI and GNI rank No. of countries
Red English Low Scale 17
Yellow English and French Low Scale 19
Blue Portugal language Low Scale 04
Green Vernacular language Low Scale 12
Table 1.3. African countries GNI and HDI rank compare with its medium of instruction
Sr.
Country MOI HDI Rank GNI Rank
No.
1. Nigeria English 157 147
2. Ethiopia English 173 174
3. South Africa English 113 187
4. Kenya English 142 153
5. Tanzania English 154 166
6. Namibia English 129 101
7. Botswana English 101 ------
8. Lesotho English 159 160
9. Gambia English 174 177
10. South Sudan English 187 184
11. Rwanda English 158 175
12. Eritrea English ------ ------
13. Liberia English 181 182
14. Sierra Leone English 184 184
15. Zimbabwe English 156 150
16. Mauritius English 65 63
17. Swaziland English 144 -------
18. Cameroon English And French 151 157
19. Senegal French 164
20. Codeliveries French Na 154
21. Morocco French 123 131
22. Chad French 186 178
23. Somalia English And Somali 158 ------
24. Niger French 189 190
25. Burkina Faso French 183 179
26. Seychelles English, French 62 54
27. Congo French 176 82
28. Burundi French 185 192
29. Togo French 165 180
30. Benin French 163 170
31. Guinea French 175 171
32. Equatorial Guinea French 141 84
33. Congo French 176 185
34. Djibouti French 172 141
35. Central African Republic French 188 ------
36. Comoros French and Arabic 165 162
37. Uganda Ugandan 162 181
38. Algeria Arabic 85 116
39. Tunisia Tunisian Arabic 95 126
40. Libya Arabic 108 90
41. Sudan Arabic 167 156
42. Mauritania Arabic 159 165
43. Egypt Arabic 115 135
44. Angola Angolan 147 128
45. Mozambique Portuguese 180 188
46. Ghana Ghanaian 140 142
47. Madagascar Malagasy 161 188
Sr.
Country MOI HDI Rank GNI Rank
No.
48. Zambia Zambian 144 -----
49. Guinea-Bissau Portuguese 177 176
50. Cape Verde Portuguese 125 ---------
51. Sao Tome and Principe Portuguese 143 -------
Source: Languages of the world – Nation online project/Human Development Report Office, 2018/Gross
national income per capita 2018, Atlas method and PPP.
Countries having a medium of instruction in the vernacular language have low HDI and
GNI rank in the world but countries having English as a medium of instruction in the higher
education have also lower GNI and HDI rank in the world. Thus, it is observed from the
above analysis that the medium of instruction in higher education, directly or indirectly
affects the Human development and National Economy.
“In 19th century, European countries colonized almost all Africa. Most present countries
in Africa emerged from a process of decolonization in the 20th century, between
increasingly powerful European countries as a key underlying cause for the Scramble for
Africa. Once one nation conquered a territory abroad, others felt compelled to respond or
they would lose power and prestige.
Some African countries had to fight for independence because Europeans had settled in
certain lands and didn't want them to gain independence because they wanted to use their
natural resources. It helped European colonial governments to gain raw materials from
Africa like crops and minerals. Europeans obtained market for their manufactured goods
from Europe.
In order to develop economic activities in Africa, colonial governments-built infrastructure
like roads, railway houses” (May 12, 2017).
Due to this, there is negative spillover effect of European countries on Africa. As like other
continents they also adopted the colonial lifestyle. So, elite section of African society
prefers European life style and hence other section of society also follow them and
government also use English, Spanish and French as official language at every place.
So, levels of education in Africa are comparatively low creating a considerable skill gap
among youth at working age. According to the African Development Bank, 25% of African
youths are still illiterate and despite a rise in primary school enrolment from 60% in 2000
to 77% in 2011, the issue of low skills levels in the workforce will continue to be a problem.
This all problems due to colonial language use in higher level of education.
4. North American continent: Medium of instruction in various country

Figure 4. Medium of instruction in various country in North America

Colour Medium of instruction HDI and GNI rank No. Of countries
Red English Data not available 15
Blue Spanish Low Scale 10
Yellow French Data not available 01
Green Multi language Low Scale 10
Table 1.4. North American countries GNI and HDI rank compare with its medium of instruction
Sr.
Country Moi HDI Rank GNI Rank
No.
1. United States American English 13 07
2. Kitts And Nevis English Na ------
3. Kingdom Of Netherland English 10 ------
4. Turks and Caicos Island English Na 33
5. Martin English
6. Virgen Island English Na -----
7. Anguilla English 70 ------
8. Barthelemy English Na ------
9. States Virgin Islands English Na -----
10. Grenada English 75 70
11. Montserrat English Na ------
12. Cayman Island English Na ------
13. Martinique English Na ------
14. Trinidad And Tobago English 69 51
15. Belize English 106 108
16. Mexico Spanish 74 74
17. Dominican Republic Spanish 94 82
18. Honduras Spanish 133 140
Sr.
Country Moi HDI Rank GNI Rank
No.
19. Salvador Spanish ------- ------
20. Guatemala Spanish 127 109
21. Cuba Spanish 73 ------
22. Costa Rica Spanish 63 64
23. Panama Spanish 66 59
24. Antigua And Barbuda Spanish Na 52
25. Dominica Spanish 103 83
26. Pierre And Miquelon French Na ------
27. Haiti French, Creole 168 173
28. Nicaragua Bilingualism 124 143
29. Puerto Rico Spanish, English --- 41
30. Jamaica Jamaican Creole And English 97 103
31. Guadeloupe French, Creole Patois Na ------
32. Bahamas English And Creole 54 31
33. Barbados English Or French 58 55
34. Canada English Or French 11 20
35. Saint Lucia English, French Potosi 90 ----
36. Bermuda English and Portuguese Na -----
37. Greenland Danish Na ------
Source: Languages of the world – Nation online project/Human Development Report Office, 2018/Gross
national income per capita 2018, Atlas method and PPP.
Countries having a medium of instruction as the English and Spanish language have low
HDI and GNI rank in the world but those countries having multi language as medium of
instruction in the higher education also have lower GNI and HDI rank in the world. Thus,
it is observed from the above analysis that the medium of instruction in higher education,
directly or indirectly affects negatively on Human development and National Economy.
5. South America continent: Medium of instruction in various country
Figure 5. Medium of instruction in various country in South America

Table A-5. of various categories of class used for study of data of Medium of instruction in higher education
for Table 1.5
Colour Medium of instruction HDI and GNI rank No. Of countries
Blue Spanish Low Scale 10
Yellow French Data not available 01
Table 1.5. South American countries GNI and HDI rank compare with its medium of instruction
Sr. Country Moi Hdi Rnk Gni Rank
No.
1. Colombia Spanish 90 91
2. Peru Spanish 89 88
3. Chile Spanish 44 57
4. Argentina English 47 62
5. Venezuela English 78 -----
6. Ecuador English 86 92
7. Bolivia English 118 ------
8. Falkland Island English Na ------
9. Uruguay English 55 53
10. Guyana English 125 106
11. Paraguay Spanish, Guarani 110 96
12. Brazil Brazilian Portuguese 79 75
13. Suriname Dutch 100 103
14. French Guiana French 125 ------
Countries having a medium of instruction in the English and Spanish language have low
HDI and GNI rank in the world but countries having multi-language as medium of
instruction in the higher education have also lower GNI and HDI rank in the world. Thus,
it is observed from the above analysis that the medium of instruction in higher education,
directly or indirectly affects the Human development and National Economy.
Colonialism brought to Latin America the quick spread and influence of Christianity into
the land, replacing traditional religions. European languages like Spanish, English,
Portuguese, Dutch, and French have been introduced to the native people. However, the
economic and social effects of colonialism are clear. I think that the three most important
long-term consequences of Columbus's encounters with the Americans were slavery,
spread disease through the Columbian exchange, and new rivalries in Europe. The early
English settlers ran into many troubles such as, poverty, food shortage, and disease.
6. Australia continent: Medium of instruction in various country

Figure 6. Medium of instruction in various country in Australia

Colour Medium of instruction HDI and GNI ranks No. of countries
Table 1.6. Australian countries GNI and HDI rank compare with its medium of instruction
Sr.
Country Moi Hdi Rank Gni Rank
No.
1. Australia English 3 13
2. Nauru English 66
3. Tonga English 98 110
4. Tuvalu English 100
5. Fiji English 92 93
6. Kiribati English 134 130
7. Marshall Island English 106 107
8. Micronesia English 131 --------
9. Palau English 60 49
10. Papua New Guinea English 153 137
11. Samoa English 104 114
12. Saolamon Island English 152 146
13. Vanuatu English And French 138 134
14. New Zealand English, Kura Kaupapa 16 -----
Countries having medium of instruction in the English language has low HDI and GNI rank
in the world but countries having multi-language as medium of instruction in the higher
education have also lower GNI and HDI rank in the world. Thus, it is observed from the
above analysis that the medium of instruction in higher education, directly or indirectly
affects the Human development and National Economy.
Result and discussion
“In the world, people are desperately trying to learn English – not simply for self-improvement,
but as an economic necessity.” It’s easy to take for granted being born in a country where
people speak the lingua franca of global business, but for people in emerging economies of
various continent like Asia, Africa and countries like China, Russia, Brazil and India where
English is official language, bad English is a critical tool, which people rightly believe will not
help them tap into new opportunities at home and abroad.” From above study researcher
found that vernacular language as a medium of instruction in the country lead to positive
impact on their economy and their HDI and GNI ranks, both are high except few countries
of the world. As in the case of United Kingdom, the English language is perhaps the greatest
and yet least-recognized international asset. It is a cornerstone of UK's identity and it keeps
UK in the mind of hundreds of millions of people around the world, even when they are
not talking to UK. Thus, the people and government of UK always seeks to get most out of
the language. They strategically think that by improving the standard of English they will
always prosper. For the nation their approach is that The UK must continue to invest in
English for the benefits, opportunities and value it brings to their trade, their culture and their
people.
In developing countries and under developed country still not growing because of under-
utilization of human resources. The Productivity growth is widely regarded as the main source
of welfare and economic prosperity. Over the last fifty years, economic literature has identified
various sources of productivity growth in an attempt to understand why countries grow at a
different rate. Historically, developed nations followed a strategy of physical and human
capital deepening in stimulating growth and higher levels of per capita income. As countries
approach the international technological frontier, to remain in a high growth trajectory they
must invest in the generation of new knowledge and ideas through R&D. Investment in R&D
is the main source of knowledge accumulation that vastly contributes to productivity growth
at industry level, although human capital has been considered to disentangle productivity-
raising innovation in aggregate level studies.
Recommendation
1. We should give more emphasis on human capital through education. At present, in the
under developed countries more than 50% population are illiterate. By providing education
to them literacy rate will increase. So, we should provide free and compulsory education
to all of them and it should be in mother tongue or vernacular language. Because use of
local language in education will create a smooth path for economic development.
2. If education is given in language other than the vernacular/local language, then students of
the first generation of the family getting educated will not find proper guidance and help
from the family and society. And at longer run this creates an economic burden on the
nation because the resource provided by the nation goes in vain.
3. It is observed that developing and under developed countries of the world provides the
higher education in the colonial language or in the English, which makes people of the
country to make extra efforts to learn foreign language. While non-colonial countries and
developed nation use their native language in the higher education and as well as in official
language. So, they get huge amount of outcome from investment made on HRD, because
at the end they become productive labor force for the country. In developing countries,
expenditure made on education is not considered as important investment because they
understand that it will not produce anything after completing study.
Flow Diagram 1
This model lead to poor economic growth because it directs the nation towards reducing
productivity.
Developing Nation At Higher Education Student don’t learn For examination
with Medium of Students do not basics and concept student mug up
Instruction – English understand the and hence couldn’t the concept or
concept and hence
or Colonial implement it in real prepare important
don’t get wisdom
Language life situations. concept ony.
For degree
Number of students Rejected from Job At the time of Job
certificate
in the unemployment interview due to lack interview student fail
students have to
category rises of communication to express in English
pass exam only.
everyday skill in English. due to shallow
No research or
knowledge
project activity
This led to the Unemployment, This hurdles to It becomes difficult

imbalance in poverty, illiteracy, growth increases for developing
economical growth of corruption and population pressure country to become
the nation various other developed country.
hurdles to growth.
The above model is the check for the developing country. It is observed that medium of
instruction in education directly affects the development of nation and generation of
productive labor force. For a developing country, youth is a future. Nation makes huge
investment in terms of time, money and resources for youth. The primary aim is to add value
and skills in the lives of people so that in return they add to the economy. But here a typical
case of English medium education directs the nation towards poor economic growth and
reduced productivity. The colonial languages are still hurdling in the growth of the nation.
Sometimes it happens that students from English medium education will produce productive
labor, since they will be able to interact at global platform. But they are not more than any
machine. They will be having information and subject knowledge but they don’t understand
how to implement it and hence they don’t add value to the economy.
Flow diagram 2
This model helps to economic growth because it directs the nation to increasing productivity.
Developing Nation At Higher Education Student learn basics In examination
Medium of Instruction Students understand and concepts and student perform
– Vernacular the concept and thus able to well and hence get
Language hence get wisdom
implement it in real good results too.
life situations.
Creates Selected in In job interview due For degree student

employment interview due to to vernacular take initiative and
opportunities in good command on language students creates innovative
various sector of language with are able to express project.
economy. communication and knowledge.
other skill.
It helps to reduce Reduce the Reduce in poverty and Developing country

other problems in corruption, Illiteracy illiteracy, reduces becomes developed
the economy. and poverty. Out population pressure. country.
flow of best brain.
Language is an indispensable tool for the communication with the society. Language has
huge economic value as it supports trade, provides a significant competitive advantage and
delivers a growing market. It is a critical component of trust building and, in turn, trade and
prosperity. According to the model shown above the students studying in vernacular
medium will understand the concepts easily and will be able to implement it. Thus, literacy
rate increases which indirectly aids to the nation growth as they reduce number of problems
arising in the country. Youth at work supports the economic growth of the country. It results
in the decrease of poverty and corruption. All the above discussion ends with the positive
results that developing country can become developed country easily.
Scope for further study

As economic development depends upon natural resources, physical capital and
technology, the growth could be measured on the bases of any of the above or developing
country.
Limitation: The objective of this research was restricted to only one factor of economic
development that is development through human resource.
Conclusion
English has developed as a global language for a range of reasons, many of them historical,
rather than anything intrinsic in the language itself. The acceptance of English language
over the globe, has made it language of primary importance. It has also proven to be an
effectively growing economic activity. The countries where the language is originated are
having higher HDI and GNI ranks in the world because of the growth measures are on high
scale. Thus, it can be directly derived that the growth of the country whether it is economic
or social depends upon the medium of instruction. So, in the case of colonial countries
where the medium of instruction is colonial languages, official languages are colonial
languages and code of conduct is colonial languages are facing difficulties in the growth
and development. The oppressed language is burden and measure hurdle in every activity.
The colonial countries may be announced independent but the education, government and
other measure conduct are still not free. Thus, from the secondary data obtained following
conclusions can be drawn:
 The countries which have medium of instruction in their vernacular language have
higher scope and opportunities for development and thus the standard of growth and
development is growing significantly. As a result, the HDI and GNI rank are on high
scale.
 The countries where oppressed language is still used on the primary bases are facing
difficulties in the growth as the society as a whole couldn’t function to add value to the
economic growth. Despite of trying to develop by employing various means they are
facing difficulties in raising the standard of living. The HDI and GNI ranks are low and
miser.
 It is also seen that the countries where vernacular languages are used have less
development but that is due to the fact that those countries are lacking in resources and
utilities. The growth rate may be slow but quite significant.
References
Arcand, J.-L. and Grin, F., 2013. Language in economic development: Is English special & is
linguistic fragmentation bad, Accessed at <https://www.researchgate.net/publication/
265185653_Language_in_economic_development_Is_English_special_and_is_linguistic_fra
gmentation_bad>
Bournakis, I., Christopoulos, D. and Mallick, S., 2015. Knowledge Spillovers, Absorptive Capacity
and Growth: An Industry-level Analysis for OECD Countries, Accessed at
<https://mpra.ub.uni-muenchen.de/63542/>
British Council, The English Effect, British Council 2013/D096, Accessed at

<www.britishcouncil.com>
Dougherty and Jorgenson (1996), and McAdam et al. (2010). Productivity growth is widely regarded
as the main source of welfare and economic prosperity.
Gross national income per capita 2018, Atlas method and PPP, Accessed at
<https://databank.worldbank.org/data/download/GNIPC.pdf>
Human Development Report Office, 2018. Accessed at <http://hdr.undp.org/en/2018-update>
Laitin, D. and Ramachandran, R., 2014. Language Policy and Economic Development, Accessed at
<https://pdfs.semanticscholar.org/a47b/426c668a943720afa41e392181a61359d981.pdf>
Languages of the world–Nation online project. Accessed at <https://www.nationsonline.
org/oneworld/african_languages.htm>
List of official languages by country and territory, Accessed at <https://en.Wikipedia.org/wiki/
list_of_official_languages_by_country_and_territory>
McCormick, C., 2013. Countries with Better English Have Better Economies, Accessed at
<https://hbr.org/2013/11/countries-with-better-english-have-better-economies>
McCormick, C., 2017. The link between English and economics, Accessed at
<https://www.weforum.org/agenda/2017/03/the-link-between-english-and-economics>
Mkwinda-Nyasulu, B. 2013/2014. Role of language in socio-economic development: the semiotics
are right, Accessed at <https://www.ajol.info/index.php/jh/article/viewFile/ 151922/141518>
Institutions and economic growth:

A comparative analysis of developing
and developed countries based on institutionalized
social technologies index
Danish Ahmed SIDDIQUI

Karachi University Business School, University of Karachi, Pakistan
daanish79@hotmail.com
Qazi Masood AHMED
Institute of Business Administration, Karachi, Pakistan
qmasood@iba.edu.pk
Abstract. This paper used index of institutionalized social technologies (lIST) developed by Siddiqui
and Ahmed (2018) as a proxy of institutions quality, to analyse their impact on Economic growth.
This index was made up of two sub-indices namely Risk reducing technologies and Anti Rent seeking
technologies. The cross sectional analysis covered 141 countries. The values are taken as average
of the period of 1990 to 2008. The effectiveness of these indices is tested in growth models along
with other variables such as inflation, human capital, savings and trade. It also factors in initial
conditions to measure signs of convergence. OLS and GMM based methodologies are employed for
estimation. The findings are robust and consistent. Institutional quality is positively associated with
economic growth suggesting that institutional reforms such as minimizing corruption, strengthen
property rights, improving law and order conditions, and enforcing contracts can leads to economic
growth. Moreover, the impact of institutions on growth is more pronounces in high income
countries, showing institutions are also crucial for sustenance of growth. Among the two forms of
institutions, the risk reducing technologies affect economic growth considerably more than anti-rent
seeking technologies. The results also confirm conditional convergence as predicted in the modern
theories of growth. Overall, these results suggest that effectiveness of institutional reforms would
depend upon how these institutions impact growth.
Keywords: macroeconomic stability, GMM technique, institutionalized social technologies.
JEL Classification: O1; O4; O5; C8.

310 Danish Ahmed Siddiqui, Qazi Masood Ahmed
1. Introduction
Some Theoretical contributions on the impact of institutions on welfare are discussed as
under. Schrag and Scotchmer (1993) traced the origin of crime and found that it’s not
exogenous nor any link found for moral characters. It is endogenously determined by
enforcement parameters of the society. Weingast (1995) narrates the historical accounts of
institutional change and claims that federalism, a self-govern system with lesser chance of
confiscation, played an impressive role in the economic growth of England and United
States in 19th and 20th centuries. Murphy, Shleifer and Vishny (1991) showed that
Occupational choice between innovative sectors and predatory sectors would depend upon
their relative returns to talent.
In countries where rent seeking has a higher reward for talent like in law sectors,
institutions would grow less as compared to countries rewarding high in innovative sectors
like engineering. A similar study conducted by Acemoglu (1995) proves that rent-seeking
activities create a negative externality on productive activities. Hence agents’ relative
reward structure is endogenously determined by rent seeking, causing multiple
equilibriums, and could let the economy trap in low level steady state equilibrium. In an
interesting study, Grossman and Kim (1996) based on inter temporal general equilibrium
model, with a predator and a prey dynasties, incorporated institutions in standard growth
model. Each dynasty has a choice to allocate inherited wealth in consumption, production
or defense and offence. Productive allocation would lead to more wealth but would, at the
same time, increase the risk of confiscation of wealth by predators. The choice of prey
would be based upon the predator’s current wealth. If it is smaller than the prey, then prey
chooses production and deter predation. However, wealth redistribution takes place
overtime as predator eats prey resources, the risk of expropriation increases to resource
allocation in defence. Overtime, all the wealth would be spent in this regard. Inter temporal
effect would be similar and prey would continue to allocate wealth to defence even though
their property is secure and production became preferable. This way, they accumulate
capital more slowly than their previous generations. Magee, Brock and Young (1989)
explore the potential of well organized rent seeking groups to exploit poorly-organized
groups by influencing irrational state through lobbying. Political parties respond to such
efforts increasing its relative benefits. The result would be that the whole economy being
captured in unproductive lobbying activities thereby creating an economic black hole. In
another entrepreneurial study, Baumol (1993) asserts that entrepreneurs could increase
technologies adoption, innovation and enable sustainable growth. However, they sometime
tend to misallocate like end mergers and acquisitions or lawsuit that could increase their
profits without being productive. Hence institutional arrangement altering the incentive
structure is needed for better utilizing entrepreneurial resources. At another place, Baumol
(1990) points out that only those firms survive that have informational asymmetries over
their competitors and the firms’ choice between technological innovations, rent seeking
and organized crime would depend upon the relative gains from these asymmetries. Strong
institutions decrease the informational asymmetries from rent seeking and organized crime,
hence only venue for survival is creating such asymmetries through innovation. Gould and
Institutions and economic growth 311
Gruben (1996) specifically focus on the impact of institutions enforcing intellectual

property rights (IPR) protections on economic growth. They find IPR protection reduce
imitation probability, increasing its rewards. However, it could also decrease competition
making innovation difficult. They argue that imperfect protection. As compared to perfect
protection, is more suited for growth. In another study, Cozzi (2001) proves that lower
protection of property rights inhibits inventions as it would result in spillover effect before
inventor can reap its rents. This could cause other to spend time searching for others’
inventions rather making their own.
Olson (1993) created a model showing that any form of government is better than anarchy.
He asserts that “Under anarchy, uncoordinated competitive theft by ‘roving bandits’
destroys the incentive to invest and produce, leaving little for either the population or the
bandits. Both can be better off if a bandit sets himself up as a dictator – a ‘stationary bandit’
who monopolizes and rationalizes theft in the form of taxes”.
In that case, he could benefit from taxes that increase in proportion to increase in
productivity. Hence it’s his interest to provide public goods that influence growth, but if he
feels he is likely to be surpassed by others, then he confiscates those assets that value more
than their expected rents that he could collect.
An interesting study by Acemoglu and Verdier (1998) points out that there are cost of
enforcement. In perfect enforcement, the costs out-weigh the benefits, hence it may be
optimal to have an intermediate level of enforcement and allowing some corruption.
Society faces a trade-off between investing in production or regulation. However, there
could be case free lunch where it is possible to simultaneously reduce corruption by
increase in public sector pay, increase investment in private sector by making it more
profitable and also improve the allocation of talent. This could be so because, better ex
ante property rights through high pays in public sector, would lead to higher profitability
and investment in private sector developing about allocation of talent of private sector in
productive activities. However, an interesting picture arises when past level of property
rights determines the present level of productivities that in turn determine the future level
of property rights. This would make initial low property rights countries fall in ‘low
property right-low investment trap’ as their initial low level of property right makes them
unproductive. Presently as they invest less in human capital, there would be lesser
potential benefits from future improvements in property rights, resulting in voting out
those future improvements. Resultantly, developing countries may opt for less effective
property rights.
2. Theoretical framework
In this study, the set of economically efficient institutions named “index of institutionalized
social technologies” is divided into two brands according to the way they impact growth,
based on the theoretical contribution of North (1981). The first is by reducing risk of doing
business thus preventing diversion of resources, and the second is by preventing predatory
rent seeking activities, thereby diverting resource toward innovation as the businesses will
not need to invest resources in protecting their property rights or earning predatory rents.
The following text elaborates these two brands of institutions:
2.1. Risk reducing technologies
The first component in index of institutionalized social technologies is Risk-reducing
technologies. Increased risk of expropriation and confiscation of property rights would
cause diversion of resources from productive activities to private arrangement in protecting
their rights that can be avoided by properly institutionalizing Risk reducing social
technologies. In other word, by institutionalizing Risk-reducing social technologies does
not mean collectively hiring guards by society proves to be cheaper, it rather means that no
guards are required in the first place. Index of risk reducing technologies covers the
institutions pertaining to property rights, law and order, and policy stability. Important
contributions in this area are by Gould and Gruben (1996); Cozzi (2001); Olson (1965,
1993, 1982); Baumol (1990); North (1990); Greif and Kandel (1995) and Weingast (1995).
For theoretical models of weak risk reducing institutions, see (Murphy et al., 1991; Schrag
and Scotchmer, 1993; Acemoglu, 1995; Ljungqvist and Sargent, 1995; Grossman and Kim,
1996). Studies such as (Magee et al., 1989; Murphy et al., 1991) explain further how
inadequate controls affect growth.
2.2. Anti-rent seeking technologies
Baumol (1990) points out that information asymmetry through rent seeking or organized
crime is curbed through strong institutions--so only venue left for competition and
dominance is through innovation. Hence in the setting of effective enforcement, these
asymmetries will lead to innovation as the only venue left to earn information rents.
The second, and perhaps more important dimension of Institutionalized social technology,
is Anti Rent seeking social technology. As shown earlier, the rent-seeking (behaviour) is
defined as “the socially costly pursuit of wealth transfers” (Tollison, 1997). Hence its
consequences on society are negative as a rent seeker benefits at the expense of others. This
in contrast to “creative destruction” where innovation leads to productive rents, rent
seeking diverts entrepreneurial efforts and resources from increase in productivity through
innovation to unproductive predatory activities by thriving on other people’s resources as
well as protection against diversion and expropriation. This is termed as “destructive
creations” by Mehlum et al. (2003). A vast literature can be found linking entrepreneurship,
rent seeking and growth (Murphy et al., 1991; Baumol, 1990, 1993; Acemoglu, 1995;
Acemoglu and Verdier, 1998). Rents can be divided into the institutional rents, which cover
institutions related to regulatory quality, corruption, and ease of doing business. The role
of government regulation has also got conflicting view. Some view regulation as a helping
hand (Pigou, 1938). Other views are based on political economics advocating that rent-
seeker in government would mould these regulations in their favour, extracting rents
(Shleifer and Vishny, 1998).
3. Empirical analysis
In this section, empirical analysis is based on OLS as well as GMM methodologies are
conducted to explore the institution growth nexus. This section is organized as follows:
Subsection 1) Describes data and Regression specification; 2) Estimation methodology;
3) Explains results.
3.1. Data description and regression specification
Specification of this study is based on combining growth theories such as Solow (1956),
Romer (1986) and Lucas (1988) with North (1981). Specifically, Modern growth theories
and their empirics provide the evidence of conditional convergence, where convergence is
conditional on factors some of which are related to institutions. The role of these
institutions in economic growth is explained by North in “contract theory” and a “predatory
theory” of the state. To assess these roles, we used standard growth regression framework
which mostly follow growth empiric literature, such as (Barro, 1991; Mankiw et al., 1992;
Leving and Renelt, 1992).
Δyi = β0 + β1Ii + β2Xi + єi
where i is the country єi is the error term. The economic growth Δyi is measured by GDP
per capita Growth in real terms, Ii stands for institutional variables, whereas Xi is the vector
of control variables for other determinants of growth. Index of institutionalized social
technologies and its two sub indices of Risk reducing technologies and Anti-rent seeking
technologies are used as a proxy of institutions. They are in 0 to 1 ranges where higher
values indicate better institutional quality. In empirical literature, there is no established
pattern of inclusion of control variable. The present study will be using variables pertaining
to initial conditions, macroeconomic stability, human capital, savings, and trade
The first control variable describes initial conditions. Testing the model for convergence
as described in neoclassical growth models (Solow, 1956) would result if country growth
rate exhibits negative relationship with its initial incomes. However, studies done by Barro
and Sala-i-Martin (1992) show that convergence is conditional upon other economic
variables. We tested the evidence of convergence conditional upon other factors like
institutions. In this regard, we used Real GDP per capita PPP in 1990. The expected sign
is negative.
Human capital increases the pace of innovation leading to greater growth and productivity
hence Education sub index of human development Index is used as its proxy.
Macroeconomic stability is another factor influencing growth. It is measured in this study
by GDP deflator. It is found that higher price instability would hamper economic growth,
hence possibly having a negative expected sign.
Saving impact growth through increasing investments (Modigliani, 1970, 1990; Maddison,
1992; Carroll and Weil, 1994) hence would show a positive relationship. It is assumed that
higher savings precede economic growth. It is represented by Gross Domestic Savings as
a percentage of GDP. Trade as a percentage of GDP is used as a proxy for trade
liberalization and openness. It tends to remove price distortions by targeting investment in

areas of competitive advantages, increases technology adoption and improves productive
and allocative efficiency thereby moving the economy to its production frontiers. This
produces a positive impact on growth and development (Jin, 2000; Sukar and Ramakrishna,
2002). Among notable variables not captured in the growth model is fiscal policy indicator,
which proved to be statistically insignificant and has, therefore not been included in the
regression specification.
Dependent and control variables expect education are taken from World Development
Indicators, while education index in taken from UNDP. Except for initial conditions, these
variables are expressed in terms of averages from the year 1990 to 2006 depending on the
availability of data. Table 1 gives detailed information about the variables and their data
source.
3.2. Estimation methodology
We first used simple OLS estimation and then used GMM instrumental estimation to
control for endogeneity. In growth models, there might be variables that are endogenous
determined from factors within model. For instance, if higher savings leads to growth,
higher growth might cause savings. Similarly, higher growth might be the reason for
improvement in institutional quality due to Halo effect where people from high growth
countries might rate their institutions high. For these cases, instruments are used to control
this potential problem of endogeneity. We employed GMM methodology for instrumental
variable estimation. GMM methodology is based on the method of moments (Pearson,
1893, 1895) and estimation techniques using instruments (Reiersol, 1941; Sargan, 1958;
Hansen, 1982). GMM estimation is based on the selection of parameters that satisfy the
theoretical relationship with minimum deviations. It is more robust estimation as compared
to other estimation methods like maximum likelihood as it is unaffected to the nature of
disturbance distribution. Instrumental variables should satisfy the assumption of
exogeneity. To satisfy the conditions, instrumental variables should be correlated with
explanatory variable but not with the error term. Meaning they are not impacted by the
dependent variable and can only affect the dependent variable through the explanatory
variables. Test of over identifying restrictions is used to assess the Validity of this
assumption.
In GMM estimation, along with any strictly exogenous variable like initial GDP, initial
values of control variables (their respective values in year 1985 for education and 1989 for
the rest), other than institutions are taken as instruments. Since institutional indices are not
in time series, we used legal origin as instruments. Apart from being exogenous since these
systems were spread primarily through conquest and imperialism, legal origin have found
to effect institutions. In particular La Porta-LLSV (1998) trace differences in legal origin
through to differences in the legal rules covering secured creditors, the efficiency of
contract enforcement, and the quality of accounting standards. Table 5 presents regressions
to find the impact of legal origin on institutions. Explanatory variables are dummy variables
for English, French, German and Socialist legal origin, relative to Scandinavian origin
(which is captured in the constant). We also control for initial level of per capita GDP. The
results show that cross country variation in institutions could largely be explained by their
legal origin hence they can be used as instruments for institutions in growth regression. The
result also shows that institutions are strongest in Scandinavian legal origin, followed by
German, British and socialists, French legal tradition tend to have less well developed
institutions.
3.3. Estimation results
The impact of institutions on economic growth is tested in two sets of equations. The first
set displayed in Table 4 estimates growth functions using OLS methodology on 130
countries while the other set in Table 6 estimates using GMM instrumental variable
technique. The results in first set in Table 4 remain consistence.
These results clearly indicate a robust positive impact of institutional index (IIST) on
growth. Initial GDP per capita shows expected negative sign and is highly significant. This
clearly indicates the sign of convergence as proposed in growth theories. Negative sign
shows the countries with lower initial GDP have experienced higher growth rate and
possibility of catching up. Among other variables, coefficient of savings also remains
positive and significant, clearly showing that saving is instrumental to growth as it increases
capital accumulation and investments. Inflation with expected negative sign is highly
significant at a 1% level in all models. This suggests that macroeconomic instability has a
negative effect on economic growth. Hence, pursuing policies of inflation financed growth
might not be fruitful in long run. Human capital, measured by education, was found to be
statistically significant and positive. This indicates investing in human capital would
produce a positive impact on growth, as it increases workers’ quality and ultimately
increase productivity.
Equations 1-3 in Table 4 estimates the impact of institutional quality on economic growth.
Three indices tested separately for institutional quality. This includes composite index of
institutionalized social technology (IIST) and two of its sub indices are index of risk
reducing technologies and rent seeking technologies. All three are positive and highly
significant at a 1% level. In a simulation, we included all three indices in one equation. It
was witnessed that when used with other institutional variables, their significance
decreased considerably probably because of high multicollinearity among these variables.
Due to this fact, we used them separately in three equations. This can also be witnessed by
extremely high correlation coefficients among institutional variables of about 0.97,
indicating that different institutional measures have high common factors on which these
measures are dependent. Overall, these finding prove robust relationship between
institutions and growth. Their estimates are large showing that marginal improvement in
institutional qualities would produce huge impact on growth. Among institutions,
coefficient of risk reducing index (4.62) is comparatively higher and more significant than
the coefficient of anti rent seeking index (2.72) this suggests that risk reducing institutions
produce comparatively larger impact on growth. However these results contradict earlier
results by Acemoglu and Johnson (2005) that proves anti-rent seeking institutions could
impact growth more. These institutional indices are comparable as they all fall within a
similar range, i.e., between 0 and 1.
Other three equations from 4-6 in Table 4 test the regional effect of institutions on
economic growth. Equation 4-6 partitioned institutions indices into High income countries
and Developing countries using dummy variables. Institutions index are multiplies by one
if the country belongs to high income group and zero if not. Second variable does the same
for nation belonging to developing countries i.e. multiply by one if they are developing
nations and zero if they are not.
These regressions clearly show that the magnitude of coefficients for high income countries
is larger and more significant. Anti rent seeking index became insignificant for developing
countries. This shows that when everything else remains the same, institutions would
impact economic growth more in high income countries as compared to developing
countries. This means good quality institutional are not only instrumental for making
nations prosper, it also be able to sustain growth in long run as high income countries with
better institutions growth comparatively more.
Regression results in Table 6 are estimated using GMM methodology to remove biases
caused by endogeneity. Models 10-12 test using basic information sets. Dependent variable
GDP per capita is regressed with institutions along with saving and initial conditions as
independent variables. These results are consistent with our earlier findings. Institutions
produce positive and significant impact on growth. These results also remain consistence
in Model 13-15 where certain other variables like trade, Inflation and education are added
to basic model.
Test of over identifying restrictions (OID) is performed to assess the validity of
instruments. Under the null hypothesis that the over identifying restrictions are satisfied.
P-value of OID show fails to rejects null hypotheses hence instruments are appropriate and
results are credible.
These results in Tables 4 and 6 are also seemed to be robust to serial correlation. Different
sets on countries are used in different regression from 130 countries used in OLS to 78
countries using GMM full information set.
4. Conclusion
This paper used index of institutionalized social technologies (lIST) developed by Siddiqui
and Ahmed (2018) as a proxy of institutions quality, to analyse their impact on Economic
growth. This index was made up of two sub-indices namely Risk reducing technologies and
Anti Rent seeking technologies. The cross sectional analysis covered 141 countries. The
values are taken as average of the period of 1990 to 2008. The effectiveness of these indices
is tested in growth models along with other variables such as inflation, human capital, savings
and trade. It also factors in initial conditions to measure signs of convergence. OLS and GMM
based methodologies are employed for estimation. The findings are robust and consistent.
Institutional quality is positively associated with economic growth suggesting that
institutional reforms such as minimizing corruption, strengthen property rights, improving

law and order conditions, and enforcing contracts can leads to economic growth. Moreover,
the impact of institutions on growth is more pronounces in high income countries, showing
institutions are also crucial for sustenance of growth. Among the two forms of institutions,
the risk reducing technologies affect economic growth considerably more than anti-rent
seeking technologies. This clearly contradicts the earlier result of Acemoglu and Johnson
(2005) that proves otherwise. The other control variables showed that human capital, physical
capital and international trade have significant impact as predicted by theory. The results also
confirm conditional convergence as predicted in the modern theories of growth. Overall,
these results suggest that effectiveness of institutional reforms would depend upon how these
institutions impact growth.
References
Acemoglu, D., 1995. Reward structures and the allocation of talent, European Economic Review 39:
pp. 17-33.
Acemoglu, D. and Johnson, S., 2005. Unbundling Institutions, Journal of Political Economy 113:
pp. 949-95.
Acemoglu, D. and Verdier, T. 1998. Property Rights, Corruption and the Allocation of Talent: A
General Equilibrium Approach, Economic Journal, Vol. 108 (September): pp. 1381-1403.
Barro, R.J. 1997. Determinants of economic growth: a cross-country empirical study, Development
Discussion Paper No. 579, Harvard Institute for International Development.
Baumol, W.J. 1993. Entrepreneurship, Management, and the Structure of Payoffs. Cambridge, MA:
MIT Press.
Baumol, W.J. 1990. Entrepreneurship: Productive, Unproductive and Destructive, Journal of
Political Economy 98(5): pp. 893-921.
Carroll, Christopher D. and David N. Weil. 1994. Saving and Growth: A Reinterpretation, Carnegie-
Rochester Conference Series on Public Policy 40: pp. 133-192.
Cozzi, G. 2001. Inventing or spying? Implications for growth. Journal of Economic Growth 6:
pp. 55-77.
Gould David M. and P. Gruben. 1996. The Role of Intellectual Property Rights in Economic Growth,
Journal of Development Economics 48(2): pp. 323-350.
Greif Avner and Eugene Kandel. 1995. Contract Enforcement Institutions: Historical Perspective
and Current Status in Russia. In Economic Transition in Eastern Europe and Russia: Realities
of Reform. ed. Edward P. Lazear. Stanford, CA: Hoover Institution Press: pp. 292-321.
Grossman, H.I. and Minseong. K., 1996. Predation and Accumulation, Journal of Economic Growth,
l (1): pp. 333-350.
Hansen, L.P., and Singleton, K.J., 1982. Generalized Instrumental Variables Estimation of
Nonlinear Rational Expectations Models, Econometrica 50: pp. 1269-1286.
Jin, Jang C., 2000. Openness and Growth: An Interpretation of Empirical Evidence from East Asian
Countries. The Journal of International Trade and Economic Development 9(1): pp. 5-17.
Levine, R. and Renelt, D., 1992. A Sensitivity Analysis of Cross- Country Growth Regressions.
American Economic Review 82(4): pp. 942-963, September.
Ljungqvist, L. and Sargent, T.J., 1995. The Swedish Unemployment Experience. European
Economic Review 39 (5): pp. 1043-1070.
Lucas R.E. 1988. On mechanism of economic planning, Journal of Monetary Economics 21(1):
pp. 3-42.
Maddison, A.1992. A long run perspective on saving, Scandinavian Journal of Economics, 94:
pp. 181-196.
Magee, S.P., Brock, W.A. and Young, L. 1989. Black Hole Tariffs and Endogenous Policy Theory,
Cambridge: Cambridge Univ. Press.
Mancur, O. 1965. The Logic of Collective Action. Cambridge: Cambridge University Press.
Mancur, O. 1982. The Rise and Decline of Nations. New Haven: Yale University Press.
Mancur, O. 1993. Dictatorship, Democracy and Development, American Political Science Review
87(3): pp. 567-576.
Mankiw, N.D., Romer P. and Weil D., 1992. A contribution to the empirics of economic growth.
Quarterly Journal of Economics 107: pp. 407-437.
Mehlum, H., Moene, K. and Torvik, R., 2003. Destructive Creativity, Nordic Journal of Political
Economy 29: pp. 77-84.
Modigliani, F., 1970. The life cycle hypothesis of saving and inter country difference in the saving
ratio. In Induction, Trade and Growth: Essays in Honour of Sir Roy Harrod. ed. Eltis, W.A.,
Scott, M.F., Wolfe, J.N. London: Clarendon Press.
Modigliani, F., 1990. Recent development in saving rates: A life cycle perspective. Frisch Lecture,
Sixth World Congress of the Econometric Society, Barcelona, Spain.
Murphy K., Shleifer, A. and Vishny, R.W., 1991. The Allocation of Talent: The Implications for
Growth, Quarterly Journal of Economics 106: pp. 503-530.
North D.C., 1981. Structure and Change in Economic history, New York, Norton.
North D.C., 1990. Institutions, institutional change and economic performance, Cambridge, MA:
Cambridge University Press, New York.
Pearson, K., 1893. Contributions to the mathematical theory of evolution. Proceedings of the Royal
Society of London 54: pp. 329-333.
Pearson, K., 1895. Contributions to the mathematical theory of evolution, II: Skew variation in
homogeneous material. Philosophical Transactions of the Royal Society of London ARRAY
186: pp. 343-414.
Pigou, A.C., 1938. The Economics of Welfare. 4th ed. London: Macmillan and Co.
La Porta Rafael, López-de-Silanes Florencio, Shleifer Andrei and Vishny Robert. 1998. Law and
Finance, Journal of Political Economy 106(6): pp. 1113-1155.
Reiersol, O., 1941. Confluence Analysis of Lag Moments and other Methods of Confluence
Analysis. Econometrica, 9: pp. 1-24.
Romer Paul, M. 1986. Increasing returns and long-run growth, Journal of Political Economy, 94(5):
pp. 1002-1037.
Sargan, J.D. 1958. The Estimation of Economic Relationship using Instrumental Variables.
Econometrica 26: pp. 393-415.
Schrag J. and Scotchmer, S., 1993. The Self-Reinforcing Nature of Crime. Working Paper
No. 93-11. Center for the Study of Law and Society, School of Law, University of
California, Berkeley.
Siddiqui, D.A. and Ahmed, Q.M., 2018. Institutionalized social technologies index: A global
perspective. Theoretical and Applied Economics, 26(4), pp. 67-96.
Solow R.M., 1956. A contribution to the theory of economic growth. Quarterly Journal of
Economics, 70: pp. 65-94.
Sukar, A. and Ramakrishna, G., 2002. The Effect of Trade Liberalization on Economic Growth: The
case of Ethiopia. Finance India, XVI(4): pp. 1295-1305.
Tollison R.D. 1997. Rent seeking. In Perspectives on public choice. ed. D.C. Mueller, Cambridge:
Cambridge University Press, pp. 506-525.
Weingast Barry R. 1995. The Economic Role of Political Institutions: Market-Preserving Federalism
and Economic Development, Journal of Law, Economics, and Organization 11 (1): pp. 1-31
(April).
World Bank, World Development Indicators (various years)
Table 1. Estimation variables' data sources and description

Variable Description Concept Coverage Source
Name Measured
1 RGDPPCG GDP per capita growth (annual %). Annual Economic 1990-2006 World
percentage growth rate of GDP per capita based on Performance Development
constant local currency. GDP per capita is gross Indicators-WDI
domestic product divided by midyear population. (2008)
GDP at purchaser's prices is the sum of gross value
added by all resident producers in the economy plus
any product taxes and minus any subsidies. (code-
NY.GDP.PCAP.KD.ZG)
2 RGDPPC90 GDP per capita, PPP (current international $). GDP Convergence 1990
is gross domestic product converted to international
dollars using purchasing power parity rates.
(code-NY.GDP.PCAP.PP.CD)
3 SAVING Gross domestic savings (% of GDP). Gross Saving and 1990-2006
domestic savings are calculated as GDP less final Investment
consumption expenditure (code-NY.GDS.TOTL.ZS)
4 INFLATION GDP deflator (base year varies by country). The Macro-economic 1990-2006
GDP implicit deflator is the ratio of GDP in current Stability
local currency to GDP in constant local currency.
(code-NY.GDP.DEFL.ZS)
5 TRADE Trade (% of GDP). Trade is the sum of exports and Openness 1990-2006
imports of goods and services measured as a share
of gross domestic product. (code-
NE.TRD.GNFS.ZS)
6 EDUCATION Education index (HDI). Human Capital 1990-2007 Human
Development
Index-HDI
(various years)
7 IIST Index Institutionalized Social Technologies Institutions Authors' own
8 Sii Index of Risk reducing Technologies Institutions calculations
9 Ri Index of Anti-Rent seeking Technologies Institutions
Table 1. Descriptive statistics

RGDPPCG IIST RI SII Education Saving Inflation RGDPPC90 Trade
Mean 1.976272 0.55797 0.566305 0.549636 0.779176 18.88961 3.97E+11 6989.239 81.48656
Median 1.846429 0.521512 0.543847 0.516084 0.848 18.98249 108.4761 4087.271 70.26988
Maximum 26.07452 0.924078 0.939835 0.918515 0.988 54.82545 5.51E+13 35780.89 417.448
Minimum -4.97836 0.058482 0.059412 0.012869 0.216 -14.6853 34.79983 245.224 3.588676
Std. Dev. 2.765753 0.186677 0.19497 0.186962 0.19022 12.11772 4.68E+12 7380.869 51.37435
Skewness 4.704483 0.059955 -0.055791 0.097168 -1.133895 0.130048 11.66221 1.388165 2.981782
Kurtosis 43.07789 2.666081 2.644591 2.600122 3.364363 3.398578 137.0072 4.532028 17.2212
Observations 139 141 141 141 136 137 139 130 137
Table 2. Correlation coefficient matrix

RGDPPCG Education Saving Inflation RGDPPC90 Trade Sii Ri IIST
RGDPPCG 1
Education 0.2826 1
Saving 0.2818 0.3986 1
Inflation -0.3339 -0.0956 -0.0928 1
RGDPPC90 0.0521 0.5938 0.4960 -0.0784 1
Trade 0.1680 0.2620 0.3867 -0.0546 0.2723 1
Sii 0.2789 0.5808 0.3776 -0.1644 0.7707 0.3393 1
Ri 0.2418 0.6567 0.2909 -0.1715 0.7010 0.2565 0.9097 1
IIST 0.2662 0.6337 0.3416 -0.1719 0.7526 0.3044 0.9766 0.9777 1
Table 4. Institutions and growth: OLS regression, 1990-2006

Dependent Variable: RGDPPCG
Model 1 2 3 4 5 6
Institutions IIST Ri Sii IIST Ri Sii
C -1.925718 -1.453419 -2.270955 -1.789766 -1.268473 -2.169534
(0.0084) (0.0395) (0.0022) (0.0187) (0.0806) (0.0049)
Institutions 4.008087 2.722701 4.620821
(0.0021) (0.0254) (0.0003)
Institutions × High Income Countries 4.10463 2.924967 4.728385
(0.0018) (0.0178) (0.0002)
Institutions × Dev. countries 3.558966 2.059017 4.312124
(0.0158) (0.1321) (0.0022)
Savings 0.048282 0.04861 0.045576 0.050358 0.051749 0.047305
(0.0006) (0.0008) (0.0009) (0.0005) (0.0005) (0.0009)
Inflation -9.71E-14 -1.01E-13 -9.57E-14 -9.91E-14 -1.04E-13 -9.7E-14
(0.0009) (0.0006) (0.0008) (0.0007) (0.0005) (0.0008)
RGDPPC90 -0.000131 -0.000108 -0.000147 -0.000148 -0.000137 -0.00016
(0.0000) (0.0003) (0.0000) (0.0004) (0.0009) (0.0002)
Education 1.927196 2.014865 2.196849 2.053523 2.245268 2.264463
(0.0422) (0.0442) (0.0146) (0.0346) (0.0287) (0.0132)
P-values of t-statistics in parentheses.
Adj. R-squared 0.28026 0.28269 0.30280 0.27687 0.25438 0.29853
F-statistic 11.04647 9.77342 12.20527 9.23193 8.33519 10.14997
Durbin-Watson stat 1.98346 1.96373 1.99150 1.98200 1.96527 1.99034
countries included: 130 130 130 130 130 130

Table 5. Legal origin and IIST, 1990-08

Model 7 8 9
Dependent Variable: IIST Ri Sii
C 0.619295 0.642296 0.596294
(0.0000) (0.0000) (0.0000)
British -0.140392 -0.149957 -0.130827
(0.0142) (0.0202) (0.0197)
French -0.182705 -0.188637 -0.176774
(0.0017) (0.0039) (0.0019)
German -0.046424 -0.053032 -0.039816
(0.5166) (0.5117) (0.5705)
Socialist -0.160933 -0.170671 -0.151195
(0.0091) (0.0142) (0.0124)
RGDPPC90 0.000015 0.0000142 0.0000158
(0.0000) (0.0000) (0.0000)
Adj. R-squared 0.592213 0.531344 0.634634
F-statistic 37.01602 26.98349 41.34019
DW stat. 2.017539 2.003731 2.065616
countries incl: 125 125 125
British = English legal origin, French = Napoleonic legal origin.
German = German legal origin, Socialist = Socialist legal origin.
Scandinavian legal origin is the omitted category.
Table 6. Institutions and growth: GMM (Generalized Method of Moments) regression, 1990-2006
Dependent Variable: RGDPPCG
Model 10 11 12 13 14 15
Information set BASIC BASIC BASIC FULL FULL FULL
Institutions IIST Ri Sii IIST Ri Sii
C -0.585179 -0.581702 -0.582298 -1.452353 -1.412784 -1.48587
(0.4892) (0.5012) (0.4833) (0.0911) (0.099) (0.0861)
Institutions 2.797519 2.651283 2.954003 3.020999 3.003407 3.036547
(0.0718) (0.0803) (0.0655) (0.0392) (0.0507) (0.0304)
RGDPPC90 -0.0000843 -0.0000801 -0.0000888 -0.000133 -0.000126 -0.00014
(0.034) (0.0362) (0.0328) (0.0000) (0.0000) (0.0000)
Saving 0.068487 0.069614 0.067061 0.060748 0.063929 0.057288
(0.0052) (0.0051) (0.0055) (0.0217) (0.0188) (0.0266)
Trade 0.002694 0.003757 0.001611
(0.4344) (0.2922) (0.6368)
Inflation -0.0000164 -0.0000154 -0.0000174
(0.1976) (0.2555) (0.1455)
Education 1.661368 1.311524 2.017376
(0.1615) (0.3326) (0.0573)
Adj. R-squared 0.220166 0.206253 0.228819 0.282594 0.239173 0.318226
DW statistics 1.790544 1.795591 1.789118 1.699283 1.705042 1.704487
J-statistic 0.034298 0.035299 0.033022 0.069047 0.069047 0.074395
Countries incl.: 114 114 114 78 78 78
OID 3.910022483 4.02409 3.7644833 5.3856948 5.000794 5.80279492
OID (p-value, 3df) (0.27134) (0.25888) (0.28804) (0.14564) (0.17174) (0.12161)
Instrument are Initial level of the explanatory variables (year 1989) except institutions and legal origin dummy variables.
Exports, imports and economic growth in India:

An empirical analysis
K. Krishna REDDY
Dr. B.R. Ambedkar Open University, Hyderabad, Telangana, India
krishnareddyou@gmail.com
Abstract. The study examines the relationship between exports, imports and economic growth in
India, spanning the period from 1980-2019. The study findings confirmed the long-run relationship
among exports, imports and economic growth and also reported that there is a unidirectional
causality from economic growth to exports; exports to economic growth; exports to imports; imports
to economic growth in the short-run. In the long-run, author confirmed that there is a bidirectional
causality relationship between economic growth and exports; and exports and imports. Finally, the
study results indicate that both exports and imports increase economic growth in India.
Keywords: India, trade, economic growth, cointegration, causality.
JEL Classification: C03, C32, F14.

324 K. Krishna Reddy
1. Introduction
In the literature, there are many studies have conducted research on international trade.
There results also inconclusive and there is a lot of scope to examine the relationship
between international trade and its related sector. International trade includes both exports
and imports, here when the countries have sufficient natural resources; those countries are
exports goods and services to foreign countries and they can get foreign remittance further
which will help to increase per-capita income of the people. Imports also increase per-
capita income when countries have insufficient natural resource, they have to import
natural resources or raw materials. Further they will use technology in the production
process and export these goods and services to foreign countries, it will improve per-capita
income. Therefore, we confirmed that international trade (exports and imports) increases
economic growth.
In the existing literature, there are many studies have reported that international trade has
a positive impact on economic growth, for example, Awokuse (2007); Erfani (1999); Vohra
(2001); Shan and Sun (1999); Sultan and Haque (2011) for different countries.
In contrast, there are few studies have reported that international trade has a negative impact
on economic growth. Moreover, the direction of causality also inconclusive, some studies
have reported that there is a bidirectional causality between international trade and
economic growth, see, Awokuse (2008); Hatemi-J (2002); Awokuse (2005). Some of them
also reported that there is a unidirectional causality between the variables, see Çetintaş and
Barişik (2009). A study by Tang (2006) could not established any long-run and short-run
relationship between the international trade and economic growth.
International trade not only increase economic growth but also create job opportunities,
increase per-capita income, improve infrastructure and increase economic scale. Based this
international trade has a positive impact on economic growth. It implies that countries have
well developmental plans, appropriate economic reforms, this will bring positive impact
on economic growth, where international trade has a negative impact on economic growth.
It implies that countries have not appropriate economic reforms, this may bring negative
impact on economic growth. Those countries have need to improve their economic policies
which is increase economic growth. The present study looks into Indian economy, whether
international trade increase economic growth or not.
Given this background, the study examines the relationship between exports, imports and
economic growth in India, spanning the period from 1970 to 2019. The applies ADF unit
test to identify the stationarity properties, Johansen cointegration test to examines the long-
run relationship between the variables and Granger causality test to identify the direction
of causality between exports, imports and economic growth.
The rest of the paper is organised as follows: Section 2 reviews literature; Section 3
discusses the data and empirical methodology; Section 4 provides empirical analysis; and
section 5 provides conclusions.
Exports, imports and economic growth in India: An empirical analysis 325
In the literature, there are many authors reported that international trade increase economic
growth. International trade has a positive impact on economic growth. For example, a study
by Kilavuz and Topcu (2012) examined the relationship between exports and economic
growth for a panel of 22 development countries during the period of 1998-2006. They result
indicate that growth of exports has a positive impact on economic growth. Awokuse (2007)
studied the causality relationship between exports, imports and economic growth for a
panel of transition economies, spanning the period from 1991:1-2004:3. His findings
reported that exports and imports increases economic growth and also shows that
international trade encourages economic growth. Ee (2016) used FMOLS techniques to
examine the relationship between exports and economic growth for Sub-Saharan African
countries covering the period from 1985-2014. Author findings confirmed that there is a
long-run relationship between the variables and also reported that exports have a positive
impact on economic growth. Gokmenoglu et al. (2015) studied the exports led growth
hypothesis for Rica during the period of 1980-2013. The study results indicated that there
is a long-run relationship between variable and also reported a unidirectional causality
relation from economic growth to exports growth. Hye (2012) examined the relationship
between exports, imports and economic growth for China covering the period from 1978-
2009. The study results indicated that there is a long-run equilibrium relationship between
exports, imports and economic growth. The empirical results also confirm the bidirectional
long run relationship between the economic growth and imports, economic growth and
exports, and exports and imports. Al-Khulaifi (2013) examined the relationship between
exports, imports and economic growth for Qatar, spanning the period from 1980-2011. The
study findings show that there is long-run relationship among exports, imports and
economic growth. There is a unidirectional causality from imports to exports. Sultan Haque
(2011) studied the relationship between exports, domestic investment and economic growth
for India during the period of 1970-71 to 2007-08. The study exists the presence of
long-run relationship between exports, domestic investment and economic growth.
Exports and investments increase economic growth.
Kibria and Hossain (2020) investigated the relationship between exports and economic
growth for Bangladesh covering the period from 1980-2018. They find a unidirectional
causality from exports to economic growth and a bidirectional causality relationship
between exports to economic growth. Raghuramapatruni and Surya Chaitanya (2020)
examined the impact of international trade on economic growth for India, spanning the
period 1991-2017. Their evidence indicated that international trade has a positive impact
on economic growth and a unidirectional causality from exports to economic growth.
Hassan (2020) studied the relationship between exports and economic growth for a panel
of two groups namely MENA and South Asian countries during the period of 1990-2018.
The findings confirmed that there is no long-run equilibrium relationship between the
variables and a unidirectional causality from economic growth to exports. Hamdan (2016)
studied the impact of exports, imports on economic growth for a panel of 17 countries,
spanning the period 1995-2013. Author findings confirmed that exports and imports have
a positive impact on economic growth. Bakari and Mabrouki (2016) studied the
relationship between exports, imports and economic growth for Turley covering the period
from 1960 to 2015. Their results showed that there is no long-run relationship between
exports, imports and economic growth and a unidirectional causality from exports to
economic growth and imports to economic growth. Bakari and Krit (2017) examined the
nexus between imports, exports and economic growth for Mauritania covering the period
of 1960-2015. Their findings indicated the presence of long-run relationship between the
variables. Exports have a positive impact on economic growth and imports have a negative
impact on economic growth while there is a unidirectional causality from imports to
economic growth. Zang and Baimbridge (2012) investigated the nexus between
imports, exports and economic growth for both South Korea and Japan during the
period of 1963-2003 and 1957-2003. They found a bidirectional causality relationship
between imports and economic growth. Baharumshah and Rashid (1999) studied the
causality relationship between exports, imports and economic growth for Malaysia
covering the period 1970:1 – 1994:4. Their empirical results reported a bidirectional
causality relationship exports and economic growth.
Given this background, in the literature, results are not uniform. Therefore, the present
study examines the relationship between exports, imports and economic growth by updated
data sets. The finding may useful for policy makers and academician.
3. Data and methodology

3.1. Data
The study uses data from 1970-2019 for India. Data on GDP per-capita 2010 constant US
dollar is used as a proxy for economic growth, exports as a percentage of GDP is used as a
proxy for exports and imports as a percentage of GDP used as a proxy for imports. The
required data is collected from world development indicator published by World Bank.
3.2. Methodology
The study examines the relationship exports, imports and economic growth. For this
purpose, author framed a model.
𝑌 𝑓 𝐸𝑋 , 𝐼𝑀 1
Where Y, EX, IM indicates economic growth, exports and imports, t is indicating time period.
The study examines the long-run relationship among economic growth, exports and
imports and long-run equation is as follows.
𝜆 𝑟 𝑇 ln 1 ̂ 2
𝜆 𝑟, 𝑟 1 𝑇 ln 1 ̂ 3
Where 𝜆 estimated characteristic and T is the number of usable observations. The 𝜆

test the null hypothesis 𝑟 0 against the alternative of 𝑟 0 and 𝜆 test the null
hypothesis is 𝑟 0 against the alternative of 𝑟 1.
The study also estimates the short-run and long-run causality between variables and
equations is as follows.
𝑌 𝛽 𝛽 𝑌 𝛽 𝑋 𝛽 𝑀 𝜆 𝐸𝐶𝑀 𝜀 4
𝐸𝑋 𝛿 𝛿 𝑌 𝛿 𝐸𝑋 𝛿 𝐼𝑀 𝜆 𝐸𝐶𝑀 𝜀 5
𝐼𝑀 𝛾 𝛾 𝑌 𝛾 𝐸𝑋 𝛾 𝐼𝑀
+ 𝜆 𝐸𝐶𝑀 𝜀 6
Where 𝑌 is the natural logarithm of economic growth at time t; 𝐸𝑋 is the natural logarithm
of exports at time t; 𝐼𝑀 is the natural logarithm of imports at time t; 𝛽, 𝛿 𝑎𝑛𝑑 𝛾 are short-
term coefficients and 𝑝, 𝑞 and 𝑟 are the log orders,𝜀 (𝑖 1,2,3) are serially uncorrelated
error terms, 𝐸𝐶𝑀 is the lagged error correction term, λ’s are the speed of adjustment
parameters.
4. Results and discussions

4.1. Stationarity
To know the stationarity properties, the study uses ADF test for economic growth, exports
and imports and results are displayed in Table 1.
Table 1. ADF test results
Variables At level First difference
t-Statistic Prob. t-Statistic Prob.
Y 8.172 1.000 -1.845 0.062***
EX 2.200 0.992 -2.926 0.004*
IM 2.146 0.991 -4.848 0.000*
Note: * and *** indicates significance at 1% and 10% level.
The study could not find stationarity at level. Therefore, the study converted economic
growth, exports and imports variables into first order difference. Then, the study confirmed
that economic growth, exports and imports variables are stationary at their first difference.
Further, all variables are stationarity at first difference.
4.2. Long-run relationship
To find the long-run relationship among the economic growth, exports and imports, the
study applies Johansen cointegration test. The cointegration results are reported in Table 2.
Table 2. Cointegration test results

Model: lnY = f (EX, IM)
Hypothesized: No. of CE(s) trace test critical values λ-max test critical values
None 32.931 29.797* 20.268 21.131*
At most 1 12.66 15.494 12.438 14.264*
At most 2 0.224 3.841 0.224 3.841
Note: * indicates cointegration at 1% level.
The study results confirmed that there is a long-run equilibrium relationship among the
economic growth, exports and imports. It implies that economic growth, exports and
imports variables are moving together in the long-run.
4.3. Direction of causality
To identify the direction of causality between the economic growth, exports and imports.
The applies Granger causality test and to estimate this method variables to be first
difference. Therefore, author converted into first order and results are reported in Table 3.
Table 3. Multivariate Granger causality results
F-Statistics for Short run causality t-statistics for long run
_____________________________________ ____________________
Dependent Variable𝑌𝐸𝑋𝐼𝑀𝐸𝐶𝑇
𝑌- 4.494 [0.016]**0.795 [0.457] -1.708(0.090)***
𝐸𝑋2.859 [0.068]* - 5.608 [0.000]***-1.674(0.096)***
𝐼𝑀3.126 [0.054]*0.993 [0.378] - -1.450(0.149)
Note: *, ** and *** indicates significance at 1%, 5% and 10% level, respectively.
The above results are confirmed that there is a unidirectional causality from economic
growth to exports; exports to economic growth; exports to imports; imports to economic
growth in the short-run. In the long-run, author confirmed that there is a bidirectional
causality relationship between economic growth and exports; and exports and imports. The
results are support the growth led growth hypothesis, exports led growth and imports led
growth hypothesis under study.
5. Conclusions
The present study examines the relationship among economic growth, exports and imports
during the period of 1970-2019. In general, both exports and imports are increases
economic growth. Given this background, the study analyzed. The results are confirmed
that there is a long-run equilibrium relationship among economic growth, exports and
imports. The direction of causality, author reported that there is a unidirectional causality
from economic growth to exports; exports to economic growth; exports to imports; imports
to economic growth in the short-run. In the long-run, author confirmed that there is a
bidirectional causality relationship between economic growth and exports; and exports and
imports. Finally, the study results indicate that both exports and imports increase economic
growth in India.
References
Al-Khulaifi, A.S., 2013. Exports and imports in Qatar: Evidence from cointegration and error
correction model. Asian Economic and Financial Review, 3(9), p. 1122.
Awokuse, T.O., 2007. Causality between exports, imports, and economic growth: Evidence from
transition economies. Economics Letters, 94(3), pp. 389-395.
Awokuse, T.O., 2008. Trade openness and economic growth: is growth export-led or import-led?
Applied Economics, 40(2), pp. 161-173.
Baharumshah, A.Z. and Rashid, S., 1999. Exports, imports and economic growth in Malaysia:
Empirical evidence based on multivariate time series. Asian Economic Journal, 13(4),
pp. 389-406.
Bakari, S. and Krit, M., 2017. The nexus between exports, imports and economic growth:
Evidence from Mauritania. International Journal of Economics and Empirical Research, 5(1),
pp. 10-17.
Bakari, S. and Mabrouki, M., 2016. The Relationship among Exports, Imports and Economic
Growth in Turkey.
Çetintaş, H. and Barişik, S., 2009. Export, import and economic growth: The case of transition
economies. Transition Studies Review, 15(4), pp. 636-649.
Ee, C.Y., 2016. Export-led growth hypothesis: empirical evidence from selected Sub-Saharan
African countries. Procedia Economics and Finance, 35(2), pp. 232-240.
Erfani, G.R., 1999. Exports and economic growth in developing countries. International advances
in economic research, 5(1), pp. 147-148.
Gokmenoglu, K.K., Sehnaz, Z. and Taspinar, N., 2015. The export-led growth: A case study of
Costa Rica. Procedia Economics and Finance, 25(1), pp. 471-477.
Hamdan, B.S., 2016. The effect of exports and imports on economic growth in the Arab countries:
A panel data approach. Journal of Economics Bibliography, 3(1), pp. 100-107.
Hassan, K.G., 2020. Empirical investigation on the relationship between exports and economic
growth in selected LDCs country groups (1988-2018). Economic Journal of Emerging
Markets, 12(1), pp. 1-12.
Hatemi-j, A., 2002. Export performance and economic growth nexus in Japan: a bootstrap approach.
Japan and the World Economy, 14(1), pp. 25-33.
Hye, Q.M.A., 2012. Exports, imports and economic growth in China: an ARDL analysis. Journal
of Chinese Economic and Foreign Trade Studies.
Kibria, M.G. and Hossain, M.S., 2020. Does export affect the Economic growth?: An empirical
investigation for Bangladesh. American Journal of Economics and Business Management,
3(1), pp. 219-226.
Kilavuz, E and Topcu, B.A., 2012. Export and economic growth in the case of the manufacturing
industry: panel data analysis of developing countries. International Journal of Economics and
Financial Issues, 2(2), pp. 201-215.
Raghuramapatruni, R. and Surya Chaitanya, R.V., 2020. An appraisal of the impact of international
trade on economic growth of India-through the ARDL approach.
Shan, J. and Sun, F., 1999. Export-led growth and the US economy: some further testing. Applied
Economics Letters, 6(3), pp. 169-172.
Sultan, Z.A. and Haque, M.I., 2011. The estimation of the cointegration relationship between
growth, domestic investment and exports: the Indian economy. International Journal of
Economics and Finance, 3(4), pp. 226-232.
Sultan, Z.A. and Haque, M.I., 2011. The estimation of the cointegration relationship between
growth, domestic investment and exports: the Indian economy. International Journal of
Economics and Finance, 3(4), pp. 226-232.
Tang, T.C., 2006. New evidence on export expansion, economic growth and causality in China.
Applied Economics Letters, 13(12), pp. 801-803.
Vohra, R., 2001. Export and economic growth: Further time series evidence from less-developed
countries. International Advances in Economic Research, 7(3), pp. 345-350.
Zang, W. and Baimbridge, M., 2012. Exports, imports and economic growth in South Korea and
Japan: a tale of two economies. Applied Economics, 44(3), pp. 361-372.
BEPS – A challenge for the increasing budget revenues

in the process of digitalization
Ionuț MIȘA
ionut.misa@yahoo.com
Meral KAGITCI
meral.kagitci@ase.ro
Abstract. In this paper we analyze the measures that led to the formation and implementation of the
action plan regarding the outsourcing profit. These measures are adopted in the European Union
and are also found in the Romanian tax code starting with 2018. The analysis is based on the action
number 1 – “Digital Economy”. In the end, we will analyze the strategy that is based on the CCCTB
(Common Consolidated Corporate Tax Base).
Keywords: fiscality, tax revenue, digitalization, tax avoidance, OECD countries.
JEL Classification: H26, H32, H87.

332 Ionuț Mișa, Meral Kagitci
1. Introduction
The fiscality must be a state attribute, but it also must take into consideration the will and
the element of consensus from the part of tax payer. The fiscality is undoubtedly necessary.
Nobody, till now, could replace this way of financing and sustaining the state.
In fiscality, there are two terms that are used by tax payers and administrative entities of
contributions, for the description of the decision taken by some tax payers in order to reduce
the taxation at minimum: tax avoidance vs. tax evasion.
Tax evasion is an illegality that includes a series of juridical consequences, while the tax
avoidance does not have any juridical consequences. The tax evasion represents violation
of fiscal legislation, while tax avoidance involves the utilization of the law gaps and acting
within the limits allowed by the law.
Due to the high rate of competition that appears in the economic sector, enterprises have
adopted strategies of fiscal planning which are exploiting the gaps in the rules regarding
fiscality in order to transfer artificially the profits to specific locations that are tax-free or
have low rates of taxation, where there does not exist or is a reduce economic activity
(through offshore companies). Thus, a system has been created that recognizes over 100
countries and jurisdictions collaborating in the implementation of the OCDE/G20 Base
Erosion and Profit Shifting package (referred to as the “BEPS package”) for multinational
enterprises.
Over 80 countries that are developing and other economies non-OCDE/non G-20 discuss
the BEPS challenges through direct participation at the fiscal business committee.
OCDE and G20 countries, together with the number of countries that participated at the
elaboration of BEPS package, establish a modern international tax framework, where the
profits are taxed when the economic activity or the created valued are identified.
BEPS package provides 15 actions that equip the governments with internal and
international instruments. These countries have now the necessary instruments for ensuring
that the profits are taxed within the jurisdictions of their economic activity that generate
the respective profits and where the added value is created. These instruments also offer a
higher security of the firms by reducing disputes regarding the application of the
international fiscal norms and standardizing compliance requirements.
BEPS actions:
 Digital Economy.
 Hybrid Arrangements.
 Controlled Foreign Companies.
 Limitation of the deduction of interest.
 Harmful tax practices.
 CEDI abuse.
 Permanent headquarters.
 Transfer prices.
 Intangibles, Risk and capital.
 Transactions with high risk.
BEPS – A challenge for the increasing budget revenues in the process of digitalization 333
 Dates.
 Disclosure of tax planning schemes.
 “Country-by-Country” report.
 The disputes resolution.
 Multilateral instrument.
In 2013, 362 out of 500 companies in Fortune were found to be transferred to tax havens
(Fortune 500 is a ranking of the largest US public companies by turnover, compiled by the
prestigious Fortune magazine). Approximately 64% of companies with any registered tax
offices have at least one in Bermuda or Cayman Islands - two notorious paradises. In
addition, the profits that all the American multinationals had, not just Fortune 500, were
earned in these islands in 2010, and accounted for 1,643% (1.60% of the country's entire
annual economic situation). Tax evasion by 30 companies with the most offshore,
collectively funded funds reaches around $ 1.2 trillion in the US. The loss due to avoiding
this tax must be offset by higher individual taxes or reductions in public investment and
public services. Therefore, this leads to a burden for the entire nation. The profitable
strategies of multinational companies raise serious issues of fairness and compliance, as
the current international tax system offers opportunities to exploit legal loopholes and
benefit from tax-free gains (Jansky and Prats, 2013). Taxes paid do not reflect the income
they have earned. In addition, finding subsidiaries in countries with tax havens is one of
the red flags of commercial money-laundering (Omara and Zolkaflila, 2015). Most tax
systems in countries operate somewhere in the spectrum between two extremes: a global
or territorial taxation. In a global tax system, corporations are subject to worldwide income
tax by their country of residence. Under a territorial system, income is subject to tax only
in the country in which it is earned. (Dowd et al., 2017).
Globalization has encouraged countries to continually assess their tax systems and public
spending to improve the “fiscal climate” for investment. Reduced taxation could initially
stimulate investment, but it will then damage all countries in the long run. There is no need
to say that tax evasion is attempted not only by multinationals, but also by individuals with
a private net worth of more than one million dollars.
In an increasingly integrated global market, national tax laws and international standards
have not kept pace with multinational companies with liquid capital and the digital
economy, the resulting gaps being exploited by tax avoidance in resident countries by
transferring activities, risks or assets to exempt or tax-free jurisdictions out of borders. This
undermines the fairness and integrity of tax systems around the world, and in particular
developing countries are devoid of an important revenue source. Moreover, these
undermines can lead to higher economic risks that can refer to the occupation of the labor
force or innovation and productivity, that can be affected if the fiscal profitability becomes
a principal stimulant for investments (Lamers et al., 2014).
The decision of companies to avoid taxes and engage in informal activities is influenced
by the policy of charging taxes, and particularly sanctions and probabilities of discovery.
Financial fraud is an incredibly dynamic phenomenon - and fraud patterns have a very short
period of validity - a simple tax system and full information on agents can reduce tax
evasion. Small businesses are particularly vulnerable to fraud because of the rare controls
they are subject to. In addition, the perception that fiscal policy is fair is associated with
low levels of tax evasion. Increased knowledge of the possibilities of tax evasion has a
negative influence on tax compliance as it contributes to its non-compliance. For tax
evasion at the international level, it is necessary to develop and implement appropriate
strategies to minimize its harmful effects. This should lead to improved tax revenue
collection by governments (Stankevicius and Leonas, 2015).
It was often believed that the existence of tax havens contributed to the financial crisis and
had a negative impact on the fiscal sustainability of the countries, even though there is no
clear evidence in this respect. All of these have contributed to the OECD/G20 BEPS
initiative on tax base erosion and the transfer of profits. This initiative does not intend to
modify the existing international standards on the allocation of taxing rights to cross-border
income sources. What it wanted to do was restart the source and residence tax on those
incomes that would otherwise have been taxed or taxed at very low rates. The rules of
foreign controlled companies (CFCs) are designed to address the shift of profits to foreign
subsidiaries that are subject to lower taxation, thereby eroding the tax base and often
delaying long-term taxation.
In addition, ineffective CFC rules are considered a key element contributing to tax base
erosion and profits transfer, especially since many countries have CFC rules, but they have
not kept pace with the changes in this environment. (Christiana HJI Panayi, 2016) The
OECD estimates that 4-10% of global income tax revenue ($ 100-240 billion annually) is
lost. An important objective of the BEPS project is therefore to provide governments with
more efficient tools to ensure the effectiveness of their sovereign fiscal policies, with other
visible goals being the correction of “distortions” that occur in commercial and investment
patterns, and the provision of conditions of fair competition between multinationals and
national companies. Another important objective of the project is to support “effective
fiscal sovereignty of countries in designing their tax systems” (OECD, 2014, p. 14). This
goal is shared by the dominant approach in philosophical literature on justice in the field
of international taxation.
In general, the literature suggests that the redistribution of revenues through the progressive
allocation of the fiscal task in countries with reduced revenues was inefficient.
For this reason, it is recommended for countries with low revenues to head towards the
expenditure side of government budget in order to achieve the desired redistribution. Since
countries with low revenues have difficulties in the implementation of the progressive
fiscal systems and increasing fiscal revenues.
3. Case study
“Base erosion and profit shifting (BEPS)” refers to tax implications that may result in
double non-taxation or erosion of the tax base in high-taxation jurisdictions.
In addition to the 15 actions, a process of monitoring the four minimum standards (Action 5,
Action 6, Action 13, Action 14) will be adopted and will implement review mechanisms
for other elements of the BEPS package. Monitoring mechanisms will be developed to
monitor compliance by the jurisdictions. These mechanisms will ensure the effectiveness
of the registration and dissemination of country reports as foreseen in the country-by-
country revision by 2020. All countries and jurisdictions adhering to this framework will
participate in this review process, allowing members to revise their own tax systems and
identify and eliminate elements that raise the risk of BEPS.
It also wants to help developing countries with reduced capacity. The G20 Development
Working Group (G20 DWG) called on the IMF, the OECD, the UN and the WBG to work
together to develop tools and guidance to help these countries address the issues of BEPS.
Sets of tools are prepared to facilitate the implementation of measures to combat BEPS as
well as other issues that developing countries have identified as priorities in regional
consultations. Countries and jurisdictions have been invited to express their interest in
joining this framework as partners, to participate on an equal footing and to commit
themselves to implementing the comprehensive BEPS package. Implementation times may
vary to reflect the level of development of the participating countries.
Relevant countries and jurisdictions are those whose adherence to minimum standards will
be necessary to ensure a level playing field. Relevant jurisdictions will be informed about
minimum standards and will be invited to engage in the BEPS package and to participate in
the review process. Regional tax organizations such as the African Tax Administration
Forum, the Centre for Refurbishments and Tax Administrations, the Tributarias
Interamerican Administration Centre will continue to play an important role in this project.
Specifically, regional tax organizations are essential for regional networks, and regional
meetings play an important role in the inclusive framework. Regional networks will provide
special support to developing countries for the implementation of the BEPS package.
The 15 actions of the BEPS plan are outlined below, in Table 1.
Table 1. The 15 actions of BEPS plan
BEPS Actions – 2020
Action 1: Digital Economy  The risks of digital economy
 It completes with the rest of actions
 Rules to counteract hybrid arrangements;
Action 2: Hybrid Engagement
 Domestic legislation (hybrid instruments) + Model Convention (hybrid instruments).
Action 3: Controlled Foreign  Rules for defining a CFC (including definition of control);
Companies (CFCs)  Exemptions and CFC thresholds;
 Definition of CFC income;
 Rules for income calculation;
 Rules for revenue allocation
 Rules to prevent or eliminate double taxation.
Action 4 - Limitation of interest  Main recommendation/default rule - deductibility according to a fixed indicator;
deduction  The secondary rule - a group-wide indicator (optional) for each country.
 There are additional optional elements, each of which has specific rules.
 Identify preferential tax regimes;
BEPS Actions – 2020

Action 5: Dangerous fiscal  Introduces mandatory automatic exchange of information on tax decisions related to
practices preferential tax regimes;
 Requires the existence of the economic substance for any preferential tax regime.
Action 6: CEDI Abuse  Preferred approach: Inclusion in the Treaties of both a benefit-limiting article and a general
anti-abuse rule in the form of a primary endpoint test.
Action 7: Permanent Headquarters  Changes to the definition of permanent headquarters (online commission/sales, e-
commerce).
Action 8-10: Transfer Prices,  Transfer of intangible assets between group members;
Intangible, Risk and Capital, High  Risk transfer/capital allocation between group members.
Risk Transactions
Action 11: Data  Does not suggest any changes in the local laws of the countries;
 Indicates a number of practices in the collection and analysis of data;
 Provides some specific recommendations for a more effective measurement in the future.
Action 12: Disclosure of  Mandatory reporting rules for tax planning perceived as aggressive or abusive.
aggressive tax planning schemes
Action 13: Reporting “Country by  Enhance tax transparency;
Country”  Reporting appropriate information for the purpose of conducting risk assessments
regarding transfer prices.
Action 14: Dispute Resolution  Improving the efficiency of the amicable settlement procedure (“MAP”) with regard to
dispute settlement on disputes arising under double taxation treaties following transfer
pricing adjustments by local tax authorities
 Required to introduce compulsory arbitration.
Action 15: Multilateral Instrument  Enhance the implementation of BEPS measures on double taxation treaties through a
multilateral instrument to modify existing bilateral treaties
Source: OECD.
Digitalization
Digitalization turns many aspects of our everyday life, as well as the way our economy and
society are organized and working. The size and speed of change caused by digital
transformation is notable and raises many public challenges. It also changes the very nature
of policy-making through the emergence of a new range of tools to support the development
and implementation of policies.
The widespread use of digital devices, connectivity and “smart” technology bring
significant changes that deeply affect relationships and markets. Information and
communication technology has become a fundamental part of business and social
infrastructure, highlighted by a strong dependence on online and efficient online
communications services, software and hardware.
An enormous amount of data is now generated by these users and devices connected
constantly. These data are collected by companies and governments and combined with
advances in data analysis and technology diffusion, providing the insights needed to
transform and shape both human behaviour and how organizations work.
Digital transformation into society has had significant effects on how we interact with each
other through social media growth and how we do business in the Internet age. The most
valuable element in the modern world is not gold or oil but data. Many of the largest
companies in the world are not manufacturers, retailers or owners, but platform providers,
data collectors.
These new business models are increasingly international and online. In this context, the
fiscal rules designed at the level of the business models of the 19th-20th centuries are
increasingly proving to keep up with the changes. Moreover, governments are always under
pressure to do more with less, and the ability of technology to increase collections and
reduce costs is extremely attractive.
Digital economy
In the digital economy, information is digitized and transmitted over digital networks, so a
new world of opportunities is emerging for business development. More and more people
and businesses are being introduced into the new information space. A huge amount of
information can be compacted and transmitted at high speed anywhere in the world.
Information and information technology are used in all economic sectors, somewhat to a
lesser extent, somewhat less, but in the end they are used everywhere. Their efficient use
allows companies to be competitive (Garifova, 2014).
The general concern in this area is the lack of a basis rather than the existence of erosion
of the taxable base. This is not a new area of concern. At the end of the 1990s, the digital
economy, formerly known as e-commerce, was considered by the OECD (Panayi, 2016).
Figure 1. Size of Digital Economy worldwide (%)
Digital Economy
40%
35%
30%
25%
20%
15%
10%
5%
0%
Source: Accenture Strategy and Oxford Economics.
According to the graphic above, the estimated size of the European digital economy is
24.5% of the European Union's GDP, equivalent to 3.6 trillion euros, and estimates show
that it will be 27.3% of GDP by 2020 of the European Union, which means 4.4 trillion
euros. But all this amount of money, though very large, fail to reach the budgets of the
respective states, but they reach tax havens, where they are improperly taxed or not at all.
Taking as a point of reference the results in Figure 1, it is very important to think of a digital
toll, because by increasing household access to the Internet, the volume of data provided
by the population to digital companies also increases, and then these data should be
remunerated by different methods.
Digital taxation
A digital tax application does not just mean converting paper forms to PDFs which are
uploaded to a government site. True digitization must be revolutionary, not only in the way
taxpayers complete their filings (the task of placing documents/electronic information in a
document), but also what is taxed and how can the authority use the powerful data pipelines
to complete and control taxes. (Pre-popular retaliations, with information collected from
third parties, fundamentally alters the structure of trust and direction of review, their
taxpayers and counselors now working on reviewing and challenging the authority's work.)
As far as fiscal aspects are concerned, it means that policy development and
implementation must be designed to change the environment, while being clear enough to
provide the certainty and clarity that facilitates sustainable economic growth and long term.
The digital tax is a tax like any other, so it must have in its calculation area both a tax rate
that is currently set by the OECD at the 3% threshold and a base taxable.
However, this basis puts states at a disadvantage, because it is not possible to accurately
quantify all digital services. One of the variants would be to approximate them to an optimal
solution for each jurisdiction.
For a number of years, political leaders, the media and civil society around the world have
voiced growing concerns about tax planning for multinationals (INEs) that take advantage
of gaps in interaction with tax systems to artificially reduce taxable income or change
profits to low-tax jurisdictions where little or no economic activity takes place.
Discussions on how to address the fiscal challenges posed by digitization have been
ongoing. Recent international efforts to address these issues have highlighted the divergent
positions of several jurisdictions.
Although the introduction of unilateral measures in several countries has highlighted the
urgency of the issue and the need to reassess some of the main international tax principles,
these divergent positions have made it difficult to obtain a consensus-based solution.
There are two proposals that are currently draft directives which, in order to be legislation,
must be adopted in a first step by the European institutions. In a second step, they must be
transposed into national law by each Member State. Therefore, they do not work at this time.
First Proposal: Implementing a “Significant Digital Presence” in Tax Law.
This directive requires each member state to amend its legislation so that profits can already
be taxed if a company has a “significant digital presence” (“virtual permanent
establishment”) in that Member State. This is decisive for taxing where the company has
significant user interaction through digital channels and where digital profits are generated.
As a result, a physical presence of this company will no longer be necessary for taxation.
A company is considered to have a “significant digital presence” in a Member State if it
meets one of the following limits: - annual revenue of EUR 7 million in one Member State;
- more than 100 000 users in one Member State in a taxable year, or - over 3,000 business
contracts for digital services that are created between the company and business users in a
taxable year. These new rules apply to all companies that provide digital services through
a digital interface. This explicitly includes, for example, cloud computing using search
engines and internet directories accessing or downloading movies providing online news
traffic information and weather reports and placing online advertising and streaming
services. The second proposal (“I Digital Services Tax”) suggests a new 3% tax on
revenues from certain types of digital services, which therefore resembles value added tax.
Taxation takes place when users play a major role in tax revenue will be collected by the
member state where the users are located and will thus generate immediate tax revenue for
that member state. The Commission estimates annual tax revenues of approximately 5
billion Euros.
This fee would apply to revenue generated from the following three types of digital services:
 the sale of online advertising space on a digital interface addressed to users of this
interface;
 providing a multiple digital interface that allows users to find and interact with other
users and also to facilitate the direct sale of goods and services among users;
 the sale of data collected and generated by user activities on digital interfaces.
Since the Commission wishes to eliminate newly-created businesses and enlargement
enterprises, this new tax only applies if the following income thresholds are exceeded:
 total annual total revenues of over € 750 million;
 total EU annual revenue of EUR 50 million.
While working on a global, consensus-based solution, a number of jurisdictions consider
introducing interim measures. Several countries believe that a provisional measure will
generate negative risks and consequences, regardless of the limits that might be imposed
for such a measure, and therefore oppose it. Other countries recognize these challenges, but
they believe digital services provided in their jurisdictions do not require fees to be paid,
and also say that designing measures would be beneficial.
There are a number of states that have begun the process of introducing interim measures,
including: France, Spain, Great Britain, Belgium, and Italy.
France
The country has decided to take a toll on technology giants in December 2018, after talks
on digital taxation across the European Union stagnated in the same month. The companies
affected by this measure are those with annual digital income of 750 million euros globally
and 25 million euros in France. It is estimated that this tax will bring the budget revenue of
500 million euros per year.
This charge could affect about 30 companies, most Americas, Uber Inc., Airbnb Inc.,
Google Alphabet Inc. and Facebook Inc. are part of the global technological giants waiting
for details on France's digital tax.
Spain
The tax proposed by Spain for digital taxation is three percent of the digital revenue
achieved on the territory of the state. The companies affected by this measure are those
with annual digital income of 750 million euros globally and 3 million euros for Spain.
The Spanish government forecasted that the financial transaction tax (FTT) and the digital
service tax (DST), which will go into effect on January 16, 2021, will bring additional
revenue of €1.818 billion ($2.1 billion) annually, while the Independent Authority for
Fiscal Responsibility (IAFR) has a lower pre-pandemic forecasting of €966 million ($1.1
billion). Legislative plans were disturbed when government failed to approve the budget,
which led to Prime Minister's request for early elections.
These elections mean that all the proposed tax laws, as well as the digital tax plans awaiting
parliamentary approval, will be dissolved with the government. The only option would be
for the new government to reintroduce the bill.
United Kingdom of Great Britain
The UK proposes a 2% digital income tax rate for companies with digital revenue of around
500 million pounds (around 580 million euros) globally and 25 million (about 29 million
euro) British pounds of digital revenue in the UK. A disadvantage would be deductibility,
companies may deduct UK tax from income tax they have in the country but will not
receive tax credits (as opposed to deductions and exemptions, which reduce the amount of
taxable income, tax credits reduce the real value of the tax due).
Chancellor Philip Hammond said the United Kingdom would implement the only digital
tax if it still stagnates at the EU level on digital taxation.
The Ministry of Finance of the country expects the law to be added to the draft law for
2019 and to enter into force in January 2020.
Italy
Italy has proposed a digital levy since 2017, and Parliament finally endorsed it in 2018.
Unlike the EU tax, the digital tax is more about service buyers than sellers. Any enterprise
that performs more than 3000 digital transactions between businesses in Italy in a calendar
year will be taxable. Companies cannot use the tax to offset Italian income tax.
Belgium
It is the last country in the European Union to publish a digital service tax bill. The country
wants to introduce a 3% provisional tax on digital income, such as the sale of user data to
companies with a total revenue of 750 million euros and revenue of 50 million euros in the
European Union.
Romania
The graph (Fig.2) with diminishing loss show how much the fiscal loss was diminished by
the NAFA (National Agency for Fiscal Administration) controls, following the application
of the transfer pricing rules. And the graph with tax on additional profit (Fig.3) shows the
additional amounts of tax on profit established by NAFA following the application of

transfer pricing rules. The idea is as it follows: a company is either on a profit or a loss.
Figure 2. Loss reduction in Romania as a result of tax inspections in which the transfer prices’ file was
requested for the period 2013-2018
Source: National Agency for Fiscal Administration.
As it can be seen in Figure 2, in 2018 the loss reduction was 771,664,471 LEI, being the
highest value recorded during the period of 2013-2018. It was noted that the evolution of
loss reduction had a substantial growth in 2015 followed by a downfall in 2016. The loss
reduction observed in 2013 was in amount of 1,505,399 LEI, considered the lowest value
from the period.
Figure 3. Additional income tax set due transfer pricing adjustment evolution in Romania for 2013-2018

Figure 3 shows the additional income tax set due transfer pricing adjustments in Romania
that started increase from 2016 when it was a downfall, doubling its value in 2017 in
comparison with the previous year and following an increase trend for the next period, in
2018 the increase being significant, more than 4 times compared with year 2017.
Figure 4. Corporate tax paid by banks in Romania for the period 2013-2019
2.500.000,00 3000%
2835%
2500%
2.000.000,00
2000%
1.500.000,00
1636%
1500%
1.000.000,00
1000%
874%
500.000,00 593% 500%
287,09% 327%
‐ 0% 0%
2013 2014 2015 2016 2017 2018 2019
Net income on corporate tax from banks % increase compared to 2013
Ensuring that our tax systems are prepared to respond to the changes brought about by
digital transformation, and to capitalize on its opportunities and provide protection against
its potential risks is a critical challenge. Reviewing international tax rules in the light of the
impact of digitization will be a significant component of this activity and will have
important ramifications for MNEs and governments as well as for the future of our tax
systems. An update of the OECD's work in these areas will form part of the Tax and
Digitalization Report, which will be prepared under the inclusive framework to be
transmitted to the G20 in 2020, The report also acknowledged that it would be difficult, if
not impossible, to stop the digital economy from the rest of the economy for fiscal purposes
because of the increasingly widespread nature of digitalization (OECD 2015).
Although the banking system has increased in terms of net assets reported by banks and
their activities, the tax on profit paid by banks was at a very low value.
In 2015, the first punctual verifications are initiated at the banks level and on the 9th of
May 2016 the tax office starts the first fiscal control at a bank in Romania. The control
targets several categories of taxes but the result is focused on the tax on profit.
During 2016, at the Department of administration of large taxpayers level, a department
that manages administratively all the banks located in Romania, several analyzes are made
and it is found that 35 of the 42 banks operating in the Romanian banking system did not
pay any tax on profit in the last 5 fiscal years.
Afterwards, new controls are initiated among the banks located in Romania.
The graph shows a sudden increase in the tax on profit paid by banks, starting with 2017.
This result is given by both the fiscal controls generated and also the overwhelming
proportion to a legislative change that clearly regulates the way in which receivables are
recorded in the accounts off the balance sheet and then transferable.
This legislative change practically begins to prevent banks from artificially generating tax-
deductible expenses, by selling high-performing or non-performing loan portfolios at a
value between 5-7% of their value. By assigning these receivables, the banks recorded
losses between 93-95% of the value of the assigned receivables. Most often these
receivables were assigned to affiliates or offshore companies.
This change practically forces the banks located in Romania to switch to profit and by
default to pay amounts representing the higher tax on profit by 2385% higher than in 2013.
With respect to the assigned receivables, the net loss representing the difference between
the transfer price and the value of the assigned receivable is deductible up to a ceiling of
30% of the value of this loss. If the transferee assigns the receivable, the net loss is
determined as the difference between the transfer price and the acquisition cost of the
receivable. When it comes to the credit institutions, if the assigned receivables are partially
or fully covered by adjustments for expected losses, as well as if the receivables are
recorded in off-balance sheet accounts and then assigned, 70% of the difference between
the amount the alienated receivable and the transfer price represent elements that are similar
to the income. (Romanian Fiscal Code, Article 25, paragraph 10)
Conclusions
The effects of these researched actions will be seen in a few years, the EU expects to be
implemented by 2020, hoping to improve the fiscal climate in the Member States.
The graphs with diminishing loss show how much the fiscal loss diminished the NAFA
controls, following the application of the transfer pricing rules. And the graph with tax on
additional profit shows the additional amounts of tax on profit established by NAFA
following the application of transfer pricing rules. The idea is this: a company is either
profit or loss.
In the wake of the implementation of work procedures on the transfer of profits, access to
databases that provide information on the transfer of profits and the acquisition of risk
analysis software, a strong increase in the amounts collected from the state budget can be
noticed.
Through operations of transfer of goods and services at overvalued prices, large companies
in Romania used to transfer profits to other countries, generally tax havens.
Through these payments at an overvalued price of the contracted goods and services, the
Romanian companies practically registered very high expenses that artificially diminished
the profit or even generated the loss of the companies.
Following the controls performed, especially during the period 2017-2019, the tax
authorities decreased the losses artificially recorded by the companies, reconsidering
certain types of expenses and thus, in the event of some of them, practically forcing them
to register profit and, therefore, to pay tax on profit. This happened considering that they
have been recording losses for years.
References
Dowd, T., Landefeld, P. and Moore, A., 2017. Profit shifting of U.S. multinationals. Journal of
Public Economics, 148, pp. 1-13.
Garifova, L., 2014. The Economy of the Digital Epoch in Russia: Development Tendencies and
Place in Business. Procedia Economics and Finance, 15, pp. 1159-1164.
HJI Panayi, C., 2016. International Tax Law in the Post-BEPs World. Singapore Management
University School of Accountancy Research Paper, (2016-S), p. 47.
Janský, P., Prats, A., 2013. Multinational corporations and the profit-shifting lure of tax havens.
Christian Aid Occasional Paper, p. 9.
Knickrehm, M., Berthon, B. and Daugherty, P., 2016. Digital disruption: The growth
multiplier. Dublin: Accenture, pp. 1-12.
Lamers, I., Mcharo, P. and Nakajima, K., 2014. Tax base erosion and profit shifting (BEPS) and
international economic law (No. BOOK). The Graduate Institute of International and
Development Studies, Centre for Trade and Economic Integration.
National Agency for Tax Administration, 2018. Annual performance report. Bucharest: NATA
OECD, 2014. Countering harmful tax practices more effectively, taking into account transparency
and substance’. OECD/G20 Base Erosion and Profit Shifting Project.
OECD, 2015. Explanatory Statement, OECD/G20 Base Erosion and Profit Shifting Project, [online]
Available at: <http://www.oecd.org/tax/beps-explanatory-statement-2015.pdf/>
Omar, N. and Zolkaflil, S., 2015. Profit Shifting and Earnings Management through Tax Haven
Subsidiaries: An Exploratory Analysis of Multinational Companies. Procedia Economics and
Finance, 28, pp. 53-58.
Stankevicius, E. and Leonas, L., 2015. Hybrid Approach Model for Prevention of Tax Evasion and
Fraud. Procedia - Social and Behavioral Sciences, 213, pp. 383-389.
US PIRG Education Fund and the Institute on Taxation and Economic Policy, 2017. Offshore Shell
Games 2017.The Use of Offshore Tax Havens by Fortune 500 Companies.US: US PIRG
Education Fund.
www.taxfoundation.org
Implementation of public policies.

The compatibility of the model of public policy
with the target space of the policy
Luminița Gabriela POPESCU

National School of Political and Administrative Studies, Romania
lgpopescu22@gmail.com
Abstract. New realities bring to the fore a series of challenges referring to the implementation of
public policies. One of the major challenges is caused by the lack of compatibility between the
chosen model for the elaboration of policy, on the one hand, and the target public space, on the
other.
That could mean that either the technical solutions formulated by the researchers, to be applied to
the target space are totally unrealistic, or in the target space there are zones of confusion, of
instability, of unpredictability, where the model chosen by the policy-makers cannot be applied. The
degree of difficulty is so much more difficult as the members of the community and society, in
general, are preoccupied, predominantly, with the problems that appear in the target public space,
while almost not at all interested in the technical problems faced by the policy-makers.
Despite this lack of interest, anyway, the problems that the community/society faces cannot be solved
but by the implementation of their initiatives and models proposed by the policy-makers. The success
of a policy is, in our opinion, firstly, a matter of compatibility between the model of the public policy
proposed and the target public space characterised by a high degree of inertia and tendencies
manifest in maintaining the status quo.
How could there be reached the compatibility between a public policy, whose vision is based on the
behavioural change of the institutions(1), organisations and people, on the one hand, and the
unreformed public space, aligned to the exigencies much exceeded by reality, on the other? Which
are the approaches necessary to a compatibility between the models of the policies and the target
space? What strategies must be chosen? How must the efforts and resources in the public space be
graduated and oriented to become compatible with the model chosen by the policy makers? These
are questions which we will try to answer in this article.
Keywords: target public space, policy-makers, evidence-based public policies.
JEL Classification: Z18, Z19.

346 Luminița Gabriela Popescu
1. Introduction
Studies on the implementation of public policies constitute a major challenge in today’s
context. Starting from the citizens’ refuse, more and more vocally expressed, to accept
policies formulated behind closed doors anymore, the rethinking of the public policies
process appears as a real priority. So that the implementation of a public policy be a success,
an imperative requirement is to give up rigid, inertial and hierarchical structures whose
main preoccupation is the preservation of the status quo in favour of adopting flexible,
transparent, adaptive structures to be consonant with the novelty brought on by a public
policy.
More precisely, O’Toole assumes that the implementation of a policy “refers to the connection
between the expression of governmental intention and actual results” (1995, p. 43) or as
DeLeon put it, “what happens between policy expectations and (perceived) policy results”
(DeLeon, 1999, pp. 314-315).
There is a multitude of variables that affect the way public policies are elaborated and
implemented, the including the political opinion, extant at national level, on the issue that
is to be solved by the initiated public policy, the public attitude on the state and the
employees, on the skills and knowledge of the managers in the public sector, on the degree
public services are citizen-oriented. Despite the fact that the boundary between the
elaboration and the implementation of the public policies cannot be clearly established,
nowadays, the relationship between the units providing public services and the structures
of the governmental policies is called into question.
The main problem is triggered exactly by the impossibility of clearly mark off the boundary
between the elaboration of public policies and the implementation of public policies. A
good policy cannot be shaped but by the actors who know how services and policies may
be managed, while insufficient experience in the domain may have impractical policies as
a result (Popescu, 2013, p. 132).
Another major implementation challenge is the lack of compatibility between the model
established for policy elaboration, on the one hand, and the space of the target public, on
the other hand. That lack of compatibility may mean that either the technical solutions
formulated by academics to be implemented in the target space are totally unrealistic or
that there are zones of confusion, of instability, of unpredictability in the target space,
where the model established by the policy makers cannot be implemented. The degree of
difficulty is much greater when the members of the community and society, generally, are
preponderantly absorbed by the problems that appear in the target public space and almost
not at all interested in the technical problems faced by the policy makers.
2. Different representations of target public space

As to the representation of the target public space, researchers have rather contradictory
opinions. Thus, for D. Shon the target public space is similar to a “swamp” where solutions
are confused, incapable of containing technical elements. In these terms, the swamp
represents the reality of complex uncertain situations.
Implementation of public policies 347
Contrary to his opinion, professor Ron Amann appreciates that in this swamp there may
also be identified zones of solid ground where real actions may develop to support both the
updating of the models of elaborating policies and their implementation (Pearson, 2002,
p. 41).
We consider that the solid ground Ron Amann is referring to is representative for the case
of the simple uncertain problems.
The issues which are the major subjects of public policy, and which concern, in particular,
the future, differ in terms of complexity and uncertainty.
Complexity refers to the number of separate or variable elements, the number of
connections between them, and the degree of variability of these connections. Obviously,
complex issues admit of a completely different approach than simple ones. Complexity
imposes the need to turn to professionals who, based on their experience and knowledge,
can develop specific strategies for solving complex problems.
In the case of simple problems, due to the sufficient volume of information available, they
are sufficient experience and intuition supported by routine (standard) procedures.
The term uncertainty refers to the availability of information and the perception of the
phenomenon essential for the decision-making process and for the adoption of the
necessary measures. The degree of uncertainty has a significant impact on the possibility
of predicting public policy. A detailed policy aimed at reducing or reducing an uncertain
problem has little chance of success.
The convenient project (optimal, recommended or preferred) chosen based on the
considerations that it will lead to the desired change, with lower costs for society than the
other variants, is a desideratum permanently subject to practical and psychological threats.
The analysis of the situation and the identification of its critical elements represent the
content of the first stage of the policy design process A clear understanding of the nature
of the problem and, implicitly, the location of the pressure sources contributes to the
generation of creative solutions. If the problem is insufficient or vaguely defined, the
solutions will be far from expected. In order to solve a problem, it is important that, from
the very beginning, the problem be formulated in a way that facilitates the discovery of
solutions.
3. Strategic models for public policy making and different typologies of public space
To answer the questions of compatibility, a first approach means configuring a link between
strategic thought, based on elaborating a public policy, the various typologies of the target
public space and the graduating of the efforts and resources on that basis. To this effect, we
formulated three working assumptions, analysis following.
Assumption 1. The restraints coming from the concrete reality are neglected, efforts being
concentrated, mainly, on the abstract domain of elaborating a policy, observing the rigours
imposed by the results of the researches and of the technical solutions.
Public policies carry values that emerge from the vision of the political management
regarding the future evolutions of the target public space. The political management is
responsible for the formalisation, dissemination and creation of a climate where the vision
and the values defining public policy taken into account will be appreciated by all the actors’
part of the target space.
Assuming such a behaviour by the political management will influence the model chosen
for elaborating a policy in a positive way (see Figure 1).
Figure 1. Strategic concepts of public policies
Vision
Values
Strategic
options
Elaborating the public policy

...
Source: Popescu (2005, p. 118).
In the present situation, requirements aim at democratisation and transparency of the public
policy process, which presumes eliminating distortions and testing perceptions and
assumptions by debates, argumentations and discussions.
In other words, elaborating public policies should become a process of public learning.
This assumption leads to the rational model, formulated by H. Simon (1945), a model
suitable for a target space where problems are linear, simple and certain and which is
compatible with the traditional bureaucratic structures (see Figure 2.a).
The tendencies of using the rational model were limited as it became ever more evident
that the phenomena and events in the real world do not follow a linear model and,
consequently, dividing in component parts and their subsequent re-assembly is not a
convenient solution for a constantly changing world.
Assumption 2. Scientific rigour is being abandoned in favour of focussing efforts and
resources on the target space where simple and uncertain problems arise.
In this case, strategic thinking requires the intuitive model, which is based only on
the experience, flair and intuition of decision-makers and not on rigorous analysis (see
Figure 2.b).
Intuitive decisions are made, inclusively, on the basis of evidence and lessons learned from
past success/mistakes, which is a return to the past, despite the rhetoric of “new” and
“modernisation” (in a past where better policy making means policy making based on
improved instrumental rationality).
Literature takes this approach as evidence-based policy approaches, for which knowledge
is the means to control unstable, high-risk situations. From this perspective, the intuitive
strategic approach marks not so much a step forward, but above all the best way to what
works. The philosophy of “what works” is a real opportunity for social sciences to influence
the process of public policies after a long period of time when they were considered
irrelevant (Parson, 2002, pp. 44-45).
Making the evidence-based policy-making model compatible with the target space requires
strategic approaches aimed at:
 Academic research. The challenge is to influence the production of relevant policy
knowledge so that they can be reproduced in a form that can be used by policy makers.
The trend in the evidence-based approach is, by excellence, a quantitative approach.
 Professional and organisational experience Focussing efforts and resources is likely
to contribute to the development of both the system, which can facilitate better
management of learning, as well as knowledge at government and departmental level.
The characteristic of the two forms of management is that they are likely to be explicit and
encoded. Evidence can be aggregated and disseminated throughout the target space. From
a certain level this means that on the one hand structures and procedures are created so that
the institutional modes of “what works” and “contribute to learning” can be extracted,
stored, retrieved and communicated, and on the other hand, academic research must be
produced and disseminated in accordance with clear specifications (Pearson, 2002, p. 46).
With regard to the evidence-based approach, Pearson considers that its opportunity to
contribute to improving the act of government is rather limited, and the contribution of this
is, in fact, to make the relationship between knowledge and policy making even more
confusing.
In support of this view, we can invoke the risk of uniformity, while the approach toward
creativity and innovation is almost completely neglected.
Thus, in evidence-based health policy, the centre of gravity is the disease and the standard
procedures laid down for treating that disease. In fact, however, the centre of gravity of
health policy is the patient (and not the disease) whose reactions cannot always be
standardized. Each patient has his own representation of how he feels the disease (Popescu,
2013b, pp. 96-98).
Another example is in the field of education; through regulations and standards, trends are
to standardize education at the expense of educational diversity.
As a conclusion, we can say that, apart from undisputed contributions, the evidence-based
approach negatively affects innovation, creativity and, by extension, competitiveness.
Politics should not be guided by dogma, but by knowledge of what works and why. It
follows that the evidence/knowledge is needed for policy makers to be able to measure the
size of the effect of A on B.
On the other hand, the evidence-based approach is to make changes fast, which will also
be positive in other areas. The relatively rapid potential for change and the visible and
measurable impact were the main criteria in the choice of transformational areas. There are
examples of good practice and successful institutional or managerial models that can serve
as a model. Poland, for example, offers a very good model in many areas, from the way it
has solved its problems in the state-owned enterprises sector to the way it attracts and
implements investment projects financed by European funds.
Without minimizing the contribution of this type of approach, we consider it important to
stress that the What works approach is not such as to ensure the success of the
implementation process, especially in the case of uncertain and complex situations, which
are also the most common in the world we live in.
Assumption 3. Scientific rigour is being abandoned in favour of focussing efforts and
resources on the target space where simple and uncertain problems arise.
We appreciate that we can assimilate Schon’s “swamp” to this constantly changing reality,
faced with complex and uncertain problems, for which scenarios different from those
already in place must be envisaged.
Under these circumstances, the requirements derived from the current requirements lead to
a reconfiguration of the target area in a way that stimulates innovative thinking and
contributes to the development of innovative models by policy makers (see Figure 2.c).
Edmund Phelps, Nobel Prize winner, pointed out in an interview with the Financial Times
in June 2014, that Europe’s problem, beyond the financial crisis, was that politicians
prevented innovation or reduced incentives for those who want to innovate. After a major
slowdown in dynamism in the ’60s, productivity growth started to decline in almost all
Western countries, halved in the US in the ’70s and almost stagnating in France, Germany
and the UK at the end of the ’90s.
So, despite an increasing need for innovative models, the response of officials is either
inadequate or is likely to be belated. They are deliberately involved in swamp, but also in
crucial matters and when asked to describe the methods used, they speak of experience,
tests, errors, intuition and muddling through (disorder, chaos) (Schon, 1983, pp. 42-43).
Figure 2. Strategic models for public policy making

Rational model (a) Intuitive model (b) Innovative model (c)
Prototype
problem
Solution
Transformation
Re-arranging Spot optimisation/ or changing
elements improvement configuration
Attempts to find an answer to complex and uncertain problems by using, for this, evidence-
based approaches, for example, are doomed to failure. The solutions for transforming the
target space continue to focus on the old paradigms and this is to the detriment of innovative
elements. In this respect, we are raising the controversy generated by evidence-based
approaches in the case of educational policies. While not minimizing the contribution of
this type of approach to improving educational process, the inherent risk of standardization
to innovation in education should be noted.
The success of the implementation of new innovative models in the target public space
implies, as stated in the beginning of this chapter, increasing/developing the target space to
the level at which it can respond to as many challenges as possible (cognitive, informational,
behavioural resource limitation, cultural) associated with innovative policies initiated. The
risk of implementation failure is due to the targeting of resources and efforts in areas other
than those through which challenges can be overcome. In other words, neglecting or ignoring
the compatibility of the target area with the “new” introduced by the public policy initiated
leads to major failures that can lead to total undermining of the public policy project. From
the panoply of the challenges listed, we bring to the fore the culture and values of the target
public space which, in our view, are most difficult to make compatible.
According to the theory developed by Lasswell, values are the essence of a public policy.
Moreover, he considers values to be real obstacles that may affect the objectivity of what
works (Lasswell, 1951).
In order to be able to understand all the implications, we will continue to look at the key
issues relating to the target area.
An organization is not just a conglomerate of people and resources. This is a set of values
and assumptions that determine its own climate and organizational culture. In addition, the
maturity of the organization is positively reflected in the organizational culture. No matter
how they are structured, organizations are not lifeless machines. The people involved, their
experience in the past of the organization, their own system of convictions and values
contribute to the vitalisation of the organisation. As the organisation gains knowledge and
manages to solve the complex problems it faces, it becomes mature. Similarly, cultural
bearers can be the engine of change or an element of maintaining the status quo, depending
on the values they make explicit by their behaviour (and not the values they publicly declare
to support).
4. The culture of target public space a challenge for implementation actions
Meeting the above-mentioned demands raises a new problem, namely the drivers of change
(organizations, authorities and officials able to put pressure on change and adaptation to an
increasingly uncertain and unpredictable world).
We stress that in our view that the success of any type of public policy is dependent on the
change of the culture of the target public space, in the sense of political socialisation. Any
other type of change, methods or processes, is favoured or. On the contrary, stopped by the
level of political socialisation.
Culture is the way that members of a group communicate with each other and with other
groups in society. In the course of the communication process, common behaviours, habits,
practices, values, beliefs and visions of the future target space are reflected and become
tangible as the policies initiated are turned into action. The importance of a new culture,
centred around the transformation sought by the implementation of a public policy, lies in
its implications for implementing the project of change. Sequential implementation, of only
those new practices that seem more useful or more generally applicable, can have
remarkable effects in the short term. Fragmented efforts, however, cannot support long-
term performance. For example, consumer policies can only succeed in the short term at
most, without a solid consumer culture being crystallised (Popescu, 2005, pp. 119-124).
The result of the completion of the concepts shown in Figure 1 with these considerations
is shown in Figure 3. In conclusion, only adaptive organizational environments compatible
with public policies (these new and changing promoters) are able to recognize the
importance of the new values promoted. From this perspective, the organisational analysis
brings the adaptive organization to the fore, characterised by:
 sophistication and maturity;
 flexibility and not rigid bureaucracy;
 proactivity;
 motivation mechanisms that promote openness and commitment to the organisation;
 a collaborative climate that facilitates team work;
 interest shown in openness to the community;
 decision transparency;
 constant concern for the development of the organisation so that it can be continuously
transformed.
Figure 3. From culture to action
Vision
Culture
Values Values
Public policy
...
Strategy
Feedback
Planning actions
Actions
Results/Evaluation
How could organisations be forced to give up the rigidity and tradition in which they seem
bound and accept the new organisational models? We need to think about how values and
identity could be maintained in a changing environment; we need to think about responding
to the challenges, how to influence and manage these changes in order to achieve a certain
degree of success.
In conclusion, efforts and resources must be allocated to the development of the target
public space, including from a cultural perspective.
The culture of public space, which we define as a patchwork of cultures of the actors
involved (public, business and civic organisations, of communities that are interested in a
given public policy), must be brought into line with the values and ethics that public policy
promotes while it is working its way among the Scylla of good intentions and the Charybdis
of bureaucracy. It is a difficult process, which needs the support of a strong culture.
However, the strength of a culture lies in its actual content and adequacy of the solutions
found to the problems of public policies. A strong culture requires the entities to agree on
how to design, implement and evaluate the public policy project.
Strong culture does not mean blind acceptance, but consensus over values and attitudes.
Building such a culture requires a lot of time and substantial efforts to learn and strengthen it.
For a better understanding, we will consider that the target public space can be assimilated
to a meta-organisation – a conglomerate consisting of all the actors involved in the
implementation process. The culture of the meta-organisation is the common places of
cultural diversity which must provide coherence and consistency to the complex structure
it defines. (In the complex structure, they are non-governmental, private and public,
national, international and transnational organisations.)
Despite the fact that cultural uniformity between these cultures is not being sought (nor
possible), there is still a need for some compatibility. This means engaging stakeholders,
which are members of the meta-organisation, in a sustained effort to create a new culture
based on the values and beliefs promoted by public policies.
5. Conclusions
The decline in societal performance appears to be a difference between environmental
opportunities and the capacity to promote public projects to capitalize on these
opportunities. Improving the quality of public projects, of forecasts of future developments
and taking management risk into account in the real-time response is the ability to close
this gap.
Under these assumptions, the mechanistic cultural model, typical of bureaucracy, is
completely out of date. The cultural model of the organic organisation is the one to which
moves to change the organisations that are part of the organisation must be concentrated.
In an overwhelming majority, to the various and serious problems faced by red tape, their
only response is to create new institutions and develop new projects, which will help them
to overcome the challenges and solve the problems. In most cases, such an approach does
not provide the expected satisfaction, due to the erroneous arguments underlying the
initiation of the response, in the sense that, for example, an attempt is made to resolve a
behavioural problem by means of a structural solution. In addition, the institutions are
rather frequent in the cases where they are waiting to replace existing programs with new
ones, in which, unfortunately, the same operating practices are used, which have already
become traditional, favouring the same customers and are subject to the same errors.
While responsible bureaucracies, so-called learning organisations, are concerned about the
added value of the processes they develop, organisations of this kind become active only if
signals are made that customer satisfaction is no longer needed.
Public policy-makers must not only rapidly acquire the innovative and creative capabilities
needed to define new horizons, but also facilitate their practical implementation. In other
words, the old paradigms need to be abandoned, but the complete commitment, in spirit
and action, to a process of long-term change, in line with the requirements of the
management of complex systems should be taken into account. Situations, where between
the culture of the actors involved in the implementation of a policy, significant differences
do not appear, are desirable and coincide with the entities have a common organisational
past.
Convictions, aspirations and behaviour are specific and reflect activities within the entity,
but retain the same general characteristics as the culture of the meta-organisation. It is
essential that there should be no too great differences or irreconcilable conflicts between
basic culture and the cultures of actors. Seeking cultural uniformity, in the absence of solid
arguments, is a source of conflict. Sometimes too much cultural differences between
stakeholders hamper communication and collaboration between them or with the various
groups of experts participating in the public policy project. Public policy management is
favourable to this kind of cultural diversity, as long as it does not affect the smooth running
of implementation (Popescu, 2005, pp. 269-271).
Note
(1)
Defined as rules guiding the behaviour of the actors involved in the development of a public
policy.
References
Laswell, H.D., 1951. The Policy Orientation, in D. Lerner and H.D. Laswell (eds.), The Policy
Sciences, Stanford University Press.
DeLeon, P., 1999. The Missing Link Revisited: Contemporary Implementation Research, Policy
Studies Review, Vol. 16, No. 3/4, pp. 311-338.
Popescu, L.G., 2005. Politici publice, 2nd edition, Economica, Bucharest.
Popescu, L.G., 2012. Strategic Responsiveness and Market Repositioning in Higher Education, LAP
LAMBERT Academic Publishing GmbH & Co., Saarbrücken, Germany.
Popescu, L.G., 2013a. Strategic approach of total quality and effects on higher education institutions.
Case of Romania, Journal of Eastern Europe Research in Business and Economics, Article ID
213298, DOI: 10.5171/2013.213298.
Popescu, L.G., 2013b. From a holistic approach of public policy to co-governance, Theoretical and
Applied Economics, Vol. XX, No. 7(584), pp. 95-108.
Schon, D.A., 1983. The Reflective Practitioner, New York: Basic Books.
Schon, D.A. and Rein, M., 1991. Frame Analysis, in P. Wagner, B. Wittrock, C. Weiss and
H. Wollman (eds.), Social Sciences and Modern States: National Experiences and Theoretical
Crossroads, Cambridge, Stanford University Press.
O’Toole, L.J. Jr., 1995. National Choice and Policy Implementation: Implications for Inter-
organizational Network Management, The American Review of Public Administration,
Vol. 25, Issue 1.
O’Toole, L.J. Jr., 1993. Multiorganizational policy implementation: Some limitations and
possibilities for rational choice contributions. In F.W. Scharpf (ed.), Games in hierarchies and
networks, pp. 1-39. London: Westview.
Fiscal sustainability in Romania
Ioan-Radu PETRARIU
radu.petrariu@rei.ase.ro
Lucian Constantin VÎLCU
vilcu_constantin_lucian@yahoo.com
Iulian Cornel LOLEA
loleaiulian@gmail.com
Liana VLADU
National University of Political Studies and Public Administration, Romania
vladuliana21@gmail.com
Abstract. In this paper we have tested the Romanian fiscal sustainability during the period between.
In this way. In this regard, we analyzed the relationship between the public debt and primary budget
deficit, and between public expenditure and public revenue in Romania. We aim to find out what if
there are any type of relationships as Afonso and Jalles (2012) defined sustainability: unidirectional
causality, bidirectional causality (perfect fiscal synchronization) and no causality. Hence, our
approach was focused on testing fiscal sustainability in Romania and we have found that there was
a sort of sustainability. In this period the policy makers took some measures in accordance with the
definition of what fiscal sustainability should mean.
Keywords: sustainable fiscal policy, primary budget balance, public debt, public expenditures, public
revenues.
JEL Classification: C22, G11.

358 Ioan-Radu Petrariu, Lucian Constantin Vîlcu, Iulian Cornel Lolea, Liana Vladu
Introduction
The economists (Blanchard et al., 1990) have argued that fiscal policy is sustainable when
the public debt does not explode and the governments are not forced to raise taxes, reduce
spending and increase fiscal deficit or public debt. According to them, the present value of
future primary surpluses must be equal to the current level of public debt. They considered
that the government should borrow money to finance the primary deficit (the difference
between primary expenditure and government revenue), the payment of previous year
interests and the public debt of the previous year.
In order to ensure the long-term budgetary resources as a premise of fiscal sustainability,
the economic and social context that will lead to increased public spending will be taken
into account.
The long-term sustainability of the public finances is also taken into account in the
assessment of stability and convergence programs. In the upcoming decades, the size and
age structure of Europe's population will undergo dramatic demographic changes. The
aging of the population will generate great economic, budgetary and social challenges. In
response, the Commission published a special report in which it concluded that if EU
governments continue to implement their current policies, the public debt will increase
sharply in the upcoming decades. Both fiscal consolidation, on the one hand, and general
economic reforms, on the other, are needed to meet the challenge of aging. Bringing, in
this way, a sustainable adjustment of the budgets, it will go a long way in improving fiscal
sustainability. Bringing, in this way, a sustainable budgetary adjustment will contribute a
lot to a long way in improving fiscal sustainability.
The sustainability of public finances has generated many debates that have led to the
identification of different economic models developed by economists such as Gupta, Keen,
Clements, Fletcher, de Mello, and Mani (2002) that involve the convergence between
economic and social growth and the natural environment protection policies. Starting from
this idea, the authors also highlighted the impact of fiscal policy on economic growth and
investment in human capital. Corsetti and Roubini (Corsetti, 1991) has shown that policy
changes or changes of relevant macroeconomic variables, such as economic growth,
inflation, interest rates, are needed in the future, when budgetary constraint is not supported
by empirical testing. (Moraga and Vidal, 2004) also highlighted the impact of fiscal
sustainability on economic growth, as well as the need to respect the intertemporal
budgetary constraint in order to ensure long-term sustainability.
Institutions such as the International Monetary Fund (IMF) and the Organization for
Economic Co-operation and Development (OECD) have also focused on sustainability
because, especially in the 1980s, many industrialized countries recorded significant public
debt, which led to numerous episodes of fiscal adjustments in order to limit budget deficits.
In addition, the member countries of the Economic and Monetary Union, as well as the new
members of the European Union, are facing legal problems of tax constraints (Stoian,
2007).
Fiscal sustainability in Romania 359
Fiscal sustainability is an important element of economic sustainability, with the following

components:
a) Stability – “The debt ratio target – as a percentage of GDP – must be such that, once
reached as the end point of a convergence process, it can always be maintained as the
average debt ratio over all subsequent economic cycles (stability)” (Hiebert, 2000).
b) Security – “must be maintained as the average debt ratio recorded in all economic cycles
that have taken place after the completion of the convergence, allowing the economy to
be safeguarded from any deficit situation that could be considered excessive in
accordance with the rules of the Treaty” (Hiebert, 2000) which it allows a country to
build its global deficit so as not to violate the 3% of GDP limit.
The conditions for fiscal sustainability, such as debt and primary surplus rates, tend to
converge to their long-term benchmarks regardless of initial conditions. The issue of
correlating budget revenues with the need for financial governance and obligations for a
public data service that leads to the association of a sustainable fiscal policy with the
solvency of government funding or clarity of liquidity. The problem of correlating the
budget revenue with the need for financial governance and its obligation to serve public
debt leads to the association of a sustainable fiscal policy with the solvency of government
funding or clarity of liquidity. In this sense, the sustainable level of public debt is that
amount of debt that could be repurchased by public authorities without adjusting revenues
and expenditures in the next period. Therefore, the fiscal sustainability is strongly
influenced by financing costs and the ability or willingness to satisfy the debt service.
The solvency conditions for assessing a long-term sustainable fiscal policy, formulated by
Artis and Marcellino in 2002 (Artis, 2002), is that public debt coverage provides prospects
for sufficient future budget surpluses. This budgetary constraint expresses an accounting
identity which, according to Mendoza (2003), involves measures to meet intertemporal
budgetary constraints affecting government debt. Therefore, the definition of fiscal
sustainability based on intertemporal budgetary constraints allows the increase of the short-
or medium-term liquidity level or long-term insolvency based on costly fiscal adjustments,
according to (Roubini, 2001).
The fiscal instruments used will take into account the dynamics of economic growth,
current account balance, balance of payments, budget deficits, interest rate and foreign
direct investment for the forecast of available resources for the interest payments on public
debt. The resources level will be compared with the value of the public debt interest and
thus one could identify the type of problems that the government has to solve regarding
solvency or liquidity. Consequently, the government must reduce its level of indebtedness
and implement measures to restore financial discipline. However, a simple analysis of the
country's indebtedness is not enough to determine the degree of sustainability of fiscal
policy. Increasing the degree of indebtedness does not necessarily imply the registration of
an economic growth rate higher than the real interest rate on government loans. Therefore,
the solvency of the state is a necessary condition, but not sufficient for the fiscal policy
(Horne, 1991) being necessary to analyze the economic environment and the origin of the
financing sources for the economic environment as a whole.
(Ballabriga et al., 2005) emphasize the importance of long-term solvency, which acts as a
constraint on fiscal policy, as the current public debt must be covered by the present value
of future primary surpluses.
Literature review
The economic literature distinguishes between different definitions of financial contagion.
However, the most widely used definition is that of Eichengreen et al. (1996) who consider
that the contagion effect is a significant increase in the probability of crisis in one economy,
conditioned by the occurrence of a crisis in another. Forbes and Rigobon (2001) have
developed their own definition that states that the term contagion represents a significant
increase of the links between markets, after a shock in one country or in a number of
countries. The significant increase in links between financial markets involves creating or
enhancing new transmission channels during the crisis, regardless the fundamental
principles and responding to a crisis in a particular country. In general, contagion refers to
the spread of financial market disruptions at regional, or even global, levels.
There are several theoretical and practical studies that have focused on the analysis of the
contagion phenomenon. Bekaert et al. (2014) analyzes the transmission of the financial
crisis from 2007 to 2009, using a factor model to predict the return in crisis times, defining
unexplained increases in factor importance and residual correlations, as an indication of
contagion.
Contagion research during Sovereign Debt Crisis in Europe, based on correlation analyzes,
indicates mixed results, but many of them are in the same direction. Some papers, such as
Claeys and Vasicek (2014), found a significant increase in correlation coefficients between
financial markets during the sovereign debt crisis in the Europe. Moreover, Horta (2013)
analyzes the contagion effect of the Greek equity market on the European equity markets
of the NYSE group in the context of the 2010 sovereign debt crisis, performing three
contagion tests based on copula functions.
Cho and Parhizgari (2008) define and measure the contagion phenomenon by analyzing
the East Asia financial crisis of 1997 on the equity markets of eight countries using DCC-
GARCH.
Cho and Parhizgari (2008) define and measure the contagion phenomenon by analyzing
the East Asia financial crisis of 1997 on the capital markets of eight countries using DCC-
GARCH. Considering Thailand and Hong Kong as alternative sources of contagion, a total
of fourteen target source pairs are analyzed using DCC-GARCH methods and median and
median difference tests. They define contagion as a statistical break in the calculated
conditional correlations, measured by the changes of the mean and median. It is also worth
mentioning the paper of Forbes and Rigobon (2002), in which they analyze the collapse of
the equity market in 1987, in addition to the 1997 crisis, with a correlation measure adjusted
according to heteroskedasticity.
In addition, the paper published by Chiang et al. (2010) identified the contagion effect
during the Asian crisis in 1997 by estimating a dynamic conditional correlation model
(DCC). Also Syllignakis and Kouretas (2009) provides further analysis of the problem of
contagion by examining correlations between seven equity markets from the CEE that have
recently became EU members. This analysis was performed using the multivariate DCC
GARCH model.
After discussing GARCH models that estimate contagion based on conditional correlations,
we should also mention the approach developed by Diebold and Yilmaz (2009), by which
they built an index that captures the volatility movement and its transmission from one
market to another. Thus, this volatility spillover index is a good measure for the contagion
effect, and it is also a way to quantify the impact of events on the equity market in one
country and the transmission through different channels to other countries’ equity markets.
Moreover, this approach used by Diebold and Yilmaz (2009) offers a possibility of exact
quantification of the contagion effect between different markets, but also of the contagion
effect at the level of a group, being able to observe this phenomenon in dynamics. This
method of analysis allows us to draw some conclusion about how certain events have
influenced the dynamics of equity markets in the recent years and the obtained results could
be very easily interpreted economically.
The experience of the most recent economic and financial crisis has shown us that markets
can be atypical in times of tension and thus the level of integration may be different from
one period to another, depending on the general economic situation. Therefore, the level of
market integration and stability may be dependent on the economic situation, but there is a
possibility that the situation may also be the opposite. We can also talk about a dependence
of the economic situation on the situation and stability of financial markets. One work that
addressed this issue is that of Mendoza et al. (2009), but the specialized literature is quite
little developed in this way.
Methodology
We will use in this analysis an approach according to (Afonso and Jalles, 2012) that will
go in two directions of analysis in order to obtain a series of conclusions regarding fiscal
sustainability. Thus, they proposed for analysis two possible relationships that could be
studied through a regression and that could provide information on fiscal sustainability. We
remind you that the analysis will be performed for the case of Romania, being a country
for which the discussions on fiscal sustainability were not numerous in the academic
environment, and the publications on this subject were quite limited.
(Afonso and Jalles, 2012) propose two types of relationships:
a) Dependent variable: primary budget balance (deficit) and independent variable: public
debt (both variables will be expressed as a percentage of GDP)
𝑝𝑟𝑖𝑚𝑏𝑢𝑔𝑑𝑒𝑓 𝛼 𝛼 𝑝𝑢𝑏𝑑𝑒𝑏𝑡 𝜀.
b) Dependent variable: public revenue and independent variable: public expenditure (both
variables will be expressed as a percentage of GDP)
𝑝𝑢𝑏𝑟𝑒𝑣 𝛼 𝛼 𝑝𝑢𝑏𝑒𝑥𝑝 𝜀.
Starting from these two, we will go further towards the realization of some tests through
which to verify a form of fiscal sustainability. These equations are constructed to illustrate
the idea of a linear relationship between variables so that later several checks can be made
using a series of concepts from the econometric literature.
First, according to (Afonso and Jalles, 2012) it can be checked the level of cointegration
between the dependent variable and the independent variable, and in this way it is sought
if there is a linear relationship between them, which would show that they will converge to
a long-term equilibrium relationship and that there is a sustainability between these
variables, i.e. between public revenues and expenditures or between the primary budget
balance and public debt. Cointegration testing can be performed using the Johansen test
implemented in Eviews.
Cointegration is a statistical property for a group of data series. Cointegration has become
an increasingly important property in contemporary data series analysis, this concept being
widely introduced by (Nelson and Plosser, 1982), noting that many time series show
stochastic trend, i.e. they have unit root or are integrated of at least order 1 (I (1)). They
also showed that processes that have a unit root do not have standard statistical properties
and cannot be modeled with classical econometric methods. Thus, they defined the
cointegration for two data series as follows: if two series are integrated by a certain order
(I (n)) and a linear combination for them is integrated by a smaller order, then it can be
stated that the series data are cointegrated.
The verification will be performed using the Johansen test for the data series used, two by
two, according to the previously defined models. The null hypotheses of this test will be
the following and will be interpreted based on calculated statistics and associated
probabilities, according to (Johansen and Juselius, 1990):
1. There are no cointegration orders.
2. There is at least one cointegration order.
3. There are at least two cointegration orders.
4. There are at least 3 cointegration orders.
Thus, this testing will be performed, and the results will provide information about the level
of cointegration of the data series, and they can be translated into a level of fiscal
sustainability according to (Afonso and Jalles, 2012). The approach was also used in a
similar way by (Gregory and Hansen, 1996) and later (Afonso and Jalles, 2012) improved
and expanded it.
If it is not possible to talk about a cointegration relationship, we will try to apply a VAR
(Vector Auto Regressive) model to analyze the relationship between the two variables and
impulse response functions. Through these impulse response functions we can see how one
of the variables reacts to a shock in another variable, and this can be very useful in our
situation when we aim to test fiscal sustainability.
Next, we will move on to the second way in which I will analyze fiscal sustainability
according to (Afonso and Jalles, 2012).This approach was developed by the two authors,
starting from a series of works in the literature. Thus, they propose to test causality in the
Granger sense for the variables considered, and the results can be put into 3 categories and
can be interpreted, especially when we talk about the relationship between public revenues
and expenditures:
a) One-way causality that can show us that the government adjusts expenditures or
revenues according to the other variable. The same interpretation applies to the
relationship between the primary deficit and public debt: the deficit or debt is adjusted
taking into account the other variable, depending on the meaning of the relationship.
b) Bidirectional causality (perfect fiscal synchronization). The existence of this result
verifies the classic hypothesis of public finances (Musgrave, 1966) according to which
decision makers correlate expenditures and revenues and make decisions that take into
account the impact on both. The same is true for the relationship between the primary
budget balance (the budget balance unaffected by interest expenditure) and the share of
public debt in GDP.
c Without any causality – this situation shows us that there are no connections and that
we cannot talk about a fiscal sustainability (situation identified by Hoover and Sheffrin,
1992) and later debated in a series of papers. This variant is consistent with the lack of
cointegration and is the most unfavorable situation, this is the situation in which it is
considered that the fiscal-budgetary policies promoted by the government are
unsustainable and will lead, sooner or later, to major imbalances in the economy.
Causality testing will be performed using the Granger test implemented in Eviews, and the
results will be interpreted from an econometric and statistical point of view, but also from
an economic point of view.
The data
For quantitative analysis based on econometric methods we will use time series for the
following variables:
 The share of public debt in GDP.
 Primary budget balance expressed as a share of GDP.
 Public revenues expressed as a share of GDP.
 Public expenditures expressed as a share of GDP.
All these data will be obtained for the case of Romania, using the databases from Ameco
and Eurostat websites. The period for which data were obtained is represented by the period
between 1998 and 2018, meaning there are 21 observations.
The data frequency is an annual one, for each of the chosen variables.
The first step in the analysis of the data series will be represented by Table 1 regarding the
descriptive statistics. The result of the calculations regarding the statistical indicators
established in the literature is presented below:
Table 1. Descriptive statistics

Primary Budget Deficit Public Debt Public Expenditures Public Revenues
Mean -1.12 26.20 36.19 32.90
Median 0.07 24.82 35.60 32.90
Std. dev 3.43 9.51 2.19 1.33
Kurtosis 0.93 -1.48 -1.04 -0.36
Skewness -1.24 -0.06 0.28 -0.02
Min -9.19 11.97 33.10 30.30
Max 2.63 39.22 40.00 35.50
Observation 21 21 21 21
Source: own computations.
Results
According to the methodology presented in the beginning of the case study, first of all, the
cointegration testing of the two groups of data series will be performed: public expenditures
and public revenues or public debt and the primary budget deficit. For this purpose, the
Johansen test in Eviews will be applied.
Before testing the cointegration, it is necessary to make sure that the tested data series are
integrated of the same order, and for this I will use the Augmented Dickey Fuller test
applied in turn for each data series.
The first relationship to be tested will be that between public debt and the primary budget
balance using the approach of authors (Afonso and Jalles, 2012).
We will test the stationarity primarily for the public debt series and for the primary budget
balance, and the results are presented in the following table:
Table 2. Stationary in level for public debt and primary budget balance series (ADF)
Public Debt Primary Budget Balance
Intercept 0.480 0.162
Trend and intercept 0.070 0.433
None 0.538 0.040
It can be seen that the probability for the ADF test is higher than 10% for 2 out of 3 results,
and which may suggest that the data series is not stationary in level. Next, we will go and
test the stationarity using the difference of order 1, meaning that we will test if the series
are integrated of order 1.
Table 3. Stationary in level for public debt and primary budget balance series (ADF)
Public Debt Primary Budget Balance
None 0.028 0.009
Following the application of the ADF test using constant, constant and trend or without any
of them, it was observed that the probability associated with the test is less than 10% for
two of these data series, meaning that we can say this time that the series of 1st order
differences are stationary and it can be said that the two data series are integrated of order
I. Therefore, it can be proceed to the cointegration test using the Johansen test according to
econometric theory. The results are presented below in the form of a summary table in
Eviews for all possibilities of testing the level of cointegration.
Table 4. Johansen cointegration test
Selected (0.05 level*) Number of Cointegrating Relations by Model
Data Trend: None None Linear Linear Quadratic
Test Type No Intercept Intercept Intercept Intercept Intercept
No Trend No Trend No Trend Trend Trend
Trace 0 0 0 1 2
Max-Eig 0 0 0 1 2
*Critical values based on MacKinnon-Haug-Michelis (1999).
It can be seen that there is a 1st order cointegration relationship, when a linear combination
is made between the two variables, if constant and trend is used in the test. Also, the results
of this test indicate that the data series are cointegrated and that a linear combination of the
two converges to a medium to long-term equilibrium level.
Therefore, it can be stated that from the point of view of the Johansen cointegration test,
the public policy promoted by the Romanian decision-makers regarding the budget balance
and the public debt was a sustainable one in a certain form. This shows that there is a linear
link between the two variables and that fiscal policy makers have taken into account
developments in the two variables to manage public debt or the primary budget balance.
The second relationship that will be tested is the one for public expenditures and public
revenues in Romania. The cointegration-based approach will be used again according to
(Afonso and Jalles, 2012). As we did in the case of the previous relationship, it will be
verified first of all that the data series have the same integration order using the ADF test.
We will present below a summary table for the probabilities for the ADF test for the two
data series using: constant, trend and constant or without either (Table 5).
Table 5. The results of the ADF test for the data series on public revenues and expenditures in Romania
Public Expenditures Public Revenues
None 0.001 0.083
The results show for both data series that they are stationary in level and it is not possible
to go further to check the cointegration. As an extension of the case study, a VAR model
can be used here to verify the relationships between these variables and to verify how
shocks are transmitted from one variable to another.
The second direction of our analysis was to test Granger causality for the two groups of
data series. The results will be interpreted according to the aspects presented in the
methodology part of the study, but we consider it necessary to remember which are the
Granger test hypotheses:
 Null hypothesis: One variable does not cause another variable in the Granger sense.
 Alternative hypothesis: One variable causes another variable in the Granger sense.
Given these hypotheses we can say that in order to validate a causality in the Granger sense
it is necessary that the probability of the test is less than the threshold of 10% to reject the
null hypothesis of the test according to which one variable does not cause another variable
in the Granger sense.
The first relationship to be tested is between the primary budget balance and the share of
public debt in GDP. The result of applying this test in Eviews is presented below:
Pairwise Granger Causality Tests
Lags: 2
Null Hypothesis: Obs. F-Statistic Prob.
PUBLIC_DEBT does not Granger Cause PRIMARY_BUDGET_DEFICIT 19 3.44751 0.0606
PRIMARY_BUDGET_DEFICIT does not Granger Cause PUBLIC_DEBT 7.99605 0.0048
Following the application of the Granger test, a causal relationship was observed from the
primary budget balance to the public debt, but also a causality from the public debt to the
primary budget balance. Thus, the primary budget balance influences the public debt in the
Granger sense and a signal can be seen for the sustainability of public finance policies
caught by this causal relationship. Also, the fact that the relationship is bidirectional shows
that public debt also causes the primary budget balance in the Granger sense.
Therefore, decision-makers must take into account and take into account in the
administration of public debt the primary budget balance, which will be reflected in an
increase in public debt, it is necessary to finance it to cover it. Thus, a certain responsibility
and fiscal sustainability at the level of public policies in Romania is highlighted.
Next, the causal relationship in the Granger sense between public revenues and public
expenditures will be tested, which are expressed as a share of GDP. The result obtained is
presented below:
Pairwise Granger Causality Tests
Lags: 2
Null Hypothesis: Obs. F-Statistic Prob.
PUB_REVENUES does not Granger Cause PUB_EXPENDITURES 19 7.05131 0.0075
PUB_EXPENDITURES does not Granger Cause PUB_REVENUES 2.23088 0.1368
Granger causality testing shows that public revenue causes public spending in the Granger
sense, and the latter does not cause public revenue. Thus, the causal relationship is one-
way. However, according to the results of the literature, this type of relationship indicates
a certain fiscal sustainability, so decision makers in Romania are taking into account public
revenues when setting the level of spending, being to some extent responsible behavior on
the part of the Government and of authorizing officers when setting the budget for the
following year.
This one-way causality is an approach to adjusting public finances according to Friedman
(1978), in our example adjusting expenditure by income level so that revenue control leads
to limited growth in the public sector: both the number of employees, as well as the level
of salary or expenses with goods and services.
Conclusion
In this article we tested two relationships: the first one is the one between public debt and
primary budget balance and the second one is between public expenditure and public
revenue, all of these variables were computed as percentage of GDP.
After we computed the ADF test, we found out that the first relationship is not stationary,
but the first differences are stationary and that the second relationship is stationarity.
Because of these ADF test results, we used Johansen test and Granger causality for first
relationship and we found out that budget balance and public debt are in close contact one
with the other and their evolution are correlated, a Granger causality in both way.
For the second relationship we used VAR (Vector Auto-Regressive). In this way, we tested
the Granger causality and we found out that only the revenues affects the expenses, not on
the other way around.
References
Afonso, A. and Jalles, J., 2012. Revisiting Fiscal Sustainability: Panel Cointegration and Structural
Breaks in OECD Countries. European Central Bank - Working Paper.
Artis, M.M., 2002. January 20-22nd. The Solvency of Government Finances in Europe. Fiscal
Sustainability. Perugia, Italy: The Bank of Italy workshop.
Ballabriga, F. and Martinez-Mongay, C., 2005. April. Sustainability of EU Public Finances,
European Economy European Commission Directorate-General for Economic and Financial
Affairs Economic Papers, No. 225.
Blanchard, O., Chouraqui, J. and Hagemann, P.a., 1990. The Sustainability of Fiscal Policy: New
Answers to Old Questions. OECD Economic Studies No. 15.
Corsetti, G.R., 1991. Fiscal Deficits, Public Debt, and Government Solvency: Evidence from OECD
Countries. NBER Working Paper No. 3658.
Gregory, A. and Hansen, B., 1996. Residual-based tests for cointegration in models with regime
shifts. Journal of Econometrics, Vol. 70, pp. 99-126.
Gupta, S.K., 2002. Fiscal Dimensions of Sustainable Development (I. M. Fiscal Affairs Department,
Ed.) Pamphlet Series, No. 54.
Hiebert, P.R., 2000. January 20-22. Close to Balance or in Surplus: a Methodology to Calculate
Fiscal Benchmarks. Fiscal Sustainability, Perugia: The Bank of Italy Workshop.
Hoover, K. and Sheffrin, S., 1992. Causation, Spending, and Taxes: Sand in the Sandbox or Tax
Collector for the Welfare State? American Economic Review, Vol. 82, pp. 225-48.
Horne, J., 1991. May. Indicators of Fiscal Sustainability. IMF Working Papers.
Johansen, S. and Juselius, K., 1990. Maximum Likelihood Estimation and Inference on Cointegration
with Applications to Demand for Money. Oxford Bulletin of Economics and Statistics, No. 52,
pp. 169-210.
Moraga, J.V., 2004. October. Fiscal Sustainability and Public Debt in an Endogeneous Growth
Model. European Central Bank Working Paper Series No. 395.
Musgrave, R.A., 1966. Principles of budget determination. Public finance: selected readings-
Random House, pp. 15-27.
Nelson, C.R. and Plosser, C.R., 1982. Trends and random walks in macroeconmic time series: some
evidence and implications. Journal of monetary economics, No. 10(2), pp. 139-162.
Roubini, N., 2001. Debt Sustainability: How to Assess Whether a Country is Insolvent? Stern School
of Business, New York, New York University.
Stoian, A.C., 2007. Fiscal sustainability based on reaction function: Case study Romania.
Macroeconomic Analysis and International Finance, Rethymno: University of Crete.

Ectap Winter - 2020

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ectap Winter - 2020

Uploaded by

Copyright:

Available Formats

Theoretical and Applied Economics

Volume XXVII, No. 4(625)/2020, Winter

What determines the portfolio investment flows

Analysis of the correlation between taxation indicators

Measuring Adobe Company performance

Welfare enhancing uncertainty

Business process modeling. Using Unified Modeling Language

Static and dynamic analysis of intra-industry trade of BRICS countries

Fossil fuel consumption, economic growth and CO2 emissions.

Methods used in risk financing

Is there a relationship between economic welfare and innovation performance?

Analysis of the natural movement of the population

Testifying the role of regulatory environment in trade facilitation:

India and trade blocs: A gravity model analysis

Predicting the volatility in stock return of emerging economy:

Portfolio optimization with VaR approach:

Income-dependent impacts of financial development

Monetary policy and exchange rate pass-through in India

Impact of medium of instruction in education on economic

Institutions and economic growth: A comparative analysis of developing

Exports, imports and economic growth in India:

BEPS – A challenge for the increasing budget revenues

Implementation of public policies.

Fiscal sustainability in Romania

Mircea Dinu Data base indexation:

Reception of texts: economia.ta@edeconomica.com

ISSN 1841-8678 (Print)

The European Union in the face of the prospect of achieving

Gabriela Victoria ANGHELACHE

Keywords: European Union, economic objectives, challenges, economic growth, harmonization of

JEL Classification: F02, F40.

Data, results and discussions

Simultaneously with the reduction of production potential, the reduction of agricultural

Source: Eurostat (data are processed by the authors).

Source: Eurostat (data are processed by the authors).

Source: Ministry of European Funds.

What determines the portfolio investment flows

Keywords: economic integration; European Union; portfolio investment; financial market.

JEL Classification: F15; F36; G15.

to establish proper institutional and corporate governance framework. Better market

Analysis of the correlation between taxation indicators

Keywords: causality, taxation, products, imports, GDP.

JEL Classification: H71, O40.

Introduction and literature review

Research methodology and data

Based on the indicators, two research hypotheses have been defined:

Results and discussions

2.2. Logarithm values

LROG = C(2,1) × LRUIT(-1) + C(2,2) × LRUIT(-2) + C(2,3) × LROG(-1) + C(2,4) ×

3. H2 testing – European Union

3.2. Logarithm values

4.2. Logarithm values

Measuring Adobe Company performance

Keywords: SWOT analysis, key performance indicators, BCG matrix.

JEL Classification: A10, G12, G17, L25, M2.

2. Complex Adaptive System in enterprise

3. Analysis and diagnostic techniques applied in an enterprise

Source: The design is done by the authors.

4. Case study: Diagnosis and analysis of key performance indicators in Adobe

Source: Refinitiv Eikon Thomson Reuters.

4.3. SWOT Analysis

Figure 3. Probability density of Adobe stock returns

Source: Processing authors in Crystal Ball.