Professional Documents
Culture Documents
Lombaerde
Lombaerde
Lombaerde
Philippe De Lombaerde
Edgar J. Saucedo Acosta Editors
Indicator-Based
Monitoring of
Regional Economic
Integration
Fourth World Report on Regional
Integration
United Nations University Series on Regionalism
Volume 13
Series Editors
Philippe De Lombaerde, NEOMA Business School, Rouen (France) and UNU-CRIS,
Bruges (Belgium)
Luk Van Langenhove, Grootseminarie, United Nations University CRIS,
Bruges, Belgium
Indicator-Based Monitoring
of Regional Economic
Integration
Fourth World Report on Regional Integration
Editors
Philippe De Lombaerde Edgar J. Saucedo Acosta
NEOMA Business School Universidad Veracruzana
Rouen, France Xalapa, Veracruz, Mexico
UNU-CRIS
Bruges, Belgium
v
vi Volumes in This Series
3 – The United Nations and the Regions. Third World Report on Regional
Integration
Lombaerde, Philippe, Baert, Francis, Felı́cio, T^ania (Eds.), 2012
2 – Aid for Trade: Global and Regional Perspectives. 2nd World Report on
Regional Integration
Lombaerde, Philippe, Lakshmi, Puri (Eds.), 2009
1 – Multilateralism, Regionalism and Bilateralism in Trade and Investment.
2006 World Report on Regional Integration
De Lombaerde, Philippe (Ed.), 2007
Acknowledgments
The editors wish to thank the United Nations University Institute on Comparative
Regional Integration Studies (UNU-CRIS) for its institutional support; the FWO
Research Network (WOG) on Globalisation, Regionalisation and Economic and
Social Inequality (GRESI) for financially supporting an author’s workshop; the
Institute for European Studies (IES) of the Free University of Brussels (Vrije
Universiteit Brussel or VUB) for hosting an author’s workshop; Lien Jaques
(UNU-CRIS) for her help with formatting the manuscript; and Esther Otten and
Hendrikje Tuerlings at Springer.
vii
Contents
ix
x Contents
Part IV Asia
13 Monitoring the ASEAN Economic Community . . . . . . . . . . . . . . . . 287
Aladdin D. Rillo
Part V Methodology
14 Opening the Black Box of Trade Agreements . . . . . . . . . . . . . . . . . 301
Tristan Kohl
15 Assessing Globalization and Regionalization Through
Network Indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
P. Lelio Iapadre and Lucia Tajoli
16 Measuring Actual Economic Integration: A Bayesian
State-Space Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
Glenn Rayp and Samuel Standaert
About the Authors
Rattan J. Bhatia is a former Deputy Director of the African Department of the IMF
and Special Representative and Director of the IMF Office at the United Nations,
New York (USA). He is currently a Senior Consultant to African governments, the
IMF, World Bank, UNDP, and the African Development Bank (AfDB).
Bruce Byiers is Senior Policy Officer at the European Centre for Development
Policy Management (ECDPM), Maastricht (the Netherlands).
Carlos Cuerpo Spanish Fiscal Council, Madrid (Spain), was formerly with the
European Commission, Directorate General for Economic and Financial Affairs,
Brussels (Belgium).
Jonas Fischer is Deputy Head of Unit, Directorate General for Economic and
Financial Affairs, European Commission, Brussels (Belgium).
J€
org K€ onig is Researcher at the Market Economy Foundation (Stiftung
Marktwirtschaft), Berlin, and former Researcher at the Georg-August University
of G€
ottingen (Germany).
Alfred Ombudo K’Ombudo is Lead Author and Project Manager, EAC Common
Market Scorecard 2014, Advocate of the High Court of Kenya, and Lecturer at US
International University, Nairobi (Kenya).
Joaquı́n Roy is Jean Monnet Professor and Director of the University of Miami
European Union Center of Excellence, University of Miami (USA).
Daniel Tanoe is the Chief of the Investment Policy Section in the Regional
Integration and Trade Division of the Economic Commission for Africa
(UNECA), Addis Ababa (Ethiopia). He was formerly the Chief of the Regional
Integration Section in the same division, in charge of the ARIA publication series.
Over the last decades, economic integration processes have proliferated in various
regions of the world, in search of appropriate conditions and institutional contexts
within which goods, services, capital and people can circulate more freely and con-
tribute to reaching higher growth. In addition, regional economic cooperation has been
organized in various policy areas, going from cooperation in building large-scale
infrastructures, over the design of regional socio-economic policies, to cooperation to
stabilize financial markets. Whether these regional policy packages are effective or not
is a matter of public, political and academic debates. It is in this context that monitoring
tools have been developed. Monitoring can be understood here as referring to:
The processes carried out by national and regional public and private sector institutions –
rules, roles, and actual physical organizations – to ensure and/or scrutinize the implemen-
tation of the regional integration process. Monitoring takes place in the ‘monitoring
system’, or the framework of national and regional public and private sector institutions
that are involved and/or employed in monitoring. [. . .] the purpose of monitoring is to
ensure that the contractual obligations assumed by all parties will be implemented. How-
ever, over time, monitoring can also come to include activities aimed at propelling the
development of new strategies and initiatives that guide regional integration project in the
right directions. And it may imply attending to the intended or unintended effects of
regional integration policies and their contribution to overall developmental goals.
[It] thus touches on various policy levels, and can take place in different stages of
integration and with various degrees of institutionalization. (De Lombaerde et al. 2008a)
In the same context, monitoring has also been defined as covering “all relevant
processes of information gathering, processing and dissemination concerning the
[. . .] integration process, performed by different kinds of actors in different moments
and lapses of time, in order to control, evaluate, correct and/or influence the integra-
tion policies and the functioning of the regional institutions” (Costea et al. 2008).
In addition, the relationship between monitoring and planning is relevant:
[. . .] monitoring provides information on achievements and/or shortfalls against planned
and agreed targets and/or indicators. Information from monitoring (and evaluation) on
implementation, outcomes (and impact) is vital to gain insight into the progress made in
implementing the REC and the status and depth of the regional (economic) integration
xv
xvi Introduction: Indicator-Based Monitoring of Regional Economic Integration
realised. Both monitoring and evaluation are important to identify necessary corrective
actions - but cover different levels and criteria. (Zelenka 2015)
1
Inter-American Development Bank.
2
The case studies that were undertaken in this project analysed the following: (i) the actors
involved (regional organizations, international/multilateral organizations, national authorities,
academics, civil society, etc.), instruments used (review processes, indicators, questionnaires,
etc.) and outputs generated (internal circulars, external reports, yearbooks, etc.) in the monitoring
processes in the respective regions; (ii) the specific institutional and/or legal structures and
instruments built-in in the agreements (or institutions) for monitoring purposes; (iii) the underlying
conceptual and methodological frameworks; (iv) the policy relevance and effectiveness of mon-
itoring; (v) specific technical issues (problems and solutions) that are of relevance to other regions;
(vi) the factors underlying the lack or deficiencies in monitoring; (vii) and the monitoring needs
that can be identified for the future. The case studies and general conclusions of the project were
published in De Lombaerde at al. (2008a). The case studies included the Andean region, Carib-
bean, Central America, Southern Cone (Mercosur), Southeast Asia, Pacific Islands, South Asia,
Gulf region, Maghreb, Eastern and Southern Africa, Europe (European Union) and North Amer-
ica. The case study on Eastern and Southern Africa was published as Hahnsohm and Adongo
(2008).
Introduction: Indicator-Based Monitoring of Regional Economic Integration xvii
worth to recall some of the main conclusions of this study (De Lombaerde
et al. 2008a):
1. Monitoring has grown more challenging with the increased complexity and
broadening of economic integration agreements.
2. The scope of monitoring has expanded well beyond coordinating the implemen-
tation of the integration commitments, it includes an important external com-
munication component with different types of stakeholders and the monitoring
tasks start before the to-be-monitored agreements are signed and extend beyond
the formally established implementation periods.
3. The odds of successful monitoring are vastly improved when the integration
agreement carries a clear, built-in agenda and processes for its administration
and implementation, and when it mandates the establishment of monitoring units
in each of the member states to interact with each other.
4. Monitoring – particularly of new regional schemes – starts at the national level.
5. Effective monitoring carried by a regional organization reduces coordination
and communication costs among member states and alleviates the workloads of
the individual member governments, although coordination costs and informa-
tional demands can be relatively high for the regional organization.
6. When the regional integration process is deepened and accompanied by the
building of relatively autonomous supranational institutions, new regional mon-
itoring mechanisms are put in place that not only perform technically sophisti-
cated monitoring tasks but that also start playing a more independent and
political role.
7. In deeper integration schemes not only are regional monitoring mechanisms put
in place, but the regional institutions themselves tend to develop and take new
monitoring initiatives, start to monitor member states’ compliance with the
regional rules and start to take overall control over the monitoring process,
resulting in a vertical two-way interaction.
8. In a developing country context, the role of external donors can be crucial
because of the public good characteristics of monitoring; they can raise the
technical capacity of the poorly performing member states, pressure reluctant
states not to oppose the development of monitoring instruments and reinforce the
impartiality of monitoring agencies; at the same time it is also a fact that when
monitoring processes are driven by external actors, the sustainability and rele-
vance of the whole effort is not guaranteed.
When looking at the monitoring systems that are in place in various institution-
alized world regions, in only a limited number of cases they involve systematic
indicator-based monitoring, and most of these cases are one-shot efforts which are
not sustained over time.
In these cases, attempts have been made to monitor the de facto and/or de jure
integration process with the help of a series of indicators.3 As with monitoring
3
On the de facto and de jure categories in regionalism studies, e.g. see Higgott (1997).
xviii Introduction: Indicator-Based Monitoring of Regional Economic Integration
regional integration in general, the experiences have been quite diverse and, so far,
their results have been mixed.4 Considerable human and financial resources have
been invested in these attempts (by both intra- and extra-regional actors) but few
have been sustained. The EU Internal Market Scoreboard, the ASEAN Economic
Community Scorecard and the EDB System of Indicators of Economic Integration
are some of the few exceptions that confirm the rule.
The variety of the experiences reflects, in the first place but not exclusively, the
observed variety in terms of complexity, depth and institutionalization of the
respective regionalization processes worldwide. Cases of regionalism vary from
simple “projects” of regional cooperation, of limited scale and importance, to large-
scale regional processes of societal transformation such as in the case of the
European Union. This obviously poses a problem of comparability, both from an
academic and from a policy perspective.5 It also explains why monitoring in some
cases can be understood in terms of a project management logic6, while in other
cases monitoring refers to an open-ended, complex and multilevel combination of
various (connected or disconnected) monitoring tools and practices in a multi-
stakeholder environment, mirroring the complexity of the process of regional
integration.7
Different cases of monitoring cannot only be distinguished by their complexity,
but specific tools can also be distinguished by their focus, scope and ambition. A
monitoring system can first of all be designed to monitor the implementation of
regionally agreed commitments (i.e. compliance) in one or more policy areas, in
principle by national authorities.8 It has been observed that in many regional
cooperation or integration schemes, this is exactly the crucial success factor for
the respective process (De Lombaerde et al.2008a). Other monitoring systems
(also) focus on aspects of the de facto regionalization process, the behaviour of
regional actors, the institutionalization process, regional policymaking and/or –
which is more complicated and ambitious – the effects of regional policies.9
Whether to include the latter category (policy effects) in a monitoring system is
directly related to the discussion on the limits of monitoring and on how far a
monitor should/could go into the analysis of the collected data.
4
For an overview of general (i.e. qualitative and quantitative) monitoring experiences in various
world regions, see De Lombaerde et al. (2008a).
5
Whether the EU is a sui generis case or not is an old debate in regionalism studies. See, for
example, Sbragia (1992), Hix (1994), Caporaso (1997), Marks (1997), Moravcsik (1997), Pollack
(1997), Checkel (2007), Warleigh-Lack and Rosamond (2010), among many others. On the issue
of comparison and comparability, more in general, see De Lombaerde and S€ oderbaum et al.
(2010).
6
For an overview of applicable tools, see, for example, World Bank (2004).
7
See, for example, De Lombaerde et al. (2008a).
8
In some cases, subnational authorities can directly engage in cross-border agreements; in other
cases, hybrid actors are involved.
9
For a further discussion of these categories, see De Lombaerde and Van Langenhove (2006).
Introduction: Indicator-Based Monitoring of Regional Economic Integration xix
10
Indicators that are used for this purpose include transposition deficits, changes in backlogs of
directives not communicated, fragmentation factor and long-overdue directives.
11
Indicators that are used for this purpose include absolute figures and changes of the number of
infringement proceedings initiated by the European Commission against member states.
12
The same is true for the exercise performed by the ECB on Mercosur (as compared to the EU)
(Dorrucci et al. 2002).
xx Introduction: Indicator-Based Monitoring of Regional Economic Integration
with CAN and Central America, based on the Madrid Declaration of 2002 (Grupo
de Trabajo Conjunto CA-UE 2005a,b,c; Joint Working Group EU-CAN 2006).
Contrary to its negotiations with the Latin American subregions, the negotiations
with the ACP subregions usually involve ex post monitoring. The so-called
EU-ACP Reviews reflect the EU’s integration-promotion strategy as it was incor-
porated in the Cotonou Agreement (European Commission 2002; European Com-
mission 2005b; Pietrangeli 2010).13 In addition, feedback mechanisms were
explicitly foreseen from the (mid-term and end-term) reviews on the (re-)
formulation of the regional indicative programmes (RIPs) and the (re-)allocation
of financial resources.14
This policy has been continued when negotiating Economic Partnership Agree-
ments (EPAs) with the various ACP subregions. The CARIFORUM-EU Economic
Partnership Agreement, signed in 2007 as the first region-to-region EPA, also
included monitoring provisions, not only referring to implementation stricto
sensu but also referring to impact on the broader development goals of the agree-
ment (see Chap. 7).15 In addition, according to the text of the Joint Declaration
annexed to the EPA, monitoring could in principle lead to adjustments in the
agreement. As the promotion of regional integration is one of the general objectives
of EU-ACP cooperation (see above) and of this specific EPA, monitoring the
Caribbean regional integration process is a logical component of the monitoring
exercise.16 However, as implementation of the EPA has been slow, setting up a
monitoring system is also not a reality yet.
Standing a bit apart in this context is the EU-funded ACP Monitoring Regional
Integration (ACP-MRI) project (Chap. 10). The ACP-MRI project was a two-year
ACP Secretariat project funded by the EU. Launched in November 2008, it aimed at
designing an information system for the measurement and monitoring of regional
integration in the ACP subregions.17 The project was conceived as a two-stage
project in which, in a first stage, a system of core indicators would be developed,
common to all the ACP regional integration organizations. In a second stage, the
system would be expanded with other indicators specific to each regional organi-
zation on the basis of their own agendas. The ACP-MRI is an interesting project,
not only from a technical point of view but also from a governance point of view.
13
See articles 28–30 of the Cotonou Agreement and articles 6–14 of Annex IV.
14
Although the Cotonou agreement did not explicitly require annual operational reviews, such
reviews were organized since 2003 for each of the programming regions in accordance with the
principle of rolling programming and by analogy with the country strategy paper review process.
15
On the role and methodology of monitoring in a context of EPAs, see Brüntrup et al. (2008).
16
Institutionalized regional integration in the Caribbean mainly involves the Caribbean Commu-
nity (CARICOM) and the Organization of Eastern Caribbean States (OECS).
17
These subregions were the 12 regional integration organizations recognized by the ACP Group
and benefiting from EDF funds. They included regional organizations in Sahel-Saharan Africa
(CEN-SAD), Southern Africa (SADC), Central Africa (CEMAC, CEEAC), Western Africa
(CEDEAO, UEMOA), Eastern and Southern Africa (COMESA, COI, EAC, IGAD), the Caribbean
(CARICOM) and the Pacific (Pacific Islands Forum). See: http://mri.acp.int/.
Introduction: Indicator-Based Monitoring of Regional Economic Integration xxi
The project cycle involved a lot of multilevel consultations, but, apparently, due to
a lack of ownership at the level of the regional organizations, the project was not
continued after the first phase and the elaborated indicator system was not adopted
by the regional organizations.
The EU-ACP cases, in contrast with the Single Market Scoreboard, are cases
where other dimensions than implementation acquire importance. Other regional
organizations involved in monitoring their regional integration processes focus
even less on implementation, to the benefit of other dimensions such as the
institutionalization of the process or de facto regionalization. Examples include
the indicator system developed by Dennis and Yusof for ASEAN and the UNECA
proposal for the African subregions (Chap. 9).
An ambitious proposal of an indicator-based monitoring system was the one
proposed by Dennis and Yusof for the ASEAN Secretariat and funded by the
Australian Regional Economic Policy Support Facility (REPSF) (Dennis and
Yusof 2003). Its objective was to measure “the progress towards economic inte-
gration of the ten ASEAN nations in the context of the aim to move towards an
ASEAN Economic Community” (Dennis and Yusof 2003:1). The proposal incor-
porated a comprehensive set of indicators, focusing on the following areas: trade in
goods, investment, trade in financial and other services, infrastructure, customs,
standards, mutual recognition agreements and conformity assessment, small and
medium enterprises, e-ASEAN and intellectual property.18 Although this report can
be considered as an important contribution to the development of indicator systems,
this monitoring system has not been implemented. However, in the report, one finds
already a suggestion to adapt the EU Single Market Scoreboard (see above) to the
ASEAN context, i.e. taking into account the intergovernmental structure of ASEAN
and avoiding the public and “confrontational” aspects of the scoreboard (Dennis
and Yusof 2003: 33). More recently the ASEAN Secretariat has built up its own
monitoring capacity, which has led, among other things, to the development of an
“ASEAN Economic Community Scorecard” (ASEAN 2012). This scorecard has
been published yet and plans exist to strengthen the capacity of the monitoring
office (with the support of the EU). However, according to Rillo and De la Cruz
(2016), monitoring in Asia is limited due to “the lack of a systematic approach to
measure its progress. This is unfortunate because integration has now become an
important policy tool in many Asian countries”.
Another proposal was the one launched by UNECA in 2001, with the aim to
assess overall regional integration in the regional economic communities (RECs) in
Africa and to analyse the performance of each country in the region (individually
and relative to other member countries) in achieving the specific objectives set by
the treaties (UNECA 2001, 2002, 2004; see also Chap. 9). Monitoring focused on
the progress made after the African Economic Community was established by the
Abuja Treaty (signed in 1991). The indicators were based on eight sectors: trade,
money and finance, transport, communications, energy, agriculture, manufacturing
18
A smaller set of “key integration indicators” was also identified.
xxii Introduction: Indicator-Based Monitoring of Regional Economic Integration
and human development and labour markets. Based on eight sectoral indices, a
composite integration index was calculated which assessed the relative perfor-
mance of a regional economic community.19
Another interesting proposal is the SADC monitoring reporting and evaluation
systems (Zelenka 2015). These systems are focused on finance and investment
(Protocol on Finance and Investment) and trade (Trade Protocol). The monitoring
systems have been financed by the GIZ Promotion of Economic Integration and
Trade programme.
There are some other cases to be mentioned that are broader in scope than the
indicator systems focusing on the implementation dimension. One is Hufbauer and
Schott’s proposal to assess regional integration in the Americas (Hufbauer and
Schott 1994). Two sets of indicators were developed by these authors: one assessing
the level of economic integration achieved by each subregional grouping and the
other examining the level of “readiness” of these groupings in order to increase the
degree of hemispheric integration. Another case is the indicator system proposed by
Feng and Genna (2003, 2004, 2005) and which is directly based on Hufbauer and
Schott’s (see Chap. 6 in this volume). The authors use “integration achievement
scores” and applied them to African, Asian and Latin American regional integration
processes. A third case is Ruı́z Estrada’s Global Dimension of Regional Integration
model which analyses the process of regional integration from a global perspective
using a multidimensional framework covering social, political, economic and
technological dimensions (Ruı́z Estrada 2004).20
A fourth case is UN-ESCWA’s regional integration index for the Arab world,
which was published for the first time in its 2006 Annual Review of Developments
in Globalization and Regional Integration in the Arab Countries (UN-ESCWA
2007). The index seeks to measure the degree of de facto integration of individual
Arab countries in the region.21 A fifth case is the ADB proposal. The indicator
system proposed by Capannelli, Lee and Petri (2009, 2010) at the ADB seeks to
19
The main objectives of the indices were (i) “ [t]o assess each country’s performance and relate it
to the goals and objectives of each regional economic community and that of Africa as a whole, as
well as to assess the performance of each economic community to that of Africa; (ii) to compare
the contributions of each member country in a regional economic community towards the
realization of such goals and objectives, in addition to the contributions that each regional
economic community has made towards the realization of goals and objectives of the continent
at large; (iii) to monitor the performance of each country, regional economic community, and the
continent as a whole for regional integration efforts over time; (iv) to enhance the quality of the
analysis by providing indices for scores and rankings at country, regional economic community
and continent levels” (UNECA 2004: 244).
20
The model also involves the calculation of the “regional integration stage index” (Ruı́z Estrada
2004:15).
21
In its current version, only four variables are used, mainly because of severe data constraints in
the region. These variables (or sub-indicators) are openness to Arab intra-regional trade, openness
to Arab intra-regional investment, openness to Arab intra-regional workers’ remittances and
openness to Arab intra-regional tourism. Country rankings are aggregated using a statistical
weighting procedure (principal component analysis).
Introduction: Indicator-Based Monitoring of Regional Economic Integration xxiii
measure regional integration in a reflexive and comparative way with the help of
quantitative indicators of market integration and policy cooperation.22 Quantitative
input indicators of political and cultural similarity are calculated to explain the
disequilibrium between institutional and economic integration in East Asia. Since
2012, the ADB also publishes an Asian Economic Integration Monitor (ADB
2012).23
Finally, also the EDB proposal falls in this category of broader indicator-based
monitoring systems (Vinokurov et al. 2010; Libman and Vinokurov 2010; EDB
2012; Chap. 5 in this volume).24
Some other initiatives are in the pipeline. They include a new attempt by
UNECA to revive its indicator system in support of the Abuja Treaty process, the
plan for an East African Community Common Market Scorecard supported by the
World Bank in collaboration with the EAC Secretariat (in order to monitor the
implementation of the Common Market Protocol) (Chap. 11), UN-ECLAC’s plans
to monitor Central American economic integration and monitoring plans at
COMESA.
A systematic review of the relevant indicator systems that have been designed so
far (see above) has been undertaken at UNU-CRIS.25 The objective of the review
was to map and compare the indicator systems, to study their technical features and
to identify best practices but also to look at the political-economy aspects of the
systems (by whom, for whom, for what purpose). The conclusions of the reviews
can be summarized as follows (De Lombaerde et al. 2008b; De Lombaerde et al.,
2010, 2011):
22
See also Capannelli (2012).
23
This new semi-annual publication evolved from the Asia Economic Monitor, which was
published since 2001.
24
See also Libman and Vinokurov (2012a) and Libman and Vinokurov (2012b: 72-85).
25
The review was quite exhaustive and included both institutional systems and purely academic
proposals. The former included systems developed by the European Central Bank (ECB)
(Dorrucci et al., 2002), various schemes proposed and/or implemented by the European Commis-
sion and its regional partners in the framework of interregional relations; UNECA’s proposal to
monitor regional integration in Africa (UNECA, 2001, 2002, 2004); the indicator system proposed
for ASEAN (Dennis and Yusof, 2003); and the Eurasian Development Bank (EDB, 2009). The
schemes proposed by the European Commission include the EU-MERCOSUR “Joint Photogra-
phy” (European Commission, 1998), the EU-CAN Joint Evaluation (Grupo de Trabajo UE-CAN,
2005a,b), the EU-Central America Joint Evaluation (Grupo de Trabajo Conjunto CA-UE, 2005a,b,
c) and the EU-ACP Reviews (European Commission, 2002; European Commission, 2005b; World
Bank, 2002; COMESA, 2002). In addition, some “academic” proposals were included in the
sample: Hufbauer and Schott’s proposal to assess regional integration in the Americas (Hufbauer
and Schott, 1994); its modified version by Feng and Genna (2003, 2004, 2005); Ruiz’ GDRI model
(2004a); and the proposal by Capannelli, Lee and Petri (2009) to measure economic integration
and cooperation in East Asia. The results of this review were published in De Lombaerde et al.
(2008b); De Lombaerde et al. (2010); De Lombaerde et al. (2011); and De Lombaerde and Salmon
(2011).
xxiv Introduction: Indicator-Based Monitoring of Regional Economic Integration
1990
1995
2000
2005
2010
2015
Hufbauer
and
Schott
SADC-
UNDP
EU-IMS
EU-M
ECB
Feng
and
Genna
UN
ECA
ASEAN
GDRI
EU-
CAN
EU-CA
UN
ESCW
A
ADB
EDB
ACP
MRI
WB
EAC
= sustained = not sustained
.
Fig. 1 Lifetime of selected indicator systems (Sources: see text and footnotes)
monitor the economic integration of countries and groups of countries in the post-
Soviet space. The system is built around indicators of regional integration in trade,
labour migration, electricity (electric power), agriculture and education to assess
the level and dynamics of market integration. Vinokurov and Libman also offer a
summary of the outcomes of the two SIEI published so far (2009 and 2014). After
presenting a consolidated index for 12 countries of the Commonwealth of Inde-
pendent States, it seems on the one hand that the level of integration has decreased.
On the other hand, EurAsEC-5 has become more integrated in the last decade.
However, from 2009 to 2012 the trends have inversed. In terms of countries,
Kyrgyzstan, Armenia and Tajikistan have taken the lead of integration, strongly
contrasting with Russia and Moldova.
Africa
Methodology
The final three chapters discuss methodological issues that are of relevance across
regional systems. In Chap. 14 (“Opening the Black Box of Trade Agreements”),
Kohl presents a coding methodology to capture the heterogeneity of trade agree-
ments and to facilitate quantitative analysis departing from qualitative legal differ-
ences in trade agreements. The coding is based on whether the provisions of the
World Trade Organization (WTO) are covered by free trade agreements, as well as
whether they are legally enforceable. Then, the author offers a heterogeneity index
that is used to analyse the content of a trade agreement. The index can be used in
gravity models to estimate the effect of FTAs on trade.
In Chap. 15 (“Assessing Globalization and Regionalization Through Network
Indices”), Iapadre and Tajoli use network analysis (NA) in order to study the impact
of globalization and regionalization on the entire structure of trade flows. NA
focuses on trade flows as a network, therefore emphasizing the relationship between
countries – the nodes – and the network structure itself. According to Iapadre and
Tajoli, this approach is particularly fit to offer a unified view of the system’s
properties, help develop trade policies and analyse changes in the world trading
system. This is why the authors present the changes in the trade networks by
studying indices that describe the network’s characteristics: density, closeness,
betweenness and degree distribution. The authors conclude that network complete-
ness is indeed achieved in some subregional components; however, the level of
heterogeneity between countries has increased.
Finally, in Chap. 16 (“Measuring Actual Economic Integration: An Outline of a
Bayesian State-Space Approach”), Rayp and Standaert construct an indicator of
regional integration based on a Bayesian state-space approach. The state-space
model is helpful in estimating the overall level of regional integration by using all
information contained in a set of indicators. The authors apply this model to the
level of regional integration between members of the OECD. The variables of the
level of regional integration – i.e. the Current Economic Integration (AEI)
(Mongelli et al. 2005) – are standardized and organized in four groups: flows of
goods, flows of services, foreign direct investment (FDI) and other financial flows
and migration. The AEI can also be used to construct a weighted directed network.
By observing the weighted directed network, the authors found that the core players
in the OECD are the United States, Germany and the United Kingdom and in
second place, France, Italy and Japan. Finally, they conclude that the level of
economic integration among OECD members has increased over the last 20 years
and that the European integration agreements and the NAFTA have had positive
effects on economic integration.
Introduction: Indicator-Based Monitoring of Regional Economic Integration xxxi
Conclusion
Whereas most analysts and policymakers will probably not question the potential of
indicator-based monitoring systems for supporting regional economic integration
processes, the current landscape is patchy and the experiences so far have not
always been successful. The monitoring efforts have very often not been sustained
over time, and it has proven difficult to secure the involvement of the relevant
stakeholders and generate visible impact in policymaking. However, cases like the
Internal Market Scoreboard in the EU show that sustained efforts in combination
with sound methodologies and clear communication clearly have an important
potential. Especially the regularly published transposition deficits have become
often cited figures in national parliaments and the mass media. The IMS is appar-
ently contributing to pressuring member states to implement timely and correctly
the regionally decided rules.
At the same time, when thinking about designing “better” indicator systems, one
should be very much aware of the distinctive features of each individual case of
regional integration. The mentioned IMS responds to a very specific monitoring
mandate, whereas the mandates can greatly vary from one situation (region) to
another and they can involve very different sets of actors and stakeholders. This
reflects in part the fact that the world of regionalism is very heterogeneous.
One-size-fits-all monitoring tools are therefore not recommended.
In addition, one should realize that designing an indicator system is not only
about making technical choices and solving data constraint problems but also about
carefully dealing with political-economy and governance aspects.
We hope that this book contributes to building the collective memory of
indicator-system builders, to exchanging experiences among them and to preparing
the ground for building more effective systems in the future.
References
ADB. (2012, July). Asian economic integration monitor. Mandaluyong City: Asian Development
Bank.
ASEAN. (2012). ASEAN economic community scorecard. Charting progress Toward regional
Economic Integration. Phase I (2008–2009) and Phase II (2010–2011), Jakarta: ASEAN
Secertariat.
Beck, U. (1999). What is globalization? Cambridge: Polity Press.
Brüntrup, M., et al. (2008). Monitoring economic partnership agreements. Inputs to the negotia-
tions and beyond. DIE Studies, 37.
xxxii Introduction: Indicator-Based Monitoring of Regional Economic Integration
Capannelli, G., Lee, J.-W., & Petri, P. A. (2009). Developing indicators for regional economic
integration and cooperation, UNU-CRIS Working Papers, (W-2009/22).
Capannelli, G., Lee, J.-W., & Petri, P. A. (2010). Economic Interdependence in Asia: Developing
indicators for regional integration and cooperation. Singapore Economic Review, 55(1),
125–161.
Capannelli, G. (2012) Recent trend of Asian economic regionalism, Paper presented at Expert
Workshop on Indicator-Based Monitoring of Regional Economic Integration, IES-VUB,
Brussels, October. 25–26.
Caporaso, J.A. (1997) Does the European Union Represent an n of 1?. ECSA Review, X(3): 1–2.
Checkel, J. T. (2007). Social mechanisms and regional cooperation: Are Europe and the EU really
all that different? In A. Acharya & A. I. Johnston (Eds.), Crafting cooperation. Regional
International Institutions in Comparative Perspective (pp. 221–243). Cambridge: Cambridge
University Press.
Costea, A.-C., De Lombaerde, P., De Vriendt, W., & Fühne, B. (2008). Monitoring and (Good)
Governance of the integration process in the European Union. In P. De Lombaerde,
A. Estevadeordal, & K. Suominen (Eds.), Governing regional integration for development:
Monitoring experiences, methods and prospects (pp. 211–245). London: Ashgate.
Cuerpo, C. et al. (2012) Scoreboard for the surveillance of macroeconomic imbalances (European
Economy Occasional Papers, (92)).
De Lombaerde, P., & Van Langenhove, L. (2006). Indicators of regional integration: Conceptual
and methodological aspects. In P. De Lombaerde (Ed.), Assessment and measurement of
regional integration (pp. 9–41). London: Routledge.
De Lombaerde, P., Estevadeordal, A., & Suominen, K. (Eds.). (2008a). Governing regional
integration for development: Monitoring experiences, methods and prospects (290 pp).
London: Ashgate.
De Lombaerde, P., Pietrangeli, G., & Weeratunge, C. (2008b). Systems of indicators for moni-
toring regional integration processes: Where do we stand? Integrated Assessment Journal, 8
(2), 39–67.
De Lombaerde, P., & Schulz, M. (Eds.). (2009). The EU and World regionalism. The makability of
regions in the 21st century. Farnham-Burlington: Ashgate.
De Lombaerde, Guinea, P. O., Pietrangeli, G., & Weeratunge c. (2010a) Systems of indicators for
monitoring regional integration processes: Where Do We stand Anno 2010?. K€ oz-Gazdasag –
Economic Theory and Policy, 5(3): 145–171. (Special issue on Evaluation of EU Membership
of New Member States)
De Lombaerde, P., S€oderbaum, F., Van Langenhove, L., & Baert, F. (2010b). The problem of
comparison in comparative regionalism. Review of International Studies, 36(3), 731–753.
Dennis, D. J., & Yusof A. Z. (2003) Developing indicators of ASEAN integration – A preliminary
survey for a road map Final Report.
Dorrucci, E., Firpo, S., Fratzscher. M., & Mongelli, F.P. (2002) European integration: What
lessons for other regions? The case of Latin America, ECB Working Paper, (185).
EDB. (2012). EDB system of indicators of Eurasian Integration, Saint Petersburg: EDB. http://
www.eabr.org/e/research/centreCIS/projectsandreportsCIS
European Commission. (1997). Single market scoreboard, European Commission, Brussels, (1).
European Commission. (1998). EU-MERCOSUR joint photography. DG Relex, European Com-
mission: Brussels.
European Commission. (2002). Informal Workshop to Discuss the Regional Cooperation Review
Process, Meeting Report, DG Development, European Commission, Brussels (DEV/B/1/WK
D(2002)).
European Commission. (2005a). Internal Market Scoreboard, DG Internal Market, European
Commission, Brussels, (14).
Introduction: Indicator-Based Monitoring of Regional Economic Integration xxxiii
European Commission. (2005b). Guidelines for 2005 Mid-Term Reviews in ACP regions under the
ACP-EC partnership agreements, DG Development, European Commission, Brussels.
European Commission. (2011). Scoreboard for the surveillance of macroeconomic imbalances:
envisaged initial design (Commission Staff Working Paper (SEC(2011) 1361)).
Feng, Y., & Genna, G. M. (2003). Regional integration and Domestic Institutional Homogeneity:
A comparative analysis of regional integration in the Americas, Pacific Asia and Western
Europe. Review of International Political Economy, 10(2), 278–309.
Feng, Y., & Genna, G. M. (2004). Domestic institutional convergence and regional integration:
Further evidence. In I. D. Salavrakos (Ed.), Aspects of globalization regionalisation and
business (pp. 23–37). Athens: Atiner.
Feng, Y. & Genna, G. M. (2005). Measuring Regional Integration, paper presented at the
Claremont Regional Integration Workshop with Particular Reference to Asia, Claremont
(25 February).
Grupo de Trabajo Conjunto CA-UE. (2005a). Primera reuni on del grupo de trabajo consunto
ad-hoc Centroamérica-Uni on Europea – Fase de Valoraci on Conjunta, San José, (31/03–1/
04).
Grupo de Trabajo Conjunto CA-UE. (2005b). II reunion del grupo de trabajo consunto ad-hoc
Centroamérica-Union Europea – Fase de Valoraci on Conjunta, Bruselas, 19–20/07. http://
europa.eu.int/comm/external_relations/ca/doc/05_doc1_19-07-05.pdf
Grupo de Trabajo Conjunto CA-UE. (2005c). III Reuni on del Grupo de Trabajo Conjunto ad-hoc
Centroamérica–Uni on Europea Fase de Valoraci on Conjunta, El Salvador, 7–8/11. http://
europa.eu.int/comm/external_relations/ca/doc/san_sal_nov05.pdf
Higgott, R. (1997). De Facto and De Jure Regionalism: The double discourse of regionalism in the
Asia Pacific. Global Society, 11(2), 165–183.
Hix, S. (1994). The study of the European Community: The challenge to comparative Politics.
West European Politics, 17(1), 1–30.
Hufbauer, G. C., & Schott, J. J. (1994). Western hemisphere economic integration. Washington,
DC: Institute for International Economics.
Jessop, B. (2000) Reflexiones sobre la (i)lógica de la globalización, Revista Zona Abierta, (92/93):
95–125.
Joint Working Group EU-CAN. (2006). EU-CAN Joint Assessment on Regional Economic Inte-
gration. Final Report of the Working Group, (12/07).
Lenz, T. (2009). From geopolitics to geoeconomics. The European Union’s promotion of regional
integration in Latin America, 1980s–today. Saarbrücken: VDM Verlag.
Lenz, T. (2012). Spurred emulation: The EU and regional integration in Mercosur and SADC.
West European Politics, 35(1), 155–173.
Libman A. & Vinokurov, E. (2010). The EDB System of Indicators of Eurasian Integration:
General Findings (UNU-CRIS Working Paper, (W-2010/6)).
Libman, A., & Vinokurov, E. (2012a). Regional integration and economic convergence in the
Post-Soviet space: Experience of a decade of growth. Journal of Common Market Studies, 50,
112–128.
Libman, A., & Vinokurov, E. (2012b). Holding-together regionalism: Twenty years of post-soviet
integration. Basingstoke-New York: Palgrave MacMillan.
Marks, G. (1997). Does the European Union Represent an n of 1?, ECSA Review, X(3): 2–3.
Moravcsik, A. (1997). Does the European Union Represent an n of 1?, ECSA Review, X(3): 3–4.
Pietrangeli, G. (2010). Supporting Regional Integration and Cooperation Worldwide: An Over-
view of the European Union Approach. In P. De Lombaerde & M. Schulz (Eds.), The EU and
world regionalism. The makability of regions in the 21st century (pp. 9–43). London: Ashgate.
Pollack. (1997). Does the European Union Represent an n of 1?. ECSA Review, X(3): 4–5.
xxxiv Introduction: Indicator-Based Monitoring of Regional Economic Integration
Rillo, A. D., & De la Cruz, V. (2016). Monitoring regional economic integration in Asia, ADBI
Policy Brief, No. 2016–1.
Ruı́z Estrada, M. (2004). Global Dimension of Regional Integration (GDRI) Model (FEA Working
Paper, (2004–7)).
Sbragia, A. M. (1992). Thinking about the European future: The uses of comparison. In A. M.
Sbragia (Ed.), Euro-Politics. Institutions and policy-making in the ‘New’ European Commu-
nity (pp. 257–291). Washington DC: The Brookings Institution.
Steger, M. B. (2009). Globalization. New York: Oxford University Press.
UNECA. (2001). Annual Report on Integration in Africa. Methodology for calculating indices of
economic integration effort in Africa. Addis Ababa: UN Economic Commission for Africa.
UNECA. (2002). Annual report on integration in Africa 2002. Addis Ababa: UN Economic
Commission for Africa.
UNECA. (2004). Assessing regional integration in Africa, ECA Policy Research Report, Addis
Ababa: UN Economic Commission for Africa.
UN-ESCWA. (2007). Annual review of developments in Globalization and regional integration in
the Arab Countries, 2006. New York: United Nations.
Vinokurov, E., et al. (2010). The system of indicators of eurasian integration. Almaty: EDB.
Warleigh-Lack, A., & Rosamond, B. (2010). Across the EU Studies-New regionalism frontier:
Invitation to a dialogue. Journal of Common Market Studies, 48(4), 993–1013.
World Bank. (2004). Monitoring and evaluation: Some tools: methods and approaches.
Washington, DC: The World Bank.
Zelenka, M. (2015). Monitoring regional economic integration in Practice. Bonn: GIZ.
Part I
Europe and Eurasia
Chapter 1
The European Commission Single Market
Scoreboard
1.1 Introduction
The Single Market of the European Union (EU), previously known as the Internal
Market, is a foundational backbone of the European integration project. It is
associated with the Single European Act, which was signed in 1986, and should
be understood as a process rather than a state. It consists of the formation of a
transnational market in which four freedoms are pursued: the free movement of
goods, services, capital and persons. The Single Market represents a core concept of
European integration as it enables European citizens to freely live, study, work and
carry out their businesses in the EU. Being a process, the EU continues to dismantle
remaining barriers between Member States and to harmonize legislation that may
hinder the performance of the integrated market and hamper the related benefits for
citizens and consumers. The Single Market, as defined in Articles 26 and
114 TFEU, consists of a large body of EU law that encompasses all measures
considered as having an effect on the market, whether it is about the four freedoms
or other supporting policies. The transposition of EU directives and regulations in
Member States’ national setting is therefore a crucial way to achieve a harmonized
set of rules for the Single Market. However, since directives ought to be transposed
V. Fernández Moriana
Toulouse Business School, Barcelona, Spain
A.M. Vida
American Chambre of Commerce (AmCham), Brussels, Belgium
Former intern, UNU-CRIS, Bruges, Belgium
P. De Lombaerde (*)
NEOMA Business School, Rouen, France
UNU-CRIS, Bruges, Belgium
e-mail: philippe.de-lombaerde@neoma-bs.fr
into national legislation by Member States following form and method consistent
with their respective constitutional settings, and taking into account national polit-
ical context and sensibilities, problems may arise with the transposition within the
established time limit. This situation may ultimately negatively affect the Single
Market’s efficiency.
As part of the general objective of enhancing the Single Market across the
Union, the European Commission has initiated the Internal Market Scoreboard
(IMS) in 1997. The Scoreboard is as such only one component of the complex
and sophisticated monitoring instrumentarium that has gradually been put in place
in the EU (Costea et al. 2008). It is essentially a monitoring record released twice
per year that synthetizes key information on the transposition of the Single Market
legislation, or lack thereof. The Scoreboard is an interesting record of data and an
instrument for public peer-pressuring and naming and shaming, sharing best prac-
tices, harmonization, and overall a system of soft power and monitoring. The initial
mandate of the Scoreboard was to depict the state of the Internal Market and to
gauge the degree to which Member States, the Council and the Commission meet
the Action Plan targets.1 Subsequently, the Scoreboard focused on tracking the
evolution of market integration by monitoring practices that (possibly) threaten its
functioning. Today, the IMS is an online platform on which the performance of
Member States regarding their timely transposition of recent directives, transposi-
tion backlog, average delays, and governance tools, are published.
This chapter offers an up-to-date description of the IMS and its evolution
throughout the years and assesses its relevance and impact on the implementation
of EU legislation. The chapter also gauges the importance of the Scoreboard as a
locus of production of public information and a basis for further analysis by
academics, governmental agencies, think tanks, etc.
The first Internal Market Scoreboard was published by the European Commission
in November 1997 and has since then been issued with a frequency of twice per
year. The Scoreboard is currently known under the name of ‘Single Market
Scoreboard’. In order to have a comprehensive view of the IMS and what it may
offer to practitioners and scholars, this first section traces the Scoreboard’s devel-
opments since its origin.
The official purpose of the IMS was to “first, offer a picture of the current state
of the Single Market and secondly to gauge the degree to which Member States, the
Council and the Commission are meeting the targets laid down in the Action
1
The Single Market Action Plan (SMAP) was endorsed by the Amsterdam European Council in
1997 which aimed at improving the Single Market by launching the IMS.
1 The European Commission Single Market Scoreboard 5
2
As stated in the first European Commission Internal Market Scoreboard, 1997, (1), Nov.
3
European Parliament, Transposition, implementation and enforcement of Consumer Law, 2009,
(IP/A/IMCO/NT/2009–02). http://www.europarl.europa.eu/RegData/etudes/etudes/join/2009/
416221/IPOL-IMCO_ET(2009)416221_EN.pdf
4
European Commission Single Market Scoreboard, 1997, (1), Nov. http://ec.europa.eu/internal_
market/score/docs/score01/score_en.pdf
5
European Commission Press Release, IP/98/441 (18/05/1998). http://europa.eu/rapid/press-
release_IP-98-441_en.htm?locale¼en
6 V. Fernández Moriana et al.
6
European Commission Single Market Scoreboard, 1999, (5), Nov. http://ec.europa.eu/internal_
market/score/docs/score05/score5_en.pdf
7
European Commission Internal Market Scoreboard, 2001, (8), May. http://ec.europa.eu/internal_
market/score/docs/score08/score8_en.pdf
8
European Commission Internal Market Scoreboard, 2001, (9), Nov. http://ec.europa.eu/internal_
market/score/docs/score09/score9_en.pdf
9
European Commission Internal Market Scoreboard, 2003, (12), May. http://ec.europa.eu/inter
nal_market/score/docs/score12/score12-text_en.pdf
10
European Commission Internal Market Scoreboard, 2008, (18), Dec. http://ec.europa.eu/inter
nal_market/score/docs/score18_en.pdf
11
European Commission Internal Market Scoreboard, 2009, (19), July. http://ec.europa.eu/inter
nal_market/score/docs/score19_en.pdf
1 The European Commission Single Market Scoreboard 7
in the 2010 Scoreboard for the first time. It was designed to provide an uncompli-
cated visualization of the overall performance of Member States regarding imple-
mentation and application of Internal Market rules by aggregating essential
indicators in multicolored tables. The IMET allows to graphically compare relative
performances of Member States and associating various transposition indicators.
The year 2010 was also the period in which the Scoreboard introduced a new
section baptized Member States’ Best Practices. It calls upon Member States to
share their methods of resolving domestic issues and therefore presents the coun-
tries with the best performance on the indicators for the given period.12 This new
section of the Scoreboard represents an additional peer-pressure tool aiming at
achieving better transposition and enforcement score.
In 2012, the Internal Market Scoreboard started to be included in the Annual
Governance Check-up report, in which progress of the integration of the Single
Market is assessed by adopting a cyclical approach.13 The Governance Cycle
consists thereby of the seven stages of EU law-making: adopt, transpose, inform,
enable, connect, solve, and evaluate. Therefore, this report goes beyond monitoring
the transposition delays. It analyzes the entire functioning of the Single Market
from the adoption of legislation to the general evaluation of the performance of
Member States in its application. Since 2013, the IMS is exclusively online and was
converted to a website platform incorporating the Governance Cycle approach,14
and a new Single Market Strategy was released in 2015 aiming, inter-alia, at adding
new indicators in the Scoreboard on public procurement efficiency.
Today, the Scoreboard gathers information from Member States and the EU, and
incorporates governance tools (such as SOLVIT, and others). It organizes the data
in four sections. The first section offers a precise overview of the performance of
each Member State, and the second monitors the functioning of various multilevel
governance instruments. The third section reviews the performance by policy area.
Moreover, the ‘traffic light chart’ offers a rapid and general overview of Member
States’ performance according to transposition, infringements, EU Pilot, IMI,
Eures, Your Europe, Solvit and Points of Single Contact. The fourth section
organizes information regarding each of the seven stages of the Governance
Cycle (Fig. 1.1). The seven stages are described below.
12
European Commission Internal Market Scoreboard, 2009, (20), Dec. http://ec.europa.eu/inter
nal_market/score/docs/score22en.pdf, and European Commission (May 2010), The Internal Mar-
ket Scoreboard No.21. and http://ec.europa.eu/internal_market/score/docs/score21_en.pdf
13
European Commission, Making the Single Market Deliver, 2011, (24). http://ec.europa.eu/
internal_market/score/docs/relateddocs/single_market_governance_report_2011_en.pdf
14
European Commission Press Release (04/07/2013), IP/13/651. http://europa.eu/rapid/press-
release_IP-13-651_en.htm
8 V. Fernández Moriana et al.
1.3.1 Adopt
The Single Market is an example of EU’s regulatory power. In this sense, enabling
the four freedoms and their outcomes relies on a vast body of EU legislation mainly
adopted through the ordinary legislative procedure. After collecting the conclusions
of policy evaluations, the Commission decides whether to delete rules, simplify
them or adopt new legislation. Information on the Commission’s planned and
prospective initiatives for future action (Commission Work Programme), a moni-
toring system of decision-making in the EU as well as legislation proposals
(Pre-Lex) and a Eur-Lex component are available under this category. Moreover,
Single Market initiatives (Single Market Act I and II) and recent governance
initiatives are also published, indicating a will to store relevant information
conjointly.15
1.3.2 Transpose
In this stage of the governance cycle, the Scoreboard offers statistics on transposi-
tion performances and number of infringement proceedings throughout Member
States. It is the section that is discussed most.
15
European Commission, Adoption, simplification or deletion of Single Market rules, 2015, Oct.
http://ec.europa.eu/internal_market/scoreboard/governance_cycle/adopt/index_en.htm. Accessed
14 Dec 2015.
1 The European Commission Single Market Scoreboard 9
The first step is to evaluate how well Member States incorporate EU directives
into national law. The main key points consist of the respect of the transposition
deadline and the correctness of transposition according to the meaning and objec-
tives of the EU directive. Therefore, the Commission calculates transposition
deficits by aggregating Member States’ notifications on their progress. Indeed,
Member States are expected to notify the Commission once the new directive has
been transposed into national law. Therefore, the basic rationale for the indicators
regarding transposition performance rests on respect of the deadline of the legisla-
tion. However, cases occur where the transposition has only been partially com-
municated or where the transposition is completed but not communicated to the
Commission.
Five indicators are used to assess transposition performance: transposition
deficit, progress over the last 6 months, long-term delay, total transposition delays,
and compliance deficit. The ‘traffic light chart’ in Fig. 1.2 shows results for the five
indicators in the 28 Member States.
The first indicator measures the transposition deficit, which is the average
percentage of Single Market legislation not yet transposed in national law. This
percentage is calculated by including transpositions that have not been communi-
cated to the Commission, directives that are considered only partially transposed,
and infringement procedures for non-notification. The objective was to maintain the
transposition deficit under 1%, as depicted in Fig. 1.3. More recently, the require-
ment has become stricter as the new goal is to reach 0.5% of delayed transpositions
(Pelkmans and Correia de Brito 2012). It is relevant to appreciate the Scoreboard’s
method of data collection for this indicator as it is exclusively based on Member
States’ capacity or will to notify the Commission of its transposition progress. It is
why it constitutes one of the main concerns among scholars when using official data
and indicators for their research in this field. This subject is further discussed in the
fourth section.
The second indicator measures the progress over the last 6 months. By measur-
ing the amount and decrease of backlog, this indicator assesses the reduction of
accumulated uncompleted transpositions and notifications over the last semester.
The goal of this indicator is to showcase Member States that are making consider-
able efforts to reduce their transposition deficit for delays surpassing 2 years
(“outstanding directives”). Figure 1.4 illustrates the existing backlog reduction
progress for the April 2015 edition. We can see that Austria has taken the lead in
curtailing the bulk of non-transposed directives while Italy had most accumulated
delay.
The third indicator gauges the directives that have been overdue for two or more
years. For example, for the October 2015 Scoreboard, this indicator of “old
directives” takes into account legislation whose transposition was due before May
2013. Because of poor records on this indicator, the Council set a “zero tolerance”
target in 2002 for these directives that have not been transposed in the long-term,
since 2002 (Pelkmans and Correia de Brito 2012). This indicator is closely related
10 V. Fernández Moriana et al.
Fig. 1.2 Indicators of transposition performance (Source: European Commission Online Score-
board. Period: 11/2014–05/2015, European Commission Single Market Scoreboard, Transposi-
tion, 2015, May. http://ec.europa.eu/internal_market/scoreboard/performance_by_governance_
tool/transposition/index_en.htm. Accessed 14 Dec 2015)
Fig. 1.3 Transposition deficits (Source: European Commission Online Scoreboard. Period
04/2014–05/2015, Ibidem)
Fig. 1.4 Transposition progress over the last 6 months (Source: European Commission Online
Scoreboard. Period 04/2014–05/2015, Ibidem)
1 The European Commission Single Market Scoreboard 11
16
European Commission Single Market Scoreboard, Transposition, 2015, May. http://ec.europa.
eu/internal_market/scoreboard/performance_by_governance_tool/transposition/index_en.htm.
Accessed 14 Dec 2014.
12 V. Fernández Moriana et al.
Fig. 1.5 Overall transposition performance (left) and Overall Infringement Performance (right)
(Source: European Commission Online Scoreboard. Period 04/2014–05/2015, European Commis-
sion Single Market Scoreboard, Transposition, 2015, May. http://ec.europa.eu/internal_market/
scoreboard/performance_by_governance_tool/transposition/index_en.htm. Accessed 14 Dec
2014, and European Commission Single Market Scoreboard, Infringements, 2015, May http://
ec.europa.eu/internal_market/scoreboard/performance_by_governance_tool/infringements/
index_en.htm. Accessed 14 Dec 2015)
The Overall Infringement Performance map takes into account three different
indicators: the number of pending infringement proceedings, the duration of
infringement proceedings (in months) and the duration since Court’s ruling
(in months). The number of pending infringement proceedings represents the
total of pending infringement proceedings per country and also compares the rate
to the last reporting period. For example, for the April 2015 Scoreboard there were
749 pending cases overall, which represents a reduction of 77 cases from November
2014 and has attained a record low. However, it could also signify that the
Commission is simply introducing less infringement proceedings. Indeed, the
data depends on whether the Commission has observed an irregularity and whether
it has decided to launch an action.
Furthermore, it is interesting to note that 95% of all infringement cases are
solved in the pre-judicial stage. Once the case on the Single Market enters the
official infringement phase, it is probable that it will remain there for a minimum
duration of 2 years (Koops 2011). It is why the duration of infringement pro-
ceedings indicator calculates the number of months that have passed since the
dispatch of a letter of formal notice; it does not include infringement proceedings
that have reached the Court of Justice. Finally, the last indicator depicts the
duration since Court’s ruling. It is the most far-reaching infringement procedure
stage. This indicator adds the cases made against a Member State and the number of
months it has taken for the Member State to comply with the Court’s judgement.
1 The European Commission Single Market Scoreboard 13
Fig. 1.6 Average duration of pending infringement cases by sector (Source: European Commis-
sion Online Scoreboard. Period 04/2014–05/2015, Ibidem)
These two stages of the governance cycle comprise a set of tools and assistance
services for citizens and businesses. The Scoreboard assesses yearly these tools that
are alternatives to infringement procedures. On the one hand, the stage relating to
inform operates websites and services such as Your Europe portal, EURES network
and European Consumer Center Network (ECC-Net). Their purpose is to inform
citizens and businesses about their rights in the Single Market, propose opportuni-
ties and give them specific advice. On the other hand, the Services Directive (2006/
123/EC) set up Points of Single Contact (PSCs) which constitute the enabling phase
of the governance cycle. PSCs embody e-governance technology aiming at
17
European Commission Single Market Scoreboard, Infringements, 2015, May. http://ec.europa.
eu/internal_market/scoreboard/performance_by_governance_tool/infringements/index_en.htm.
Accessed 14 Dec 2015.
14 V. Fernández Moriana et al.
1.3.4 Connect
The tools presented in this stage (the Internal Market Information System (IMI) and
Consumer Protection Cooperation network (CPC)) work to boost administrative
cooperation between national authorities. This initiative was put in place to offset
the vertical approach placing the EU above Member States. Competent bodies in a
Member State can conveniently contact their counterparts for cooperation and
assistance. To measure performance in this stage, the Scoreboard uses indicators
which essentially monitor the speed of communication between different adminis-
trations and the efforts made to improve the IMI system. For the CPC Network,
responsible for enforcing EU consumer protection laws, the indicator consists of the
number of CPC cases.19
1.3.5 Solve
The Single Market Scoreboard also monitors EU-Pilot and SOLVIT performance.
Indeed, although these two tools and the Scoreboard differ in their stated purposes,
the expected medium to long term objective is similar: to improve Member State
compliance with EU law and prevent costly infringement procedures (Koops 2011).
18
European Commission Single Market Scoreboard, Your Europe, 2015, May;. European Com-
mission Single Market Scoreboard, EURES, 2015, May; European Commission Single Market
Scoreboard, European Consumer Centre Network, 2015, May; European Commission Single
Market Scoreboard, Points of Single Contact, 2015, May. http://ec.europa.eu/internal_market/
scoreboard/governance_cycle/index_en.htm#gov_cycle
19
European Commission Single Market Scoreboard, Internal Market Information (IMI) system,
2015, May; European Commission Single Market Scoreboard, CPC Network, 2015, May. http://
ec.europa.eu/internal_market/scoreboard/governance_cycle/index_en.htm#gov_cycle
1 The European Commission Single Market Scoreboard 15
They are online tools used to solve Single Market issues between EU citizens and
businesses and a Member State (SOLVIT), or between the Commission and a
Member State if a European law has been inappropriately applied by the country
(EU-Pilot). The Scoreboard therefore assesses the performance of these practical
tools in order to ensure their relevance. The effectiveness of EU-Pilot is predom-
inantly based on the Member States’ average response time. SOLVIT is measured
by the efficiency of the office in terms of service and resolution of issues.20
ECC-Net and CPC Network are also included in this section.
1.3.6 Evaluate
20
European Commission Single Market Scoreboard, SOLVIT, 2015, May; European Commission
Single Market Scoreboard, EU PILOT, 2015, May. http://ec.europa.eu/internal_market/score
board/governance_cycle/index_en.htm#gov_cycle. Accessed 14 Dec 2015.
21
European Commission Single Market Scoreboard, Evaluation and Assessment of Feedback,
2015, May. http://ec.europa.eu/internal_market/scoreboard/governance_cycle/evaluate/index_en.
htm. Accessed 14 Dec 2015.
16 V. Fernández Moriana et al.
Fig. 1.7 Evolution of average transposition deficit (Source: European Commission Online Score-
board. Period 04/2014–05/2015, European Commission Single Market Scoreboard, Transposition,
2015, May. http://ec.europa.eu/internal_market/scoreboard/performance_by_governance_tool/
transposition/index_en.htm. Accessed 14 Dec 2015)
1.4.1 Achievements
Out of wariness of the relevance of soft power measures and data reliability, the
Scoreboard has been developing its methods of data collection and issue coverage
since its creation in 1997. One can observe that the rate of transposition deficit has
decreased in the past 16 years, as illustrated in Fig. 1.7, although a direct causality
can obviously not be asserted. There is an inherent difficulty in assessing the impact
of soft power policy, consisting of pressuring Member States or simply publishing
data for improved future policy decisions.
As depicted in Fig. 1.7, the EU average transposition deficit has been progres-
sively curbing to reach stability. Since November 2011, the percentages of directives
that have not been notified as transposed, or perceived as incorrectly transposed,
oscillate between 0.5% and 0.7%. The Single Market Act’s proposal (2011) for an
average of 0.5% is therefore reached as a whole and attained by half of the Member
States. The first challenge of the Single Market - keeping the transposition deficit
under 1% - has been attained by the total of the Member States for 3 years in a row
since 2012. Nevertheless, other related outcomes are contrasting. For example, the
rate of incomplete implementation of directives in countries remained at 5% in 2013.
Therefore, seventy-three directives do not produce their full effect in the Union and
the Single Market is limited in its opportunities. This situation mostly concerns three
sectors: financial services, environment, and transport.
However, results are more positive in what concerns the number of infringement
proceedings. As seen in Fig. 1.8, the number of infringement proceedings, whether
it concerns non-timely or incorrect transposition of a directive, is at a record low in
1 The European Commission Single Market Scoreboard 17
Fig. 1.8 Number of pending infringement cases (Source: European Commission Online Score-
board. Period 04/2014–05/2015, Ibidem)
2015 and concerns 749 cases. All Member States have reduced or maintained their
number of pending cases, except new Member State Croatia.22
It is interesting to note that Member States’ performance regarding infringement
varies greatly depending on the indicator employed (Number of pending infringe-
ment proceedings, Duration of infringement proceedings and Duration since court’s
ruling). For example, Malta is the best performer for the indicator relating to the
total number of pending infringement proceedings, yet it is last in regard to the
duration of the proceeding. Italy and Greece spend most time between the Court’s
ruling and achieving conformity to the Court and the directive itself. Yet, the
overall number of pending cases has been steadily decreasing since November
2008 (see Fig. 1.8). It is believed that the implementation of early problem-solving
systems, such as SOLVIT (2002), EU-Pilot (2008), and IMI have addressed the
issue reasonably effectively. Indeed, these compliance mechanisms aim at
curtailing the number of Court cases by encouraging the search for informal and
administrative solutions beforehand. They are means for Member States’ adminis-
trations to better understand the provisions to be implemented by dialoguing with
the Commission and by contacting fellow national, regional and local administra-
tions from other Member States. The use of IMI and EU-Pilot helped to alleviate
complications between Member States, such as health systems coordination,
licenses wrongly issued to sell in other EU markets, cross-country verification of
professional qualifications and business licenses or residence permits issues. In
addition, Member States may simply not be aware of a flawed situation; the
22
Data is accessible in: European Commission Single Market Scoreboard, Transposition, 2015,
May. http://ec.europa.eu/internal_market/scoreboard/performance_by_governance_tool/transposi
tion/index_en.htm. Accessed 14 Dec 2015.
18 V. Fernández Moriana et al.
Scoreboard and other mechanisms may help in annunciate the situation (Koops
2011).
Nonetheless, some sectors are more subject to struggle in implementation than
others, regardless of compliance mechanisms. Transport, environment, and taxa-
tion, and more precisely, air transport, atmospheric pollution and indirect taxation
constitute the majority of infringement proceedings and are most time-consuming
in solving. These three sub-sectors alone account for almost a quarter of all
infringement proceedings.23
When analyzing infringement procedures, it is key to distinguish between the
distinct cases leading to infringement (Fig. 1.9). Late transposition of directives
accounts for the largest part of the number of pending infringement cases. These
“non-communication cases”, borrowed from the transposition tool, indicate a lack
of notification on the Member States’ part of a transposed directive. Wrong
application of Treaty articles, regulations and decisions come second in this pie
chart, which suggests a poor implementation of primary law. Thirdly, bad applica-
tion of directives constitutes 22.2% of total infringement proceeding cases. They
indicate a deficient application and implementation of the transposed provisions of
a directive, whereas non-conformity of transposition refers to the incomplete legal
transposition of the directive itself.
Regarding the impact of the Scoreboard on infringement for non-notification
cases, the results are mitigated but can be explained. As expressed in Fig. 1.10, the
number of infringements for late transposition dwindled until 2002, and reached
two peaks: in 2004 and in 2007. This can be interpreted in various ways. The two
EU enlargements represent the first distortion: new Member States had a limited
time to absorb the acquis communautaire. Second, the deadlines of the directives
themselves have a role to play. Indeed, an unusual number of directives included a
transposition deadline near the end of the year 2002. This fact led to a bump in late
transpositions notifications in 2003 instead of 2002 (Koops 2011).
It is important to keep in mind that data can be influenced by a number of
reasons. For example, the number of new or revised Single Market legislation that
has been adopted can influence the information. Indeed, if lesser directives are
endorsed, there will be less pressure for transposition. This can be illustrated in light
of the data of the second indicator relating to outstanding directives. The rate of
directives that have not been transposed in several years is far from its 0% target.
The percentage of directives overdue for two or more years, whilst lesser than
before, wavers at 4%, which represents 46 directives that have not been transposed
on time. The EU aims at eliminating this legislative loophole that makes some
sectors of the Single Market intangible in daily life. Moreover, compliance moni-
toring is a newer task for the Scoreboard and it is unknown if the Commission has
sufficient resources to oversee Member States’ correct transposition performance. It
23
European Commission Single Market Scoreboard, Infringements, 2015, May. http://ec.europa.
eu/internal_market/scoreboard/performance_by_governance_tool/infringements/index_en.htm.
Accessed 14 Dec 2015.
1 The European Commission Single Market Scoreboard 19
Fig. 1.9 Number of pending infringement cases by type (Source: European Commission Online
Scoreboard – May 2015, Ibidem)
Fig. 1.10 EU Compliance 1995–2009: number of infringement procedures (Source: Koops 2011:
27)
is why the principal long-term indicator only concerns the timeliness of transposi-
tion. Nevertheless, it is believed that the possibility of new financial sanctions
launched by the Lisbon Treaty has played a role in this configuration. Since 2011,
financial sanctions could be imposed in the wake of the first referral to the Court by
reason of failing to notify a transposition.
The ratio of directives to regulations also proves the lesser relative importance of
directives. On the one hand, there have been more regulations in the last years, and
therefore fewer transpositions. On the other hand, the number of EU directives is
steady. “[. . .] [T]herefore, a given transposition deficit is relatively less problematic
in 2012 than it was 10 years ago” (Pelkmans and Correia de Brito 2012:49), since
the weight of directives is lesser in comparison to regulations. Indeed, directives
have been consolidated, reviewed, deleted; or transformed into regulations
(Pelkmans and Correia de Brito 2012).
20 V. Fernández Moriana et al.
In addition to all the indicators used to show Member States’ performance, the
Single Market Scoreboard may have also influenced its own results by sharing the
different practices and initiatives that countries have implemented to improve
transposition and reduce infringements. Sharing “success stories” was a strategy
first used by businesses. It has since been applied in international organizations and
particularly in the EU in order to “inspire and instruct Member States as to how one
can most effectively tackle a certain problem” (Koops 2011:13). By exchanging
these specific practices, all Member States have the possibility to learn from one
another and benefit from a set of available policy ideas. A reform of an Italian law to
help transposition of EU law into domestic legislation and the creation of an Inter-
departmental Committee in Ireland,24 a centralized database for directives in
Greece,25 a network of “Euro-Coordinators” throughout governments and contact
points in Belgium26 are just some examples of Best Practices being shared in the
Single Market Scoreboard.
24
European Commission Internal Market Scoreboard, 2013, Feb. http://ec.europa.eu/internal_
market/score/docs/score26_en.pdf
25
European Commission Internal Market Scoreboard, 2012, Sep. http://ec.europa.eu/internal_
market/score/docs/score25_en.pdf
26
European Commission (September 2011), Internal Market Scoreboard. http://ec.europa.eu/inter
nal_market/score/docs/score23_en.pdf
1 The European Commission Single Market Scoreboard 21
The Single Market Scoreboard has not only been, and is not only, a useful tool to
record Member States’ performance on transposing European law regarding the
Single Market; it has also developed into an instrument for scholars and researchers
interested in analyzing the Single Market and compliance from various angles. As
explained above, the Scoreboard has been gradually refined since 1997. Today, it
communicates indicators by sector, Member State and stage in the governance
cycle as well as “governance tools”. The indicators used in the Scoreboard offer a
descriptive picture of the current situation regarding transposition, infringement
and general Single Market street-level reality and provide scholars with a signifi-
cant amount of disaggregated information, packaged in relatively long time series.
These data may inspire new problematics, offer a basis for explanatory and analyt-
ical theorization, and enable researchers to go beyond the descriptive platform.
Many questions may arise from a simple observation of Scoreboard data. Why do
particular sectors have a tendency to show higher transposition deficits? Why do
specific countries have recurrent transposition delays for more than 2 years while
others comply within 6 months? What group of countries are faster on solving
infringements, and why? Searching for the underlying reasons for these questions is
crucial. After identifying and confirming explanatory hypotheses, the Commission
and Member States may find the correct measures to address these issues. The
Scoreboard thus represents an easy and practical way to visualize differences
between countries and policy areas, offering academics a clue for the direction of
the necessary research, as well as providing data to test hypotheses presented
elsewhere.
The scientific literature on Member States’ compliance and transposition per-
formances is very diverse and touches upon a variety of disciplines, methodologies,
theoretical premises, and conclusions. The principal common characteristic among
the array of scholars is their interest in having access to an “excellent laboratory to
study (non-)compliance across policy fields and countries” (Angelova et al. 2012:
1270). A review of the literature enables to perceive the diversity of proposed
research questions and explanatory hypotheses. For example, does a preference-
based approach, focused on the Member State’s political preferences (Versluis
2005; Thomson 2009) explain non-compliance better than national administrative
capacity? Moreover, several approaches such as the management approach
(Haverland et al. 2011), the perception of national inclusion in the EU decision-
making procedures (Mastenbroek 2005), and the focus on national decision-making
structures (K€onig and Luetgert 2009) might shed light on compliance issues and
provide insights for the EU to further develop its monitoring mechanisms and
infringement tools tailored to address non-compliance. In a more specific way,
other works have tested the legislative and institutional misfit hypothesis (Thomson
2009) and domestic veto-player theory (Tsebelis 2002) as well as the more recent
“worlds of compliance” model (Falkner et al. 2007a).
22 V. Fernández Moriana et al.
27
The Annual Report is similar to the IMS but it concerns all EU legislation.
1 The European Commission Single Market Scoreboard 23
transposition. In some cases, only the first few provisions of the directive are
transposed and the country notifies it as complete. In other cases, only an insignif-
icant fraction is left and the Commission launches an infringement procedure
(Falkner et al. 2007b). This is why there might be a great difference between
these quantitative results and independent qualitative research. According to
Falkner et al. (2007b) there would be an average difference of 28 months between
quantitative data and independent qualitative data. More recently, Zhelyazkova
(2013) has criticized the reliability of EU data regarding the number of infringe-
ment proceedings and, again, transposition rates. Indeed, the first indicator depends
on whether the Commission detects the non-compliant case and on whether it
moves to act upon it and launches an infringement procedure. Therefore, the two
main focal points of the Scoreboard, infringement and transposition rates, would
not be formal indicators of compliance. Overall, these authors highly favor
collecting and re-using independent academic data rather than relying on the EU
Scoreboard and other official points of information. Importantly, their subjects of
study revolve around implementation and enforcement of EU directives and not on
their delay of transposition.
Scholars using the data from the Scoreboard attempt to control them and use
them as a starting point for their research. The use of the Scoreboards received a
push by the work of Kaeding and others who studied the factors influencing delay in
transposition. Data from the Scoreboard are taken as a starting point for the analysis
and are then combined with data from Communitatis Europeae Lex (CELEX or
Eur-Lex today) for collecting additional information on infringements. Sprungk
(2013) also starts her analysis with the help of information retrieved from the
Scoreboard. Analyzing the role of national parliaments in the transposition process,
Sprungk is of the opinion that the EU data is more relevant for studying delays in
transposition rather than for studying their legislative correctness. Kaeding and
Voskamp (2010) also use Scoreboard data on delayed transpositions, and use
distinct sources when analyzing incorrect transposition. Thomas (2013) has studied
the impact of national elections on timely transposition of EU directives. She argues
that “timely transposition sends a signal of reliability, respect, and support for EU
initiatives. At a practical level, it increases efficiency and allows the optimal policy
goal to be realized”. K€onig and Luetgert (2009) examine transposition failure and
delay of directives by way of separating domestic and EU-related variables. They
use data from the Scoreboard and compare them with other sources, such as
CELEX. Finally, these two authors concede that their logic follows the Commis-
sion’s positivist and legalistic approach, that mainly focuses on what is reported and
therefore, on timeliness of transposition.
Several authors thus use official EU data (Scoreboard and CELEX/Eur Lex) to
analyze the reasons of transposition delays. Their focus is therefore not on the
correctness and completeness of transposition, nor on their substantial implemen-
tation and enforcement. The transposition of directives is taken as their dependent
variable. This might give a hint on the usefulness of the Scoreboard from particular
angles of analysis, as every source of information has its benefits and drawbacks.
Nonetheless, some studies have used the Scoreboard for data on late and incorrect
24 V. Fernández Moriana et al.
1.6 Conclusions
The Single Market Scoreboard is one of the monitoring instruments that the
European Commission has in its toolbox to assess national compliance with EU
Directives in the Single Market. The essential goal of the Scoreboard is to improve
the functioning of the transnational market by showcasing the progress made by
Member States in the transposition of EU Directives relating to the Single Market.
Since 1997, the content of the Scoreboard has evolved, and today it comprises a
variety of dimensions. These are, for example, the Governance Cycle, the traffic
light chart, indicators on supranational governance tools, and public procurement
performance indicators. In all of these features however, the main focus of the
Scoreboard has dwelled on indicators measuring the timely and correct transposi-
tion of directives as well as data on the status of infringement procedures. Through
trial and error, the Scoreboard has taken, forgotten and created new indicators and
sections for the past 18 years.
The impact of the Scoreboard on its two goals is difficult to measure because of
the importance of external factors. The first and official objective, which is show-
casing the performance of Member States’ transposition of EU directives, is mainly
based on notification from the Member States. Although assessing timely notifica-
tions is feasible for an institution as the Commission, assessing the correct trans-
position of a directive is arduous. Indeed, one must take into consideration the
different national legislative structures and content, all the provisions of the direc-
tive as well as its general meaning. Concerning the underlying goal of the Score-
board, which is to improve Member State compliance, results are varied.
Implementation deficit and infringement proceedings have altogether decreased
1 The European Commission Single Market Scoreboard 25
in the past years, although some countries still have more trouble than others. Yet,
lengthy infringement proceedings are still a problem for the implementation of
some directives. EU directives on air transport, atmospheric pollution and indirect
taxation do not legally exist in some countries. Moreover, positive results can be
skewed by the fewer number of adopted directives in the past 10 years and the
relative higher number of regulations. Nonetheless, inter-state peer-pressure, Lis-
bon Treaty dispositions, governance tools (such as SOLVIT), and fear of criticism
emanating from consumers and businesses, may all have had an impact in improv-
ing compliance, although it is inherently difficult to assess.
Scholars are divided regarding the academic use of the Scoreboard. Whereas
some (B€ orzel 2001, Mastenbroek 2005, Falkner et al. 2007b, Thomson 2009)
explicitly dismiss its relevance, others (Kaeding 2013, Kaeding and Voskamp
2010, Steunenberg 2010, Sprungk 2013, Howarth and Sadeh 2010) endorse its
information, and yet others do not make mention of the European monitoring tool.
The main criticism of the Scoreboard lies in its incomplete nature in terms of
Member State implementation of EU law - and therefore, its street-level visibility.
However, scholars that use the Scoreboard principally capture its data as a starting
point in their research problem and study the question of timeliness of transposition
of EU Directives into national legislative frameworks. It seems that for many, the
Scoreboard is indeed a useful tool and data source for exploratory research. The
second criticism is the question of Member States’ accuracy regarding their noti-
fications to the Commission; yet the availability of quantitative information on
Member States’ compliance as well as on the performance of governance tools may
be worth investigating further. Finally, the Scoreboard may be a promising subject
of study in what regards the efforts of the Commission to harmonize and integrate
the Single Market, as well as for studies on soft power and peer-pressure. It is a tool
that is in constant evolution within a legislative European framework that also
renovates itself; therefore it needs constant reevaluation of its performance and
relevance.
References
Angelova, M., Dannwolf, T., & K€ onig, T. (2012). How robust are compliance findings? A research
synthesis. Journal of European Public Policy, 19(8), 1269–1291.
B€
orzel, T. (2001). Non-compliance in the European Union: Pathology or statistical artefact?
Journal of European Public Policy, 8(5), 803–824.
Costea, A.-C., De Lombaerde, P., De Vriendt, W., & Fühne, B. (2008). Monitoring and (Good)
governance of the integration process in the European Union. In P. De Lombaerde,
A. Estevadeordal, & K. Suominen (Eds.), Governing regional integration for development.
Monitoring experiences, methods and prospects (pp. 211–245). Aldershot: Ashgate.
Falkner, G., & Hartlapp, M. (2008). Problems of operationalization and data in EU compliance
research (WZB Discussion Paper, SP I 2008–104).
Falkner, G., Hartlapp, M. & Treib, O. (2007a). Worlds of compliance: Why leading approaches to
European Union implementation are only ‘sometimes-true theories’. European Journal of
Political Research, 46(3), 395–416.
26 V. Fernández Moriana et al.
Falkner, G., Hartlapp, M., Leiber, S., & Treib, O. (2007b). In search of the worlds of compliance:
Promises and Pitfalls of quantitative testing. Wien: Institute for Advanced Studies, Vienna,
political science series.
Haverland, M., Steunenberg, B., & van Waarden, F. (2011). Sectors at different speeds. Analyzing
transposition deficits in the European Union. Journal of Common Market Studies, 49(2),
265–291.
Howarth, D., & Sadeh, T. (2010). The ever incomplete single market: Differentiation and the
evolving frontier of integration. Journal of European Public Policy, 17(7), 922–935.
Huggins, R. (2009). Regional Competitive Intelligence: Benchmarking and Policy-Making.
Regional Studies, 44(5), 639–658.
Kaeding, M. (2013). Towards an effective European single market: Implementing the various
forms of European Policy Instruments across Member States. Wiesbaden: Springer.
Kaeding, M., & Voskamp, F. (2010). Better implementation of EU Legislation is not just a
question of taking Member States to Court (EIPA Working Paper, 2011/W/01).
K€onig, T., & Luetgert, B. (2009). Troubles with transposition? Explaining trends in member-state
notification and the delayed transposition of EU Directives. British Journal of Political
Science, 39(1), 163–194.
Koops, C. (2011). EU compliance mechanisms. The interaction between the infringement pro-
cedures, IMS, SOLVIT and EU-Pilot (Amsterdam Centre for European Law and Governance
Working Paper Series, 2011–08, 1–38).
Mastenbroek, E. (2005). EU Compliance: Still a ‘Black Hole’? Journal of European Public Policy,
12(6), 1103–1120.
Pelkmans, J., & Correia de Brito, A. (2012). Enforcement in the EU single market. Brussels:
CEPS.
Sprungk, C. (2013). Legislative transposition of directives: Exploring the other role of National
Parliaments in the European Union. Journal of Common Market Studies, 51(2), 298–315.
Steunenberg, B. (2010). Is big brother watching? European Union Politics, 11(3), 359–380.
Tallberg, J. (2002). Paths to compliance: Enforcement, management, and the European Union.
International Organization, 56(3), 609–643.
Thomas, M. (2013). The relationship between national elections and the delay in transposition of
E.U. directives. Politics & Policy, 41(6), 911–946.
Thomson, R. (2009). Same effects in different worlds: The transposition in EU directives. Journal
of European Public Policy, 16(1), 1–18.
Toshkov, D. (2007). In Search of the Worlds of Compliance: Culture and Transposition Perfor-
mance in the European Union. Journal of European Public Policy, 14(6), 933–954.
Tsebelis, G. (2002). Veto players: How political institutions work. Princeton: Princeton University
Press.
Versluis, E. (2005). Compliance problems in the EU. What potential role for agencies in securing
compliance? Paper presented at the 3rd ECPR general conference, Budapest, 8–10 September.
Zhelyazkova, A. (2013). Complying with EU directives’ requirements: The link between EU
decision-making and the correct transposition of EU provisions. Journal of European Public
Policy, 20(5), 702–721.
Ziller, J. (2006) Making European policies work: Research into the implementation of EU Law in
the 21st century. EIPASCOPE, Spec., 25–30.
Chapter 2
Scoreboard for the Surveillance
of Macroeconomic Imbalances
in the European Union
2.1 Introduction
1
To this can be added the recent “two-pack” which aims to further strengthen surveillance
mechanisms for euro area Member States, including budgetary surveillance and stronger moni-
toring of countries with financial stability issues or countries requiring financial assistance.
C. Cuerpo (*)
Spanish Independent Authority for Fiscal Responsibility, Madrid, Spain
e-mail: carlos.cuerpo@airef.es
J. Fischer
European Commission, Directorate General for Economic and Financial Affairs, Brussels,
Belgium
The key aspect of the MIP presented in this chapter is the alert mechanism
scoreboard of indicators used to single out countries for which a more in-depth
study of risks is needed to identify whether imbalances exists or not. Section 2.2
describes first the rationale of the MIP, how it works, the role of the alert mecha-
nism, guiding principles for the selection of the indicators and also the overall
economic rationale in which they should be read. Section 2.3 then comments on the
outcome of the reading of the scoreboard in the first three Alert Mechanism
Reports. Section 2.4 concludes. Further details on the indicators in the alert
mechanism are provided in the Annexes.
This section first describes the role of the scoreboard in the overall MIP. Then the
chosen indicators are described ahead of a presentation of the guiding principles
used for their selection and a discussion of their basic economic rationale.
The overall design of the Macroeconomic Imbalance Procedure follows the implicit
logic of the Stability and Growth Pact, with a ‘preventive’ arm and a stronger
‘corrective’ arm for more serious cases. For euro-area countries, the corrective arm
is supplemented by an enforcement mechanism including the possibility of finan-
cial sanctions2. To detect macroeconomic imbalances, the procedure relies on a
two-step approach (Fig. 2.1). The first step consists of an alert mechanism aiming to
identify Member States where there are signs of a potentially building up boom-
bust cycle or in which the adjustment to busts requires more in-depth examination.
In the second step, for the identified Member States, the in-depth reviews (IDR)
assess whether there are imbalances and, if so, their nature and extent.
The aim of the alert mechanism is to identify countries and issues for which
more in-depth analysis is required. More specifically, the alert mechanism consists
of an indicator-based scoreboard complemented by an economic reading thereof
presented in an annual Alert Mechanism Report (AMR). The conclusions of the
AMR are discussed in the Council and the Eurogroup to enable the Commission to
obtain feedback from Member States. The Commission then decides on the final list
2
The Macroeconomic Imbalance Procedure rests on two pieces of legislation. The first Regulation
(EU 1176/2011) sets out the details of the new surveillance procedure and covers all the Member
States. The second Regulation (EU 1174/2011) establishes the enforcement mechanism, including
the potential use of sanctions, and only applies to the euro-area Member States.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 29
3
The initial proposal contained ten indicators and envisaged an additional indicator of the banking/
financial sector vulnerabilities, which was finally agreed upon in the spring of 2013.
4
Final consumption expenditure of households and non-profit institutions serving households
deflator.
5
An overview of precise formulas used in the computation of the transformations for each headline
indicator is presented in Annex A.2.2, together with their corresponding thresholds, the indicator
tables from the most recent AMR including the reading indicators. It can be noted that this year
eight additional indicators on social issues have been added to the list of auxiliary indicators used
for the economic reading. Annex A.2.1 offers additional information on every indicator: data
sources, indicative thresholds and additional indicators used for economic interpretation.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 31
internal imbalances, and encompasses variables where both the economic literature
and recent experiences suggest associations with economic crises.
Secondly, the scoreboard (indicators and thresholds) are chosen as to provide a
reliable signaling device for potentially harmful imbalances and competitiveness
losses at an early stage of their emergence. This has led to a combination of stock
and flow indicators which can capture both shorter-term rapid deteriorations as well
as the longer term gradual accumulation of imbalances. Moreover, it has led the
Commission to set indicative thresholds at prudent levels, which on the one hand
avoid excessive numbers of ‘false alarms’ but which on the other hand are not set so
stringently that they only identify problems once they are entrenched. To this end,
thresholds have generally been established via a statistical approach based on the
distributions of the indicators’ values, by identifying the thresholds as the lower
and/or upper quartiles of the distributions: such thresholds are generally consistent
with the values found in the empirical literature.
Thirdly, the scoreboard has an important communication role. For this purpose,
the scoreboard consists of a limited number of indicators. Moreover, the choice of
indicators and transformations is kept as simple and straightforward as possible.
Data transformations are transparent and tractable so that they can be replicated by
third parties. The choice of indicators complements indicators/targets used in other
EU surveillance exercises. For transparency reasons, the Commission will make the
scoreboard indicators publicly available on its website.6
The fourth principle requires indicators to be of high statistical quality in terms
of timeliness and comparability across countries. To this end, they are derived from
data compiled according to the principles of the European Statistics Code of
Practice of the European Statistical System (ESS). Where available, Eurostat
sources are used so that the data comparability and statistical quality can be
ensured. Otherwise, when Eurostat data are not available, the highest quality
alternative data sources are chosen (e.g. the ECB).
The criteria described above illustrate how the selection of indicators has taken into
account a number of practical concerns such as availability of data or simplicity for
better communication. However, the reliability of their economic signaling is the
fundamental guiding principle for macroeconomic surveillance. To this end it is
important to have a consistent analytical framework in mind. First the headline
indicators are conceptually divided into two main areas; internal and external
imbalances. While discussed in more detail in Box 2.1 below, internal imbalances
6
http://ec.europa.eu/economy_finance/economic_governance/macroeconomic_imbalance_proce
dure/mip_scoreboard/index_en.htm
32 C. Cuerpo and J. Fischer
are traditionally defined in the economic literature along the two dimensions
covered by the Phillips Curve, unemployment and inflation. An economy is con-
sidered domestically balanced or in a sustainable capacity utilization position
whenever its rate of unemployment does not translate into inflationary pressures;
i.e. it has reached its NAIRU or non-accelerating inflation rate of unemployment.
When extending the analysis to an open economy context, the external balance
must be factored in. In the long-run, current account positions must be compatible
with economic fundamentals as countries cannot maintain an under or overvalued
currency ad infinitum. External balance does not necessarily imply zero current
account levels, as deficits/surpluses may result as natural responses to changes in
underlying structural characteristics and the related adjustment in saving and
investment decisions of economic agents. For instance, countries in a catching-up
phase often run current account deficits as investing in productive activities
increases the prospects of future income. Borrowing from abroad allows them to
smooth the inter-temporal profile of consumption. Similarly, countries with ageing
population may find it opportune to save today, i.e. run current account surpluses, to
avoid a drop in consumption in the future. Against this background the rationale for
each of the indicators can be discussed within the two dimensions.
The current account balance (or the current external balance)7 is the major driver of
net lending/borrowing of the economy as a whole and thereby provides important
information about the economic relations of the country with the rest of the world.8
Moreover, current account imbalances often reflect other types of disequilibria,
e.g. excessive credit expansions in some countries led to rapid asset price increases
and fed back into large external imbalances. The current account balance is
therefore an important indicator which provides information about the potential
existence of macroeconomic imbalances.
Based on an extensive literature review of 83 papers, Frankel and Saravelos
(2010) point out that the current account balance is one of the most frequent
statistically significant indicators in explaining crisis incidence. Current account
deficits can be a sign of an imbalance, if, for instance, the volume of borrowing is
such that it leads to an unsustainable external debt position. In turn, a high current
account surplus may be considered worrisome when it reflects weaknesses in
domestic demand. Surveillance under the MIP covers both current account
7
These terms reflect the same economic concept but are usually associated with different data
sources for this indicator (current external balance refers to National Accounts while current
account balance refers to Balance of Payments data).
8
Net lending/borrowing versus the rest of the world comprises both the current and the capital
account (the latter recording mainly capital transfers, which in the case of EU Member States may
be relatively sizeable due to transfers under EU structural funds).
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 33
surpluses and deficits which, from an economic point of view, pose different types
of policy challenges. In particular, unlike current account deficits, large and
sustained current account surpluses do not raise the same concerns about the
sustainability of external debt and financing capacities, concerns that can affect
the smooth functioning of the euro area (which is a key criterion for triggering the
corrective arm of the MIP). This means that surveillance under the MIP encom-
passes all Member States, but that a greater degree of urgency is required in
countries with large current account deficits and competitiveness losses.
In order to allow for a stock-flow analysis of external positions, the MIP
scoreboard looks into the Net International Investment Position (NIIP) as the
stock counterpart to the current balance.9 Typically, highly negative NIIPs result
from persistently high current account deficits. In this respect, a number of the
conceptual issues discussed in the section on the current account balances apply to
NIIP as well. The net international investment position (NIIP) records the net
financial position (assets minus liabilities) of the domestic sectors of the economy
versus the rest of the world. It provides an aggregate view of the net external
position of a country and it is also frequently used in economic analysis and
research, focusing on external vulnerability of countries and the risk of crises (see
for example Frankel and Saravelos 2010; or Furceri et al. 2011a, b).
Persistent shifts in price and non-price competitiveness of each Member State
relative to its major trading partners are also captured in the scoreboard via the
inclusion of the real effective exchange rate based on consumer prices, export
market shares and nominal Unit Labor Costs (ULC). First, the real effective
exchange rate based on consumer prices casts a comprehensive picture of global
‘price’ pressure on domestic producers in a medium-term perspective.10 Since it is
closely related to the terms-of-trade concept, this indicator also exemplifies the
attractiveness of imports over domestic production11. In the economic literature, the
REER has often been found to be a statistically significant predictor of the inci-
dence of economic crises: it is thus frequently considered among early warning
indicators (Reinhart et al. 1998). In particular, Frankel and Saravelos (2010)
identify the REER as a very important leading indicator in 48 out of 83 studies
on crises occurring before 2008.
9
Plus the capital account balance. However the current account balance represents in most cases
the bulk of the net lending and borrowing position.
10
Given that this indicator is meant to monitor the global competitiveness of each member state, it
is very relevant not to exclude the influence played by the exchange rate developments so to assess
the relative price developments conditional on exchange rates. This indicator will not be used as a
trigger to discuss exchange rate policy that is outside the scope of the entire exercise.
11
Terms of trade are country-specific and defined as the ratio of export to import prices, which in
principle can be understood as a REER for a particular choice of deflators. In contrast to pure
external competitiveness indicators such as export market shares, the REER thus not only
embodies price features of exported goods and services to external markets, but also the attrac-
tiveness of imports versus domestically produced goods. As a two-sided indicator, it is therefore
frequently related to current account developments (cf. Salto and Turrini 2010, for an overview).
34 C. Cuerpo and J. Fischer
The financial crisis has highlighted the dire implications of excessively high debt
stocks and rapid credit expansion on financial stability and economic growth.
Moreover, there is a potentially important link between credit growth and external
imbalances. Stronger relative demand pressures in some Member States fuelled
import demand, triggered capital inflows and contributed to the widening of current
accounts deficits. Excessive credit dynamics matched these domestic demand
pressures, leading to the rise in household and corporate debt (European
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 35
12
Credit growth in the quoted literature refers to outstanding credit growth, i.e. at the growth in the
stock variable which represents the flow plus valuation effects.
36 C. Cuerpo and J. Fischer
13
For an extensive discussion on the channels through which banks affect the sovereign and vice
versa see BIS (2011).
14
Acknowledging that no indicator can capture all potential risks stemming from the financial
sector, like the vulnerabilities that are related to the size of the financial sector, its expansion, risks
of liquidity and risks that are related to the funding structure. The ESRB will look at the financial
system from the perspective of systemic risk. The scoreboard will however approach it from the
point of view of resource misallocation and macroeconomic imbalances at country level, which are
essentially sources of risk to sustainable economic growth but can, if left unchecked, evolve also
into sources of systemic risk.
15
With respect to risks stemming from cross-border exposures, they are difficult to grasp with
domestic-oriented indicators. For deeper analyses, a breakdown of cross-border exposures by
counterpart country and sector can be a useful tool in depicting concentrations of risks and over-
exposures as stated in Borio and Drehmann (2009).
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 37
creditworthiness16. There are also feedback effects from banks to the government
as banks are large creditors to sovereigns, making them vulnerable to fiscal woes.
Moreover, a high level of general government debt increases the vulnerability of a
Member State and weakens its room of manoeuvre to deal with crisis situations. An
indicator for general government debt is therefore included in the scoreboard not to
monitor risks of unsustainable public finances, which are covered by the Stability
and Growth Pact, but to be considered together with the indicator on private debt
and thereby to offer a broader picture of Member States’ indebtedness.
16
The most prominent example is Ireland where the banking support induced a sharp deterioration
in public finances with a fiscal deficit exceeding 30% of GDP in 2010 (nearly two thirds of it
related to banking support) and a public debt level rising from 25% in 2007 to close to 100%
in 2010.
38 C. Cuerpo and J. Fischer
Current Current
account surplus account deficit
θ* Domestic Domestic
underutilisation overheating
Y* Y
17
The four program countries (Greece, Portugal, Ireland and Romania) were not covered in the
assessment as they are already under an enhanced program-based surveillance regime.
18
It should however be noted that while the Commission identified excessive imbalances the
corrective arm of the MIP was not initiated (which is a choice at discretion and not automatic). The
Commission gave the benefit of the doubt given the ambitious National Reform Programs
presented in the context of the European semester.
40 C. Cuerpo and J. Fischer
found to have excessive imbalance together with Slovenia. Germany was found
to have an imbalance.
• The 4th AMR published in November 2014 and identified 16 countries for an
IDR, adding to the list of the 14 countries with an imbalance from the previous
cycle also Portugal and Romania (both leaving program status entering the
standard EU surveillance frameworks).
It can be noted that the number of countries selected for an IDR have not been
decreasing so far. The basic rationale for this is that since imbalances are identified
after the detailed analyses in the previous IDRs, the conclusion that an imbalance
has been overcome should also take place only after duly considering all relevant
factors in another in-depth review, which then could potentially lead to the closure
of the MIP for some Member States.
The issues for which countries have been selected for an IDR over the four
published AMRs mirrors the broad scope of the procedure as reflected in the
scoreboard:
• External sustainability. In some countries, the external position has been in
focus due to large negative net international investment positions (NIIP) despite
in some cases rapid adjustments in the current account in recent years
(eg. Hungary, Bulgaria, Spain). A key mitigating factor in this context is the
role of Foreign Direct Investment (FDI) in catching up economies.
• Export performance and competitiveness. For several other countries, focus is
more on the weak export performance (eg. Italy, Belgium, France, Finland)
visible both in large losses of global export market shares as well as a trend wise
deterioration in current account positions seen over a longer period (even if
external sustainability as such may not be a key concern). Indeed, several euro-
area countries lost export market shares well beyond what would be explained
by the rapidly increasing competition from emerging economies. The drivers
behind this are discussed in the IDRs and are in some cases the losses in cost
competitiveness, both due to high wage increases but even more low productiv-
ity growth, while in other cases issues linked to non-cost competitiveness issues
dominates.
• Private sector indebtedness. Potential risk from high levels of private sector
indebtedness has been a key factor for a number of countries (for example
Denmark, Spain, the Netherlands, Sweden, UK). While credit flows have been
rapidly adjusting in recent years, many Member States are left with high levels
of private sector indebtedness and are in a likely prolonged process of
deleveraging and adjustment of sectorial balance sheets. In a number of cases,
the deleveraging challenge for households and/or businesses is compounded by
the high levels of public debt. It was highlighted that the impact of deleveraging
in the private sector could be magnified by the fiscal pressures.
• Housing and asset markets. Linked to the continuous build-up of indebtedness
in the private sector, several countries have displayed developments in asset
markets, in particular housing, that warrants further analysis (for example Spain,
Netherlands, Denmark, Sweden, UK) where very high increases in house prices
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 41
together with a rapid build-up of household debt has been a concern, even if in
recent years house prices have been adjusting.
In other words, the reading of the scoreboard and the identification of imbalances
is not an automatic exercise and there are large differences across countries, both in
terms of scope of the problems and the severity of the risks involved. While some
countries need correcting internal and external imbalances accumulated in the past
others with an overall better performance still have risks in specific parts of the
economy with a macroeconomic dimension.
The discussion above already points to changes in the economic conditions over the
last 4 years and how these are visible in the scoreboard. Indeed, the four consecutive
readings of the scoreboard allow looking at how the different indicators have
performed in terms of “flashes” (that is where the indicator value exceeds the
scoreboard value). It should be noted that in the 2014 AMR indicators were
reported on the basis of the new statistical standards (BPM6 and ESA2010). This
of course had an impact on individual values, in some cases significant, but did
overall only have a very marginal impact in terms of flashes and did not have any
impact of the overall reading of the scoreboard.19
Some of the trends are noteworthy and reflect economic developments over the
last few years and the characteristics of the progress with euro area rebalancing and
recovery process from the crises well (Table 2.1). For example:
• There has been a remarkable improvement in the current account positions of the
Member States which used to have large deficits which shows in the number and
composition of flashes. Many Member States that, until a few years ago,
registered the largest current account deficits and were experiencing
unsustainable developments. However, the external sustainability of the most
vulnerable economies has not yet been firmly re-established and most Member
States register large negative NIIP. At the same time current account surpluses
have increased in several cases and are now very high, and well above the
indicative threshold.
• There have been gains in price and non-price competitiveness in several coun-
tries, in particular the most vulnerable. In the latest updates of the MIP score-
board, there has been quite some volatility with strong depreciation recorded in
the AMR-2014 then being neutralized in the AMR-2015.
• Export performance has improved for several countries, but most Member States
keep losing market shares globally. However, over the latest years, the export
19
For details on the impact and implications of this changeover of statistical standards see the
statistical annex of the AMR-2015.
42 C. Cuerpo and J. Fischer
Table 2.1 Number of “flashes” in the scoreboard in AMR 2012, 2013, 2014, 2015.
AMR 2012 2013 2014 2015 Comment
CA 11 10 9 5 Declining, now surplus flashes dominates
NIIP 15 15 16 16 Steady (stocks take time)
REER 4 1 9 0 Volatile (appreciation early on vs. depreciation later)
EMS 15 17 19 17 Increasing (and less differentiated across)
NULC 8 4 1 5 Adjustment, cyclical
HP 2 0 1 2 Correction cycle
CF 1 1 0 1 Credit in correction cycle
PSD 15 15 14 15 Steady (stocks take time)
GGD 14 14 14 16 Steady (stocks take time)
U 7 9 11* 14 Increasing reflecting crises
TFSL – 1 0 0 Correction cycle
performance has become less heterogeneous among the EU countries, and there
has been an improvement in the export performance in some countries. The
assessment of developments in export market share should take into account that
the relative losses are related to the expansion of big emerging economies, like
China, Brazil, Russia, India, among other.
• Balance sheet adjustments continue in many Member States but the private debt
stocks remain high. Despite ongoing deleveraging efforts, the private sector debt
still exceeds the indicative threshold in most Member States.20 The pace and
extent of the ongoing adjustment varies, however, across countries. The
deleveraging dynamics is influenced by the underlying credit market conditions
and access to finance. Credit growth is now generally below the threshold.
Reduction in house prices became widespread and gained pace. Real house
prices further adjusted in recent years and has now bottomed out.
• The employment and social situation deteriorated in a number of countries
during the rebalancing process. In particular, unemployment has grown very
substantially in several Member States. Very few Member States have recorded
contained increases in joblessness in the latest number of years.
In the 2014 European semester the MIP has been applied for three full cycles. While
the procedure is still relatively new and continues to be shaped in a context of
“learning by doing”, valid experience has been gained. As regards the scoreboard
and its role in the alert mechanism stage of the procedure, it has in many aspects
performed its role as planned.
20
Cf. ‘Refining the MIP Scoreboard – Technical Changes to the Scoreboard and Auxiliary
Indicators,’ op. cit., on changes in the definition of private sector debt in the scoreboard.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 43
It should be recalled that the scoreboard is not an instrument for policy conclu-
sions and advice, it is a tool to filter out and signal different key issues related to
macroeconomic imbalances. It must of course also be kept in mind that while the
scoreboard in some sense is designed as a preventive tool, aiming for the MIP to be
applied early on in a process of building up imbalances, it has effectively first been
applied in a phase after the bust where economies have been in an adjustment phase.
This explains why stock indicators currently show more warning signals than flow
indicators. Moreover, despite being limited in scope, the indicators in the score-
board have allowed to make an economic reading with qualified argumentation
looking at interlinkages between stocks and flows, internal imbalances and external
performance and competitiveness. Clearly, there is no automaticity in the reading of
the scoreboard, and the selection of countries for an IDR. That there is a role for
judgment is not only well recognized but also intended.
How the scoreboard will develop in the future is not set in stone. There is a
continuous work on development every year where advances in statistics and
availability of data are taken into account. Such ongoing work is discussed with
legislators (Council and Parliament) and communicated in the annual Alert Mech-
anism Reports. For example, recently, there has been an agreement to include a
number of social indicators as auxiliary indicators and the indications that the role
of social indicators may increase even further.
The legislation calls for a general review of the MIP legislation end of 2014 and
the discussion will develop over 2015 Looking even further out, how the MIP and
its scoreboard will develop also depends on the work to develop the general
surveillance framework and the deepening of EMU.
Annexes
This annex looks more in detail at the scoreboard indicators, their defnitions and
also reports some of the concerns at the time they were selected, including alter-
native indicators considered. In several instances, a number of indicators/trans-
formations were considered, each having particular strengths and weaknesses. After
careful consideration of the pros and cons of these alternative options, the most
appropriate indicator was chosen. Nevertheless, recognizing the critical importance
of taking due account of country-specific circumstances and institutions, the eco-
nomic reading of the scoreboard is complemented by additional information and
indicators. This inter alia includes the general macroeconomic situation, such as
growth and employment developments, nominal and real convergence inside and
outside the euro area and specificities of catching-up economies. Additional indi-
cators are considered that reflect the potential for the emergence of imbalances as
well as the adjustment capacity of an economy, including its potential to sustain
44 C. Cuerpo and J. Fischer
The scoreboard indicator is the 3-year backward moving average of the current
account balance expressed in percent of GDP, based on Eurostat data from Balance
of Payments statistics, with the indicative thresholds of +6% and 4%. The average
over 3 years is used so as to control for short-term fluctuations of the annual figures
and to provide indications of the persistence of a potential imbalance.
Data on the current account balance21 are derived from the Balance of Payments
(BoP) statistics reported by Eurostat. This source is widely used by other interna-
tional institutions as well as academics. BoP (and International Investment Posi-
tion) statistics are the statistical tools expressly built to monitor the relations of a
country with the rest of the world. An important advantage of this data source is also
its quick availability and high frequency. BoP data also allow decomposing external
imbalances by counterpart area, hence giving an idea of a possible spill-over of a
crisis from a given country to another.
An alternative data source on current transactions balances is the Rest-of-the-
World Accounts (RoW) in the National Accounts (NA). This data is consistent with
other indicators derived within the NA framework and also with the Commission
forecast for the current account balances. However, there are discrepancies between
the data derived from the NA and the BoP data. These differences occur despite the
fact that “current external balances” from the NA and “current account balances”
from the BoP describe the same economic concept. The issue has been closely
monitored by Eurostat together with ECB and national statistical institutes and it
appeared that the differences stem from compilation practices, methodological
reasons, different data vintages and revisions, errors and omissions22. BoP data
are compiled first, and subsequently incorporated in relevant external account
components of NAs. When compiling NAs, data related to the RoW sector have
to be reconciled with those related to the domestic economy (the focus for NAs).
21
The current account covers all transactions occurring between resident and non-resident entities,
and refers to international trade in goods and services, income and current transfers.
22
In 2011, the fourth survey on the discrepancies between the BoP/RoW data will be conducted.
The past surveys (2009) analysed in detail the reasons for existing discrepancies and formulated
recommendations. Some Member States already implemented some of Eurostat’s recommenda-
tions. The methodological differences will hopefully disappear after 2014, but some discrepancies,
due to the different compilation practices, will remain.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 45
Indicative Threshold
23
It should nevertheless be noted that attempts to identify thresholds beyond which current account
imbalances pose a problem are mired with conceptual and methodological difficulties.
24
On the basis of the AMECO data, the average current account deficit at the onset of a reversal
(as defined by the IMF) would be 3.2% for the EU countries.
46 C. Cuerpo and J. Fischer
(median 3.4% of GDP) and the average “justified” surplus around 3.7% of GDP
(median 3.1% of GDP).
Finally, much research has focused on assessing the sustainability of current
account imbalances. This strand of literature typically attempts to estimate values
of current accounts which would stabilize the external position of a country at the
current or a predetermined level (e.g. Milesi-Ferretti and Razin 1996; Edwards
2001). These results are typically country-specific and do not deliver a general
benchmark.
The upper value of the threshold is set at +6%. The upper quartile of the
distribution of the 3-year backward average of current account balances corre-
sponds to +2%. To this an additional 4% margin has been added in line with the
“intelligent symmetry” approach to current account balances. This allows tackling
both current account surpluses and deficits but recognizes that the urgency for
policy intervention is clearly greater in the case of current account deficits. It also
reflects the fact that the risk of negative spillover effects of current account deficits
is more prevalent than for current account surpluses due to sustainability
considerations.
Complementary Indicators
In the discussions it was also agreed that it is important that the economic inter-
pretation will take due account of additional relevant information, in particular the
specificities of catching up economies. The potential risks from external deficits
need to be qualified by taking into account capital transfers in the form of EU
structural funds, as they can finance in part current account deficits. Similarly, the
destination of the capital flows is relevant as strong FDI inflows help to provide a
relatively safe financing of current account deficits in many of these Member States.
To account for the inflows of EU structural funds, the sum of current account and
capital account will be considered for Member States for which this information is
relevant. Conceptually, the sum of current account and capital account determines
the net lending/borrowing of a country and is thus the flow counterpart of the net
foreign financial asset position/net international investment position. The capital
account comprises (a) capital transfers receivable and payable between residents
and non-residents (e.g. debt forgiveness), and (b) the acquisition and disposal of
non-produced, nonfinancial assets between residents and non-residents (e.g. natural
resources, licenses, contracts, leases or marketing assets). The net size of the capital
account is typically rather small. However, in a number of catching up Member
States, capital account can be non-negligible as a part of structural/cohesion funds
is recorded here.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 47
Indicative Threshold
The statistical analysis of the NIIP distribution yields 35% of GDP as an indic-
ative threshold. It is difficult to establish a level of net external assets which can be
considered as risky and economic literature attempting to do this is rather scarce.
This is due to the fact that next to the absolute level of net foreign liabilities, it is in
particular the composition of both gross assets and liabilities in terms of types or
maturities, which determine the overall vulnerability of the external position of a
country.
Unlike large negative NIIP positions, large positive external asset positions are
not a priori considered to be problematic for a Member State or the functioning of
EMU. Therefore, the scoreboard contains an indicative threshold for negative
NIIP only.
Complementary Indicators
25
Data on the NIIP cover stocks of direct and portfolio investments, financial derivatives and other
investment and reserve assets.
48 C. Cuerpo and J. Fischer
26
FDI is indeed a less risky source of external financing, although it can be argued that high inflows
of FDI increase the vulnerability of an economy as FDI can flow out of the country too. This is
particularly the case of undistributed profits which are considered as FDI inflows. FDI also
generates dividend flows which are reflected in the external position of a country.
27
It should also be noted that NED only excludes the equity part of FDI but still includes “other
capital” FDI which covers borrowing and lending of funds (loans, debt securities) between the
direct investor and its subsidiaries abroad.
28
Nevertheless, the components of NIIP that are not considered in NED also carry potential risks.
The non-debt components of NIIP excluded from NED essentially consist of equity and financial
derivatives. While the investments underlying these flows do not generally need to be repaid at a
certain point in time, such investments can be rather volatile and generate sudden capital outflows
which can complicate macroeconomic management. Furthermore, some of these components can
also partially reflect the existing external as well as internal imbalances and ignoring them would
mean missing part of the overall picture.
29
REER are based on the harmonized index of consumer prices (HICP) where available. For
(non-EU) trade partners without HICP methodology, the respective headline Consumer Price
Index (CPI) is used.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 49
In order to derive the REER from the NEER, several options were discussed
during the design of the scoreboard. The competitiveness of each supplier relative
to its trading partners can be measured by the REER expressed either in terms of
production costs (ULC), export prices or economy-wide prices (HICP or GDP
deflators).
First, the REER based on broad measures of prices or costs, such as HICP or
GDP deflators, provides the most comprehensive picture of price competitiveness
of domestic producers in a medium-term perspective. The basket of goods on which
these price indexes are calculated includes both tradable and non-tradable goods
(excluding capital goods). Additionally, given that price indexes also include the
price of imported goods, countries with different import-dependency will have
different relative price effects of nominal exchange rates changes. Such effects
need to be accounted for when interpreting the REERs. Second, the ULC-based
REER shifts the focus of the assessment of relative competitiveness in terms of
consumer prices to relative production costs. This important notion is also picked
up by the ULC scoreboard indicator (see below). For tightly integrated economies
in a monetary union, a ULC-based REER would capture a similar notion as the
headline ULC indicator. Third, the REER based on relative export prices, while
being a rather intuitive measure of market competitiveness, suffers from a number
of weaknesses; (i) the calculation of export prices is strongly influenced by the
composition of exports and by the price dynamics of exported goods; (ii) REERs
based on export prices convey information on how producers set prices in order to
maintain market shares in case of nominal exchange rate variation (pricing to
market) even at the expense of profits, providing a short-term picture that might
be out of line with the dynamics of REERs calculated with different deflators.31
Indicative Threshold
Concerning the indicative thresholds, symmetric thresholds are considered for the
REER indicator. The focus is put on detecting harmful imbalances, which may be
captured by an unsustainable appreciation meaning a loss of competitiveness, or
depreciation signaling potential problems related to domestic demand or the poten-
tial of harmful future price convergence. Furthermore, a differentiation of thresh-
olds between euro-area and non-euro-area countries is adopted in line with the
30
In the case of the scoreboard, the NEER is obtained from a weighted average (by double export
weights) of the exchange rate versus a panel of the most important trading partners of the euro-area
(EU-28 plus Australia, Canada, United States, Japan, Norway, New Zealand, Mexico, Switzer-
land, Turkey, China, Brazil, Russia, South Korea and Hong-Kong). The indicator takes into
account about 76% of the world exports instead of only 58% with the current panel.
31
High productivity in ICT for example has been reflected in falling prices of ICT goods relative to
others. For countries heavily specialized in those goods (see Japan) this kind of price dynamics
will tend to limit the increase of REER based on export deflators with respect to the REER based
on other deflators.
50 C. Cuerpo and J. Fischer
Herman Van Rompuy Task Force (Task Force 2010). Differentiated thresholds
reflect nominal exchange rate variability, catering thus for that countries with
flexible exchange rates may be subject to non-persistent swings in the REER due
to nominal exchange rate fluctuations with their most important trading partners.
The differentiation between euro-area and non-euro-area Member States also
reflects the trend real appreciation in catching-up countries. This can be explained
by increases in wages in the tradable sector due to productivity growth that are
transferred to the wages and prices of the non-tradable sector (Balassa-Samuelson
effect) where productivity does not increase commensurately. Countries that have
undergone economic transitions (e.g. liberalized trade and capital flows), and have
been catching-up to the levels of development of the EU-15 countries, typically
have experienced a trend appreciation in terms of the REER indicator. If REER
appreciation is due to the Balassa-Samuelson effect, with productivity improve-
ments in tradable goods, this should not threaten international competitiveness. The
most recent empirical studies find a Balassa-Samuelson effect for new Member
States of only 1% per year, on average (Égert et al. 2005). This is a rather modest
contribution that is not sufficient to explain the observed REER appreciations in
catching-up countries.
Overall, with a REER indicator calculated as a 3-year percentage change, the
transformation looks at medium-term developments in relative prices. To also cater
for exchange rate flexibility, one standard deviation is added to the value of the
thresholds derived from the distribution in the sample of euro-area countries. The
standard deviation is larger than the value on the Balassa-Samuelson effect esti-
mated in the literature, i.e. 1% change per year as signaled above. The thresholds
corresponding to the lower and upper quartiles of the distribution are /+5% for the
3-year percentage change. These thresholds would apply to euro-area countries32.
For the non-euro area countries, the standard deviation of the distribution is
subtracted from the lower quartile and added to the upper quartile. The resulting
thresholds for non-euro-area countries are therefore /+11%.
Complementary Indicators
The REER indicator captures persistent price changes in a common reference unit
(HICP/CPI) relative to major trading partners and thus illustrates the magnitude of
developments in price and cost competitiveness. Significant deviations of the
REER based on HICP/CPI from the benchmark indicate that prices have grown
out of line with productivity for some time without compensation via the nominal
32
The thresholds for non-euro area countries cannot be derived from the distributions of the
percentage deviations from the three-year percentage changes for non-EA member states because
these distributions are heavily influenced by the strong appreciations occurred in the past 15 years
in many transition economies.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 51
exchange rate, i.e. the country has lost or gained labor cost competitiveness with
respect to its trading partners.
In particular for euro-area Member States, persistent divergence in price and cost
competitiveness versus their EMU peers is a concern as this may hamper the
smooth functioning of the monetary union. In order to monitor such structural
losses or gains in competitiveness and trade, the additional indicators complement
the economic reading with a REER indicator that focuses on euro area trading
partners instead of the broader set of 36 countries in the headline REER indicator.
Moreover, REER developments are analyzed in conjunction with other scoreboard
indicators on competitiveness (in particular the development of ULC and export
market shares) to gain insight on the cost, price and non-price competitiveness
performance of Member States.
The scoreboard indicator is the percentage change of export market shares over
5 years, based on Balance of Payments Eurostat data, with a lower indicative
threshold of 6%. For each country, the export market shares are computed as
the share of the country’s export revenues in total world export revenues, in current
prices. The indicator thus adds many aspects of competitiveness to the scoreboard
that are not captured by price and cost competitiveness alone (that is monitored with
the real effective exchange rate based on HICP/CPI and the nominal ULC).
There are a number of options available as regards the definition of the indicator.
Firstly, one aspect to take into account is the time variation to apply: changes over
one, 3 or 5 years. Given the high volatility of year-on-year changes in view of
idiosyncratic trade shocks, this option was excluded in favor of a longer assessment
period which would better reflect structural losses/gains in export performance. The
percentage change over 5 years of the value of goods and services exports for each
country as share of the world exports of goods and services appears to be the most
opportune data transformation to measure long-term competitiveness development.
There is an important caveat, though: the short time series available permits to
calculate 5-year export market shares changes only from 1999 onwards.
Indicative Threshold
The indicative threshold of the export market share indicator has been obtained
from the lower quartile of the data series distribution. This threshold corresponds to
cumulative losses of 6% over a period of 5 years. For this indicator, no upper
threshold has been considered because in the context of the MIP, since the focus is
on the detection of the harmful imbalances that may jeopardize the healthy func-
tioning of the EMU. In that context, the key concern is the detection of Member
52 C. Cuerpo and J. Fischer
Complementary Indicators
33
The volume indicator has therefore been calculated by using for each country export of goods
volumes indexes derived from EUROSTAT and for the world export of goods volume indexes
derived from UN-Comtrade. Cf. UN, 2010 International Trade Statistics Yearbook, Volume II-
Trade by Commodity world trade tables covering trade values and indices up to the year 2010
(December 2011) and UN, 2009 International Trade Statistics Yearbook, Volume II- Trade by
Commodity world trade tables covering trade values and indices up to the year 2009 (December
2010). See, http://comtrade.un.org/pb/
34
With respect to non-price competitiveness, the quality differentiation and the characteristics of
exported products are often mentioned; however no aggregate and widely-used measure is
available to quantify the concept.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 53
The scoreboard indicator is the percentage change over 3 years of nominal unit
labor cost based on Eurostat data, with the indicative thresholds of +9% and +12%
for euro-area and non-euro-area countries, respectively.
The ULC index used in the scoreboard corresponds to the ratio of compensation
per employee to real GDP per person employed (labor productivity). The original
data on nominal compensation per employee, GDP and employment stem from
Eurostat and the index is calculated by DG ECFIN (AMECO database).36 In order
to capture the medium/long term developments of labor costs, the scoreboard
indicator for the ULC is calculated as the 3-year percentage change, as it dampens
35
This does not rule out cost competitiveness because the higher the productivity, the more output
will be produced for the same amount of inputs, which corresponds to lower marginal costs of
production.
36
The series used are: compensation of employees (total economy), employees (total economy),
gross domestic product at constant market prices, employment (total economy that also includes
self-employed). When available, full-time equivalents of employees and employment are used.
54 C. Cuerpo and J. Fischer
Indicative Threshold
The threshold corresponding to the upper quartile of the statistical distribution over
the sample of euro-area countries is 9%.37 For non-euro-area countries 3 percentage
points have been added to obtain a 12% threshold. This differentiation is not based
on the statistical distribution over the non-euro area sample, but was made since
historical data reflect the fact that the majority of non-euro area countries have
experienced a major trade liberalization in the period covered by our data (since
1995), which entails a natural process of factor price equalization towards the levels
of the trade partners. These strong adjustment processes due to trade liberalization
should however be considered to weaken over time and in the future. In that respect,
catching up transition economies are of particular concern as they can experience a
trend increase in ULC because the increases in wages in the tradable sector linked
to productivity growth are transferred to the wages and prices of the non-tradable
sector (Balassa-Samuelson effect), where productivity does not necessarily
increase. However, recent empirical studies gauge this effect to be limited (Égert
et al. 2005; European Commission 2008; and Peters 2010). No upper threshold has
been considered, because in the context of the MIP, the focus is on the detection of
the harmful imbalances that may jeopardize the smooth functioning of the EMU,
such as unsustainable increases in the cost of labor.
Complementary Indicators
37
Following suggestions by the ECB, thresholds were also calculated with a convergence approach
methodology (i.e. for each year the average of the three best performers plus a fixed percentage)
however such year-specific thresholds resulted to be very cyclical and heavily influenced by
outliers.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 55
sharp increases in ULC. Similarly, the developments in REER, which show price
and cost competitiveness relative to the main trading partners, point to increased
divergence.38 This may signal potential structural rigidities in product and labor
markets but partly reflects the catching-up process in several Member States. To
account for the longer-term losses in cost competitiveness, percent variations over
longer time periods (up to 10 years) are also considered in the economic reading of
the scoreboard.
Moreover, scoreboard indicators on competitiveness and trade are
complemented by a set of additional indicators. Persistent divergence in price and
cost competitiveness among euro-area countries is of particular concern, provided
that ensuing external debt problems may hamper the smooth functioning of EMU.
For this reason, an effective ULC deflator39 indicator versus the rest of the euro-
area countries is included among the additional indicators. The importance of
monitoring effective ULC deflators as an indicator of competitiveness develop-
ments was recognized by the Heads of State or Government of the euro area in their
Council Conclusions of 11 March 2011. As part of efforts to assess whether wages
are evolving in line with productivity developments, the Pact for the Euro Area
calls for ULCs to be monitored over a period of time, by comparing developments
in other euro-area countries and their main trading partners.
The scoreboard indicator is the year-on-year growth rate of the deflated house price
index (HPI)40, data source Eurostat, with an indicative threshold of 6%. The
consumption deflator is used to reflect the value of house prices relative to the
whole consumption basket. This way of computing real house prices is widely used
in the literature and by other international organizations (e.g. OECD).
Data on house price indices are provided by various institutions. The only
harmonized index is, however, the Eurostat experimental house price index
38
In a number of Member States with high external deficits, the increases in labor costs and REER
appreciations were concentrated, although not exclusively, in the non-tradable sectors. This, in
turn, induced a reallocation of resources towards these sectors, exerting further pressure on
external positions.
39
The effective ULC deflator relative to 35 trading partners is calculated by DG ECFIN. Reference
countries were selected on the basis of their importance for euro area exports. The effective ULC
Q Dj wi
deflator relative to partners i is computed as Di where wi are the trade weights (double export
i
weights, 1999¼100) and Dj, Di are deflators for home country j and partner country i. The effective
ULC deflator uses “double-export-weighting” The general idea of using the “double-export-
weighting” procedure is to reflect (i) competitors’ shares in export markets; and (ii) the relative
importance of a particular market for the country and industry under consideration.
40
Household and NPISH final consumption expenditure (P31_S14_S15).
56 C. Cuerpo and J. Fischer
Indicative Threshold
Given the scarcity of time series data, it is difficult to derive a threshold based on the
statistical distribution. Using the OECD dataset of 19 OECD countries on a long
series of historical data (1970–2007) gives a lower upper quartile of the distribution
of 6%. This compares with the threshold derived from the information provided by
the house price cycle. For instance, a recent study by Agnello and Schuknecht
(2009) looks into house price cycles and identifies phases of booms and busts in
18 industrialized countries. The 25 most severe booms are characterized by an
average expansion of real house prices of 40% over an average period of 7 years.
(The severity is judged based on an index which gives an equal weight to the
magnitude and the duration of the house price in the boom phase.) This translates
into an annual increase of close to 6%. Given that only the top 25 most severe
41
At the same time, Eurostat is also working on the Owner Occupied Housing (OOH) index.
Unlike the HPI, it measures the cost of owner occupiers in a HICP framework. For details on the
differences between the two, see Eurostat (2010a) and Eurostat (2010b).
42
30 January 2012.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 57
booms over the period 1970–2007 are selected amongst the total of 100 identified
booms, the associated 6% threshold could be seen to be at the high end.
Complementary Indicators
As part of the economic reading of this scoreboard indicator, real house price
growth over longer time periods will also be considered, as a complement to the
short-run indicator. To this end, 3-year average price growth rates are used as an
additional indicator. Moreover, the analysis of the house price cycle proves to be
very informative. The cumulated house price growth from the latest through to peak
and the average annual pace of growth can illustrate the scale of house price
developments. Coupled with information on house price determinants, such as
credit growth, cost of credit and demographic factors, these could provide indica-
tions of future house price developments.
During the process of designing the scoreboard, the nominal house price index
was also discussed; this indicator is likewise used in the economic reading. For
instance, if nominal house price inflation occurs at the time of final consumption
inflation, and thereby the real house price growth does not pick up the acceleration
in nominal house price inflation, potential risks of a house price bubble will be
grasped through economic judgment. In order to put house prices into perspective,
it is useful to assess them against households’ capacity to repay and alternative
options such as rental markets. In this vein, affordability (price-to-per capita
disposable income) and dividend (price-to-rent) ratios will also be assessed.
Although their findings have to be considered with caution due to their simplifying
assumptions and their crude approach, they provide a useful qualifier.
Volume indicators, in particular residential construction and value-added in
construction (as percent of GDP), are a useful complement to assess house prices.
The responsiveness of supply to changes in prices plays an important role in
shaping housing markets. A responsive housing supply reduces house price vola-
tility but at the potential cost of greater fluctuations in residential investment, with
the net impact on overall economic activity being unclear (Andrews et al. 2011).
Thus, it seems that during boom periods, inelastic housing supply reinforces house
price overvaluation while high supply elasticity coupled with expectations of future
housing price rises may lead to overshooting in construction activity.
The scoreboard indicator is the stock of private sector43 debt in% of GDP, defined
as the sum of loans and securities other than shares, consolidated. The threshold of
private sector debt is 133%.
58 C. Cuerpo and J. Fischer
Private sector debt is a stock variable defined as the sum of loans and securities
other than shares (excluding financial derivatives)44 and is expressed in percentage
of GDP. The data stem from the annual financial accounts and balance sheets
(AFA) collected by Eurostat and the quarterly financial accounts (QFA) collected
by the ECB.45
The envisaged indicator is currently based on consolidated data, i.e. excluding
intra-sector liabilities such as intra-enterprise loans. When the scoreboard was
initially designed (European Commission 2012), however, it was decided to use
non-consolidated data mainly because of lack of consolidated data for all Member
States. One drawback of non-consolidated data is that it is not known to which
extent intra-sector liabilities are dominated by intra-group transactions. If intra-
group loans form the bulk of intra-sector credit, non-consolidated data may be
biased due to national and multinational (non-financial) corporate accounting
practices. For example, in Member States where each unit/branch of an
enterprise-group reports on its credit/debt, the non-consolidated data would prob-
ably show higher figures than in Member States where the headquarter reports on
total group consolidated credit/debt. Thanks to the technical work by Eurostat and
the Member States’ statistical institutes, consolidated data are now available for all
Member States and used as sources for the headline scoreboard indicator.
Indicative Threshold
The threshold of private sector debt is 133% of GDP, as derived from the upper
quartile of the statistical distribution of the indicator. Annual data for the period
1995–2007 were used to establish the threshold.
Economic Interpretation
The selection process of the indicator dismisses the category “other accounts:
payable”. Although it is a non-negligible subcategory for several Member States,
it exhibits high volatility and may therefore introduce noise in the data that is
difficult to justify. The item reflects valuation effects as well as volume effects
43
Private sector is defined as non-financial corporations, households, and non-profit institutions
serving households. The non-financial corporations sector includes both private and public
corporations. Referring to the proposed indicator as private sector debt may, therefore, be partly
misleading as it also includes public non-financial corporations (which are market producers).
However, in the absence of a more refined indicator, the current definition will have to be used.
44
In order to get a clearer economic interpretation of the indicator, financial derivatives were
excluded from the definition as of 2013, after consultation with Member States and the European
Parliament.
45
Both data sets deliver fairly consistent data, as QFA is broadly the quarterly equivalent of the
AFA data series.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 59
The scoreboard indicator is private sector credit flows (transactions) expressed in%
of GDP, and it includes loans and securities other than shares (excluding financial
derivatives), consolidated data. It is the flow counterpart of private sector debt
(which is a stock indicator). The indicative threshold of private sector credit is 14%.
The sources of data are the annual financial accounts and balance sheets (AFA)
collected by Eurostat and the quarterly financial accounts (QFA) collected by the
ECB. The source data used for debt and credit flows is the same. Therefore, data,
methodological and technical issues pertaining to these two indicators largely
overlap.
46
Please note that the debt remains unconsolidated within the household sector, and between the
non-financial corporation sector and the household sectors.
60 C. Cuerpo and J. Fischer
Two other indicators were considered and discarded. Firstly, initial consider-
ations aimed at an indicator measuring the year-on-year percentage change in credit
flow. The rationale behind this choice of data transformation was that it can detect
rapid increases in credit flows that could be associated with credit bubbles, which in
turn may contribute to crisis situations. However, interpretation difficulties arise
since credit flows typically evolve in a cycle. This induces a risk that by using this
indicator the gradual build-up of a credit bubble is concealed when credit flows
remain high but steady (“high speed but no acceleration”) and thus its early-
warning properties are jeopardized. Secondly, the year-on-year change in private
sector debt as percent of GDP was considered, as it represents the most straight-
forward flow counterpart of the indicator on private sector debt. Notwithstanding its
consistency with the stock variable, this indicator is heavily influenced by Other
Economic Flows (OEF), which is a non-directly interpretable residual. OEF con-
sists of nominal holding gains and losses (changes in prices) and other changes in
volume (mainly reclassifications). However, distinguishing between changes in
prices and changes in volumes is difficult, and it seems that OEF is heavily
influenced by reclassifications.
Indicative Threshold
The indicative threshold of private sector credit is 15% of GDP, as derived from the
upper quartile of its historical distribution. Annual data for the period 1995–2007
are used to establish the value of the threshold.
Complementary Indicators
As in the case of private sector debt, the subcategory “other accounts: payable” is
not included, although this item is potentially interesting to be considered as an
additional indicator to qualify debt developments. Moreover, as discussed for the
private debt indicator, an important issue is the choice between consolidated or
non-consolidated data for the scoreboard indicator. In order to ensure consistency
with the stock counterpart of credit flows, the latter is also based on consolidated
data, i.e. excluding intra-sector liabilities. Non-consolidated data will be used as an
additional reading indicator.
value outstanding at the end of the year and consolidated between and within the
sectors of general government. The threshold is 60%.
The definition of general government consolidated gross debt is the one used for
the purpose of the Excessive Deficit Procedure (EDP) as well as for the Stability
and Growth and Stability Pact (SGP). The Maastricht Treaty, together with Council
Regulation (EC) No 3605/93 define the general government debt as the total gross
debt at nominal value outstanding at the end of the year and consolidated between
and within the sectors of general government. Other accounts payable and financial
derivatives are not included in the definition, mainly for measurement reasons.
Indicative Threshold
As regards the threshold for the general government’s indebtedness, the Treaty
reference value of 60% of GDP will be used (as a separate indicative threshold for
public debt under the MIP would be confusing).
Complementary Indicators
General government debt is assessed for its contribution to the general indebtedness
of a Member State, being thus looked at together with private sector debt. A high
level of general government debt is more worrying when it accompanies large
private sector debt. Nevertheless, high general government debt represents a vul-
nerability per se. A high level of government sector debt cannot in any way
compensate for a low level of the non-financial private sector debt (and vice versa).
Unemployment Rate
The scoreboard indicator is the 3-year backward moving average of the unemploy-
ment rate, based on Labor Force Survey from Eurostat47, with an indicative
threshold of 10%.
Given the focus on the adjustment capacity of the economy and the ability of
labor markets to reallocate labor resources, the average over the last 3 years is
preferred to yearly figures which are strongly influenced by short term volatility. In
this sense, the selected indicator can be seen as a proxy of the structural
47
The unemployment rate is expressed conforming to International Labor Office definitions: the
labor force is the total number of people employed and unemployed. Unemployed persons
comprise persons aged 15 to 74 who are without work during the reference week, are available
to start work within the next 2 weeks, and have been actively seeking work in the past 4 weeks or
had already found a job to start within the next 3 months.
62 C. Cuerpo and J. Fischer
unemployment rate, which is, however, an unobservable variable and the estimates
of which are subject to numerous caveats. Similarly, the indicator considers levels
of unemployment rather than changes, as increases/drops in unemployment tend to
be highly correlated with GDP growth.
Indicative Threshold
The statistical approach delivers an indicative upper threshold of 10% based on the
upper quartile of the historical distribution. Due to the focus on adjustment in labor
markets and not on cyclical fluctuations, only an upper threshold was considered in
the scoreboard.
Complementary Indicators
The scoreboard indicator is the growth rate of total financial liabilities of financial
corporations, non-consolidated data, with an indicative threshold of 16.5%.48
This indicator has the advantage of being the simplest to grasp and thus easy to
communicate. The indicator is not covered explicitly by existing financial regula-
tion and therefore there is no risk of overlap. Given that it does not require
instrument disaggregation which is more prone to disentangling difficulties and
reclassifications, the indicator provides also a fairly reliable basis for comparison.
Moreover, it does not discriminate against different funding specificities of Member
States. And last but not least, it is not specific to the business model of a specific
subsector. In principle, the size of the financial sector can be measured on the asset
or the liability side of financial sectors’ balance sheet. The correlation between
financial assets and liabilities’ growth rates is very high (above 0.9), and thus it does
not make a large difference whether financial assets or liabilities are considered.
48
Liabilities include: Currency and deposits, Securities other than shares, Loans, Shares and other
equity, Insurance technical reserves and other accounts payable. The coverage of institutional
sectors includes central banks and other institutions, MFIs, insurance companies and hedge funds.
2 Scoreboard for the Surveillance of Macroeconomic Imbalances in the. . . 63
Indicative Threshold
Complementary Indicators
Existent scoreboard indicators, such as credit transactions and housing price devel-
opments, already provide information on the financial sector’s efficiency of allo-
cating resources and potential imbalances. For example, strong credit growth
coupled with excessive increases of housing prices indicate a possible misallocation
of credit and the build-up of an asset bubble. However, indicators capturing the
change in size of financial sectors’ balance sheets, leverage indicators or soundness
indicators also provide meaningful information.
Signals given by the growth rate of financial liabilities would be interpreted in
the economic reading in line with developments in the real sector. For instance, if
the headline indicator credit flow in the non-financial private sector (non-financial
corporations and households) develops at a slower pace, one should look at what
else drives growth in financial sector’s liabilities. Is this growth backed by a
balanced development across liabilities’ categories, in other words, is equity grow-
ing at a similar pace with debt or is the maturity structure biased against short-term
liabilities?
The debt-to-equity ratio is therefore a natural complement to be included within
the list of auxiliary indicators, as the ratio indicating the relative proportion of
shareholders’ equity and debt used to finance assets. The ratio provides information
on the excessive leverage building up within the financial sector which can have an
amplifying impact on the economic cycle (Kollmann and Zeugner 2012). A highly
leveraged financial system may amplify unfavorable economic developments, like
a recession, or doubts on the solvency of the sovereign. Excessive leverage by
banks is widely believed to have contributed to the global financial crisis. While
leverage is key for growth, excessive leverage carries the threat of the amplification
effect of the volatility of returns: since the absolute increase in value of returns is
accelerated when leverage is employed, so are the losses. As long as the money is
optimally allocated in economic terms, high leverage levels do not necessarily
imply high risk. It is only when the credibility of the borrower or the return of the
underlying instrument is more uncertain that high leverage will increase the
intermediary’s risk profile (European Banking Federation 2010).
Annex A.2.2 Scoreboard indicators and their indicative thresholds
% y-o-y Gross Fixed Gross Domestic Net Lending / Net External Inward FDI Inward FDI Net Trade Balance % Change % Change (5 years)
Year Change in Capital Formation Expenditure on R&D Borrowing Debt as Flows as Stocks as of Energy Products (3 years) in Export Performance
2013 Real GDP as % of GDP as % of GDP as % of GDP % of GDP % of GDP % of GDP as % of GDP in REER vs. EA vs. Advanced Economies
BE 0.3 22.3 na 0.1 86.8 11.1 191.2 4.4 0.6 2.6
BG 1.1 21.3 na 3.7 25.5 3.3 95.9 6.4 0.4 13.2
CZ 0.7 24.9 na 0.1 3.8 3.9 77.9 4.9 2.4 1.1
DK 0.1 18.4 na 7.1 9.8 0.2 29.5 0.1 1.3 12.1
DE 0.1 19.7 na 6.9 12.2 1.4 40.2 3.5 0.7 4.3
EE 1.6 27.3 na 1.4 6.4 3.6 95.4 2.4 5.5 22.1
IE 0.2 15.2 na 4.4 425.3 16.0 166.9 3.3 3.1 1.9
EL 3.9p 11.2p na 2.3 130.9 1.2 11.1 3.4p 3.4 22.1
ES 1.2p 18.5p na 2.1 91.2 3.0 54.8 3.3p 0.4 0.4
FR 0.3 22.1 na 1.3 32.4 0.2 40.8 3.1 1.3 6.8
HR 0.9 19.3 na 0.9 60.3 1.7 55.1 5.0 2.9 15.3
IT 1.9 17.8 na 1.0 59.2 0.9 23.8 3.3 1.0 12.6
CY 5.4p 13.4p na 1.7 115.4 15.6 275.5 6.3p 1.0 22.0
LV 4.2 23.3 na 0.1 35.8 3.2 53.0 5.4 0.3 16.1
LT 3.3 18.2 na 4.6 28.4 1.5 37.7 7.3 1.3 30.8
LU 2.0 17.1 na 3.8 2072.9 724.1 5206.0 5.6 1.7 9.5
HU 1.5 19.9 na 7.8 58.6 3.2 215.9 6.4 3.2 13.4
MT 2.5 17.5 na 4.9 115.5 5.7 132.4 9.6 0.1 2.8
NL 0.7p 18.2p na 8.6 41.5 37.1 537.7 1.4p 1.4 2.7
AT 0.2 22.2 na 0.9 20.2 3.6 77.0 3.5 1.6 11.0
PL 1.7 18.8 na 1.0 35.7 0.0 49.0 2.7 3.3 6.7
PT 1.4e 15.1e na 2.3 102.3 3.9 69.2 3.6e 0.1 1.5
RO 3.5p 24.7p na 1.3 34.6 2.0 42.8 1.9p 0.7 24.8
SI 1.0 19.7 na 5.9 34.9 0.2 29.2 5.5 0.0 10.6
SK 1.4 20.4 na 3.6 23.1 0.6 59.1 5.9 2.5 4.8
(continued)
% y-o-y Gross Fixed Gross Domestic Net Lending / Net External Inward FDI Inward FDI Net Trade Balance % Change % Change (5 years)
Year Change in Capital Formation Expenditure on R&D Borrowing Debt as Flows as Stocks as of Energy Products (3 years) in Export Performance
2013 Real GDP as % of GDP as % of GDP as % of GDP % of GDP % of GDP % of GDP as % of GDP in REER vs. EA vs. Advanced Economies
FI 1.2 21.2 na 1.3 35.1 4.5 46.8 2.6 1.9 27.3
SE 1.5 22.1 na 6.4 63.3 1.2 86.9 1.5 6.0 8.9
UK 1.7 16.4 na 4.2 26.0 1.6 73.6 0.9 4.3 5.4
References
Adalid, R., & Detken, C. (2007). Liquidity shocks and asset price boom/bust cycles (ECB Working
Paper, 732).
Agnello, L., & Schuknecht, L. (2009). Booms and busts in housing markets: Determinants and
implications (ECB Working Paper, 1071).
Andrews, D., Caldera Sanchez A., & Johansson, A. (2011). Housing markets and structural
policies in OECD countries (OECD Economics Department Working papers, 836).
Berkmen, P., Gelos, G., Rennhack, R., & Walsh, J. P. (2009). The global financial crisis:
Explaining cross-country differences in the output impact (IMF Working Papers, 09,280).
BIS. (2011). The impact of sovereign credit risk on bank funding conditions (CGFS Papers 43),
Committee on the Global Financial System.
Borio, C. & Drehman, M. (2009, March). Assessing the risk of banking crises – Revisited. BIS
Quarterly Review.
Buiter, W. H. (2010). Housing wealth isn’t wealth. Economics: The Open-Access, Open-Assess-
ment E-Journal, 4, 1–29.
Carlin, W., Glyn, A., & van Reenen, J. (2001). Export performance of OECD countries: An
empirical examination of the role of cost competitiveness. The Economic Journal, 111,
128–162.
Case, K., Shiller, R. J., & Quigley, J. M. (2001). Comparing wealth effects: The stock market
versus the housing market (NBER Working Papers, 8606).
Chinn, M. D., & Prasad, E. S. (2003). Medium-term determinants of current accounts in industrial
and developing countries: An empirical exploration. Journal of International Economics, 59,
47–76.
Coricelli, F., Mucci, F., & Revoltella, D. (2006). Household credit in the new Europe: Lending
boom or sustainable growth? (CEPR Discussion Paper, 5520).
Duenwald, C., Gueorguiev, N., & Schaechter, A. (2005). Too much of a good thing? Credit booms
in transition economies: The cases of Bulgaria, Romania, and Ukraine (IMF Working Paper,
05, 128).
Edwards, S. (2001). Does the current account matter? (NBER Working Paper, 8275).
Égert, B., Halpern, L., & MacDonald, R. (2005). Equilibrium exchange rates in transition
economies: Taking stock of the issues (William Davidson Institute Working Paper, 793).
European Banking Federation. (2010, December). Bank leverage and its economic implications.
European Banking Federation report.
European Commission. (2008). EMU@10. http://ec.europa.eu/economy_finance/emu10/index_
en.htm
European Commission. (2009). Competitiveness developments in the euro area. In European
Commission, DG ECFIN, Quarterly Report on the Euro Area, 8(1). Brussels.
European Commission. (2010). Surveillance of Intra-Euro-Area competitiveness and imbalances.
European Economy, 1.
European Commission. (2012, February). Scoreboard for the surveillance of macroeconomic
imbalances (European Economy, Occasional Papers 92).
Eurostat. (2010a). Draft technical manual on owner-occupied housing for harmonised index of
consumer prices. Draft version 1.9.
Eurostat. (2010b). Experimental house price indices for the euro area and the European Union
(Eurostat Research Paper).
Eurostat. (2013). Handbook on residential property prices indices (Eurostat Methodologies and
Working Papers).
Frankel, J., & Saravelos, G. (2010). Can leading indicators assess country vulnerability? Evidence
from the 2008–09 global financial crisis. Harvard Kennedy School: mimeo.
Freund, C. (2005). Current account adjustment in industrial countries. Journal of International
Money and Finance, 24, 1278–1298.
72 C. Cuerpo and J. Fischer
Furceri, D., Guichard, S., & Rusticelli, E. (2011a). Medium-term determinants of international
investment positions: The role of structural policies, OECD Economics Department Working
Papers, 863. Paris: OECD Publishing.
Furceri, D., Guichard, S., & Rusticelli, E. (2011b). Episodes of large capital inflows and the
likelihood of banking and currency crises and sudden stops, OECD Economics Department
Working Paper, 865. Paris: OECD Publishing.
Gerdesmeier, D., Roffia, B., & Reimers, H. (2009). Asset price misalignments and the role of
money and credit (ECB Working Paper, 1068).
Ghosh, S.R., & Ghosh, A.R. (2003). Structural vulnerabilities and currency crises (IMF Staff
Papers, 50(3)).
Gourinchas, P., & Obstfeld, M. (2011, January). Stories of the twentieth century for the twenty-
first. American Economic Journal: Macroeconomics, American Economic Association, 4(1),
226–265.
IMF. (2007). Chapter 3: Exchange rates and the adjustment of external imbalances. In IMF (Ed.),
World economic outlook. Washington, DC: IMF.
Jordá, O., Schularick, M., & Taylor, A. (2011). Financial crises, credit booms and external
imbalances: 140 years of lessons. IMF Economic Review, 59, 340–378.
Kollmann R., & S. Zeugner. (2012, August). Leverage as a predictor for real activity and volatility.
Journal of Economic Dynamics & Control, 36(8), 1267–1283.
Melitz, M. (2003). The impact of trade on intra-industry reallocations and aggregate industry
productivity. Econometrica, 71, 1695–1725.
Milesi-Ferretti G. M., & Razin, A. (1996). Sustainability of persistent current account deficits
(NBER Working Paper, 5467).
Peters, D. (2010). Price competitiveness in Central and Eastern Europe: A case study for
transition economies. IMK Studies. Macroeconomic Policy Institute: IMK at the Hans
Boeckler Foundation.
Reinhart, C., Kaminsky G., & Lizondo, S. (1998). Leading indicators of currency crisis (IMF Staff
Papers, 45(1)).
Sachs, J., Tornell, A., & Velasco, A. (1996). Financial crises in emerging markets: The lessons
from 1995. Brookings Papers on Economic Activity, 27(1), 147–216.
Salto M., & Turrini, A. (2010). Comparing alternative methodologies for real exchange rate
assessment (European Economy – Economic Paper, 427).
Setzer, R., van den Noord P., & Wolff, G. (2010). Heterogeneity in money holdings across euro
area countries: The role of housing (European Economy – Economic Paper, 407).
Summers, L. (1996). Commentary. In R. Hausman & L. Rojas-Suarez (Eds.), Volatile capital
flows: Taming their impact on Latin America (pp. 53–57). Baltimore: Inter-American Devel-
opment Bank/Johns Hopkins University Press.
Task Force. (2010, October 21). Strengthening economic governance in the EU. Report of the Task
Force to the European Council.
Chapter 3
The EU Index of Integration Effort
J€org K€
onig
The European Union (EU) is a unique economic and political integration project.
Often referred to as a system sui generis, the EU is unique in both the scope and
depth of its integration efforts. What has begun as a peacekeeping endeavor among
six European countries struggling from the aftermath of World War II has evolved
into a complex network of meanwhile 28 member states and numerous European
institutions with supranational authority.
Hereby, European integration has passed through nearly all of the so-called
stages of economic integration. According to Balassa (1961), formal
economic integration takes place in several stages that envisage successive
market liberalization between the participating economies, accompanied by
the formation of common rules and institutions. Whereas the first stages are
concerned with lowering tariffs and non-tariff barriers to trade and factor
movements to finally create a common market, the later stages are engaged in
allocating necessary policy prerogatives to the supranational level, eventually
culminating in the creation of a single economic and political entity. Beginning
with the lowest stage, these are a preferential trade agreement (PTA), a free trade
area (FTA), a customs union (CU), a common market (CM), an economic and
fiscal union (EFU), an economic and monetary union (EMU) and a political
union (PU).
J. K€onig (*)
Market Economy Foundation (Stiftung Marktwirtschaft), Berlin, Germany
Georg-August-University of G€ottingen, G€
ottingen, Germany
e-mail: koenig@stiftung-marktwirtschaft.de; Joerg.Koenig@wiwi.uni-goettingen.de
Table 3.1 assigns the individual stages of economic integration to the respective
steps of European integration. Since the pooling of coal and steel production in
1951, the EU and its predecessors have almost gradually developed from a mere
PTA to EMU. This development is worth mentioning as only less than 5% of all the
FTAs that have been notified to the World Trade Organization (WTO) have
succeeded in further deepening their integration process.1 The recent
implementations of the European fiscal compact and the European banking
union (as well as Britain’s vote to leave the EU) further show that the institutional
design of the EU is still developing, leaving the potential completion of PU for the
future.
Tinbergen (1954) once defined that ‘integration may be said to be the creation of
the most desirable structure of international economy, removing artificial hin-
drances to the optimal operation and introducing deliberately all desirable elements
of coordination or unification’ (p. 95). Generally speaking, the past integration
efforts of the EU have laid down the floor towards achieving this desirable structure
1
According to the WTO website, http://rtais.wto.org/UI/PublicMaintainRTAHome.aspx, there are
currently 6 CUs (besides the EU) out of 211 FTAs notified under GATT Art. XXIV or GATS Art.
V: the Caribbean Community and Common Market (CARICOM), the Southern African Customs
Union (SACU), the Eurasian Economic Community (EAEC), the Central American Common
Market (CACM), the East African Community (EAC) and the Southern Common Market
(MERCOSUR).
3 The EU Index of Integration Effort 75
Italy
Greece
Poland
Germany
United Kingdom
Sweden
Romania
Czech Republic
Hungary
Slovakia
Denmark
Slovenia
Estonia
Lithuania
0 10 20 30 40 50 60 70
Fig. 3.1 Pending infringement cases (Source: European Commission (2013), p. 21)
2
See also Fernández Moriana, Vida and De Lombaerde (this volume) on the legal compliance of
the EU member states.
76 J. K€
onig
65%
55%
EU importance
45%
EU openness
35%
25%
1993 1996 1999 2002 2005 2008 2011
Fig. 3.2 Share of Intra-EU trade in goods (Source: Own calculations based on Eurostat data.
Notes: ‘EU importance’ refers to the sum of imports and exports of goods traded within the EU-15
as a share of total trade in goods. ‘EU openness’ refers to the sum of imports and exports of goods
traded within the EU-15 as a share of GDP)
is that due to expected higher marginal revenues, the free movement of capital and
labor allows for the optimal allocation of production factors, thus enhancing the
productive efficiency of the firms. The rise in product specialization through a
reduction in average costs (economies of scale), in combination with the elimina-
tion of tariffs and non-tariff barriers to trade, pave the way for larger trade flows
between the member states. In turn, increasing trade is expected to have significant
positive effects on the economic performance of the member states – such as greater
market efficiency and product innovation due to increased competition – finally
leading to a reduction in price levels, a rise in consumption and, hence, long-term
economic growth.3
However, not all the member states seem to be able to make use of such
improvement in market efficiency. Even 20 years after the launch of the EU’s
single market, immense heterogeneities between the members’ trade patterns exist.
For example, whereas both Greece and Portugal show intra-EU trade balance
deficits in relation to their gross domestic product (GDP) of 5% in 2012, Ireland
and the Netherlands have surpluses of 12 and 28%, respectively. Furthermore,
when examining only the export ratios, even larger disparities appear: Belgium
and the Netherlands possess internal export ratios of nearly 65% in 2012, while
Greece and the United Kingdom hold ratios of only 6 and 9%, respectively. As
presented in Fig. 3.2, the EU’s internal trade in goods as a percentage of GDP (‘EU
openness’) has increased by 10% since 1993. When considering the share of
European trade over the total world trade (‘EU importance’) in goods, though, the
3
See Cecchini et al. (1988) and Baldwin (1989) for rather optimistic ex ante analyses of the
potential single market effects, and Ilzkovitz et al. (2007), Boltho and Eichengreen (2008) and
Badinger and Breuss (2011) for ex post analyses of the European integration effects on trade and
growth.
3 The EU Index of Integration Effort 77
0.8
0.6
0.4
0.2
-0.2
Fig. 3.3 Average levels of business cycle symmetry (1999–2012) (Source: Own calculations
based on Eurostat data. Notes: The figure presents the average correlation coefficient (and the
respective standard deviation) over 5-year moving windows between the domestic real GDP
growth rate and the average growth rates of the remaining EMU-11 countries. The growth rates
refer to quarterly data, which have been adjusted to seasonal and trend effects using the Hodrick-
Prescott filter)
ratio has steadily declined over the same period by roughly 10%. Despite offering a
large CM, the EU seems to lose attraction in comparison to markets outside the
union.4 The loss of market efficiency in some member states does not only generate
large macroeconomic imbalances within the union but also challenges the external
economic competitiveness of the EU.5
The same is true for the specific case of EMU. The national loss of autonomous
monetary and exchange rate policy demands for a certain degree of similarity in the
development of important macroeconomic variables. If large macroeconomic
imbalances appear, the member states become more prone to asymmetric shocks
and the ‘one-size-fits-all’ monetary policy becomes less effective. Figure 3.3, which
measures the symmetry of real GDP growth rates across the EMU-11 countries,
reveals large heterogeneities over the period 1999–2012. The correlation coeffi-
cients are high in some countries but very low in others, such as Greece and
Portugal. The respective standard deviations are also very different and even
point to negative correlations in some years. Hence, a well-functioning EMU
seems to be less likely if the member states do not increase their integration efforts
in that regard.
Likewise, increasing economic heterogeneities between the member states may
also pose a serious threat to the EU’s primary aim of ‘creating an ever closer union’
4
Such as the emerging markets of BRICS (Brazil, Russia, India, China and South Africa) and
MIST (Mexico, Indonesia, South Korea and Turkey).
5
See also Cuerpo and Fischer (this volume) on the measurement of macroeconomic imbalances
within the EU.
78 J. K€
onig
(Preamble, TEU). The results of the 2014 elections to the European Parliament
give cause for particular concern. Populists and anti-European parties have been
gaining ground in most of the European countries and the average voter turnout
was again far below 50%; in some countries even below 20%. Whereas many of
the populists in Northern Europe took advantage of the people’s fear that the EU
might be on the verge of becoming a one-sided fiscal transfer union and/or a union
of mass migration, many parties of the Southern countries blamed the EU for
imposing tough reforms and austerity policies on their economies. As a conse-
quence, the EU now has to deal with parties such as the Front National (France),
UKIP (Britain), Freedom Party (Netherlands), Lega Nord, Five Star (both Italy),
People’s Party (Denmark), Finns (Finland), Jobbik (Hungary), Golden Dawn,
Syriza (both Greece), FPÖ (Austria) and AfD (Germany) – all pursuing disinte-
gration of the EU.
These and other potential effects of the members’ different integration efforts
and capabilities underline the importance of measuring European economic inte-
gration. A separate investigation of these effects, however, does not allow for
general statements on a country’s overall level of integration efforts. Hence, the
various effects of EU integration should be merged into one statistic. Such a
composite indicator combining the most relevant aspects of European economic
integration would be able to verify the degree of a country’s overall level of
integration and to highlight those dimensions that need further integration efforts.
The countries’ overall integration levels would become numerically tangible,
making European economic integration operational for further empirical research.
This also allows the identification of member states that tend to fall behind the
general speed of European economic integration and of others that determine and
accelerate the speed of integration as a ‘core group’.
Recently, K€ onig and Ohr (2013) have developed such a composite indicator –
the ‘EU Index’.6 Their index covers various relevant aspects of EU integration
that also go beyond the Internal Market Scoreboard and the Macroeconomic
Imbalance Procedure of the European Commission. Appropriate statistical tech-
niques combine the data to manageable indices which offer both general and very
specific insights on a country’s integration efforts. In the following sections, the
composition of the EU Index is briefly presented and the statistical methods of
aggregating its indicators are discussed. As the original index covers only the
EU-15 countries until the year 2010, an extended and updated version of the EU
Index is introduced and analyzed, capturing the EU-25 countries until 2012.
Some recommendations for future integration policies are also derived from the
results.
6
For more information on the EU Index see www.eu-index.org.
3 The EU Index of Integration Effort 79
7
Other criteria referring to the ex ante optimality of EMU include the flexibility of domestic prices
and wages, the mobility of capital and labor, and the responsiveness of fiscal transfers (Mundell
1961). For an evaluation of the EMU’s constitutional design see Ohr (2009) and De
Grauwe (2013).
80 J. K€
onig
V i, t
I i, t ¼ 100 ð3:1Þ
V maxðj;T Þ
V i, t
I i, t ¼ 100 ð3:2Þ
V world
i, t
8
See, for instance, Romer (1986), Lucas (1990) and Krugman (1991) on divergence effects and, on
the other hand, the ‘law of one price’, the Lerner-Samuelson theorem and the traditional neoclas-
sical growth theory by Solow (1956) and Swan (1956) to explain economic convergence.
3 The EU Index of Integration Effort 81
Table 3.2 Description and sources of indicators measuring a country’s European Integration
Level
Indicator Description Source
EU Single Market
EU openness
Trade in goods Intra-European imports and exports of goods Eurostat
as a percentage of GDP.
Trade in services Intra-European imports and exports of ser- Eurostat
vices as a percentage of GDP.
Capital movement Intra-European stocks (inward and outward) Eurostat,
of foreign direct investments as a percentage (UNCTAD)
of GDP.
Labor migration Amount of European employees (ILO defini- Eurostat
tion) as a percentage of the total number of
employees (foreign and national).
EU importance
Trade in goods Intra-European imports and exports of goods Eurostat
as a percentage of total trade in goods.
Trade in services Intra-European imports and exports of ser- Eurostat
vices as a percentage of total trade in services.
Capital movement Intra-European stocks of foreign direct Eurostat,
investments as a percentage of total FDI. (UNCTAD,
OECD)
Labor migration Amount of European employees (ILO defini- Eurostat
tion) as a percentage of the total number of
foreign employees.
EU Homogeneity
Per capita income Real GDP per capita at constant prices Eurostat
(2005¼100, in PPP) in relation to the
respective EU average.
Purchasing power Purchasing power standards (EU-15¼1) in Eurostat
standards relation to the respective EU average.
Labor cost Labor costs (wage costs and payroll costs) per Eurostat
hour (in PPP, for the manufacturing sector
and for companies with 10 or more
employees) in relation to the respective EU
average.
Long-term interest Long-term interest rates according to the Eurostat
rate Maastricht criteria (10-year government
bonds) in relation to the respective EU
average.
Public debt ratio Gross government debt as a percentage of Eurostat
GDP in relation to the respective EU average.
Consumer tax rate Implicit tax rate on consumption (consump- Eurostat
tion tax revenues in relation to private con-
sumption spending) in relation to the
respective EU average.
(continued)
82 J. K€
onig
where the difference between a country’s value and the average value of the
remaining EU countries represents the degree of heterogeneity. If the difference
between the two variables is 0, the maximum degree of homogeneity is achieved.
Absolute values are considered in this equation as for homogeneity it is irrelevant
whether a value deviates positively or negatively from the EU average.
The co-movement of business cycles between the member states in the dimen-
sion ‘EU symmetry’ is measured as:
I i, t ¼ corr V i, τ ; Vj, τ 100 ð3:4Þ
where the correlation coefficient between a country’s values and the average values
of the remaining EU countries indicates the level of integration. The correlation
takes into account period τ, covering the preceding 5 years (20 quartiles) for each
value. A positive correlation of 1 represents the highest possible level of European
economic integration in this field.9
A country’s participation in the Schengen Agreement and its EMU membership
gives the following index scores:
9
Negative correlation values are tolerated here because a value of less than zero represents an anti-
cyclical behavior of a country’s figures and should therefore be treated as disintegration.
84 J. K€
onig
8
< 0, if having flexible exchange rates
I i, t ¼ 50, if participating in the Exchange Rate Mechanism II ð3:5Þ
:
100, if being a member of the European Monetary Union
0, if staying out of the Schengen Agreement
I i, t ¼ ð3:6Þ
100, if participating in the Schengen Agreement
Weights can have a large effect on the outcome of the overall index and country
rankings. Selecting an appropriate weighting procedure is therefore fundamental to
the successful construction of a composite indicator. The a priori weighting proce-
dure performed by some indices (e.g., the Kearney/Foreign Policy Globalization
Index or the Human Development Index of the United Nations) is not considered
accurate weighting due to the lack of objectivity in the assigned weights. Here, the
importance of one indicator over others rests solely upon the subjective belief of the
expert. This approach leads inevitably to a bias in the final results. The results of
indices using a priori weighting are further criticized in terms of sensitivity to
alternative weighting schemes (Lockwood 2004). Instead, the weights should be
generated on statistical grounds. Considering the statistical structure of the data set
used in the index ensures the calculation of objective weights that are not influenced
by the expert’s opinion. A sound statistical procedure, therefore, respects the
statistical relevance and informative value given by each indicator with regards to
the relative contribution to the overall index (OECD 2008).
The EU Index uses such a statistical weighting scheme by computing the
weights with the principal component analysis (PCA). Originally designed by
Pearson (1901), Spearman (1904) and Hotelling (1933) to analyze and reduce the
3 The EU Index of Integration Effort 85
10
See, for instance, the index of Economic Freedom by the Fraser Institute, the CSGR Globaliza-
tion Index or the KOF Index of Globalization.
11
Other common measures such as the Kaiser-Guttman criterion or the Parallel Analysis do not
lead to reasonable results here as there are too many components with eigenvalues close to 1.
86 J. K€
onig
4
Eigenvalues
0
1 4 7 10 13 16 19 22 25
Principal components
Fig. 3.4 Scree test of the principal component analysis (Notes: The obvious ‘kink’ at component
4 indicates that only three components should be extracted. The smooth decrease in eigenvalues
after component 4 points at random correlations and can be neglected)
than 1 and where two or more factor loadings of an indicator reach similar sizes. Thus,
the horizontal sum of all three factor loadings – each squared and multiplied by the
respective share of total variance of the component – eventually assigns the overall
weight to each indicator. These weights are presented in the last column of Table 3.3.
Multiplying the weights with the respective indicator finally leads to the indi-
vidual index scores of each country. The country rankings are calculated for each
dimension of EU integration as well as for a country’s overall level of integration
efforts. The results of the EU Index are presented in the next section. It is further
briefly analyzed whether the EU countries are on the verge to become a more
homogeneous or heterogeneous community.
Table 3.4 presents the results of the EU Index for the EU-25 countries. As ten
countries entered the union in 2004, the EU Index presented here is calculated for
the years 2004–2012.12
Belgium is the top performing country with respect to overall European eco-
nomic integration in 2012, accomplishing 75.3 of 100 possible index scores. At
some distance, Ireland and Austria reach second and third places, followed by
Germany, France and the Netherlands. Four of the six best performing countries
12
The original EU Index developed by K€ onig and Ohr (2013) was calculated for the EU-15
countries over the period 1999–2010. As Luxembourg shows many extreme values (e.g. GDP
per capita) it is not considered in either EU Index.
Table 3.3 Rotated factor loadings and computed weights of the EU Index
Rotated factor loadinga Weight (%)b
Comp 1 Comp 2 Comp 3 Comp 1 Comp 2 Comp 3 Overall weight (%)c
Single market EU-openness to goods 0.434 0.039 0.049 7.1d 0.1 0.1 7.2
EU-openness to services 0.281 0.100 0.093 3.0d 0.4 0.2 3.6
EU-openness to capital 0.390 0.020 0.081 5.7d 0.0 0.2 5.9
3 The EU Index of Integration Effort
Table 3.4 Results of the EU-25 Index for 2004 and 2012
EU Index 2004 EU Index 2012
Rank Country Score Rank Country Score
1 Belgium 66.3 1 Belgium 75.3
2 Netherlands 59.9 2 Ireland 70.5
3 Ireland 58.3 3 Austria 69.6
4 Finland 57.8 4 Germany 66.1
5 Cyprus 56.1 5 France 66.0
6 Germany 56.1 6 Netherlands 65.5
7 Austria 56.0 7 Slovakia 65.5
8 Spain 55.3 8 Finland 65.1
9 France 52.9 9 Spain 64.3
EU-25 51.9 10 Slovenia 63.3
10 Denmark 51.7 11 Malta 62.5
11 Portugal 51.5 EU-25 61.4
12 Sweden 50.9 12 Czech Republic 61.3
13 Italy 50.1 13 Cyprus 60.8
14 Malta 50.0 14 Italy 60.4
15 United Kingdom 48.2 15 Portugal 59.6
16 Slovenia 47.8 16 Denmark 57.3
17 Estonia 47.6 17 United Kingdom 56.8
18 Czech Republic 47.4 18 Estonia 56.3
19 Slovakia 46.9 19 Sweden 55.3
20 Greece 46.2 20 Lithuania 53.5
21 Poland 44.9 21 Hungary 51.9
22 Hungary 43.7 22 Latvia 51.7
23 Lithuania 39.9 23 Poland 50.7
24 Latvia 36.2 24 Greece 46.8
Source: Own calculations, www.eu-index.org
belong to the founding members of the European Coal and Steel Community. Only
the founding member Italy (ranked 14th in 2012) does not belong to the top group.
With Slovakia and Slovenia, also two of the Central and Eastern European Coun-
tries (CEECs) that joined the EU in 2004 are among the ten best performing
countries. Most of the CEECs, though, show very low integration efforts when
compared to the EU-25 average. Lithuania, Hungary, Latvia and Poland show
especially low levels of integration. The three EMU opt-outs (Denmark, UK and
Sweden) also show fairly low levels of integration efforts. At the very end of the
2012 ranking is Greece, achieving less than 50 index scores.
This relatively large discrepancy between the most and least integrated countries
was already present in 2004, yet at lower levels. Belgium achieved 66.3 index
scores reaching first place again, and being followed (more or less) by the same
countries as in 2012. The CEECs are skewed towards the very bottom of the index
scale – together with Greece. When compared to 2012, the lack of integration
90 J. K€
onig
90
80
70
60
50
40
30
2004 2005 2006 2007 2008 2009 2010 2011 2012
Fig. 3.5 Development of EU Integration Dimensions (Notes: EU-25 average index scores on the
vertical axis)
efforts made by Greece since 2004 becomes considerably visible. More precisely,
Greece still shows very low levels of single market integration (particularly in the
goods and capital sections), high debt ratios and long-term interest rates, low
symmetry in GDP growth and unemployment rates, and too many open infringe-
ment cases. With 46.2 and 46.8 index scores in 2004 and 2012, respectively, Greece
has hardly made any additional efforts in total EU integration.
The EU-average, on the other hand, has increased by roughly 20% since 2004.
The largest boost in integration efforts was achieved in Slovakia and Slovenia,
followed by Latvia, the Czech Republic, Austria and Lithuania. Hence, it seems
that some of the CEECs are catching up to the top performing countries in terms of
EU integration, even if most of the CEECs are still below the EU-average in 2012.13
Does the increase in integration efforts of almost all member states also imply that
the EU is becoming a more homogeneous community? Figure 3.5 raises some doubts
in this regard. By illustrating the average development of each EU dimension over
time, it becomes evident that the homogeneity dimension (economic convergence) of
the EU Index has not improved over the years. Whereas the symmetry and the
conformity dimension have made substantial improvements in integration efforts,
the homogeneity dimension even sees a slight decrease in integration efforts. These
disintegration tendencies have occurred especially due to different labor costs, long-
term interest rates, public debt ratios and capital tax rates.
By performing cluster analysis it is further shown that the EU has not become
more homogeneous but rather heterogeneous with several (relatively) homoge-
neous country groups. Figure 3.6 presents the results of the cluster analysis using
the 2004 data set. Squared Euclidean distances are used to measure the relative
13
Among the CEECs, Hungary and Poland have made the lowest integration efforts since 2004.
For a more elaborate view on the integration profiles for Hungary and the CEECs see Palankai and
Miklos (this volume).
3 The EU Index of Integration Effort 91
Fig. 3.6 Dendrogram for 2004 (Notes: Cluster analysis based on 25 indicators used to calculate
the EU Index in 2004 (using Ward’s clustering))
distances between the countries (and country groups): The lower the measured
distance, the more homogeneous is the country pair, respectively the country group.
It is very well shown that the EU-25 of 2004 was mainly divided into two parts: the
EU-15 countries on one side and the ten new EU member states (plus UK and
Ireland) on the other. The EU-15 group consisted of two subgroups each containing
six EU countries, one led by Austria and Germany and the other by Finland and the
Netherlands. The other (new EU member) group is also marked by two subgroups:
one of the CEECs and another consisting of the island states of Malta, Cyprus, the
UK and Ireland.
When performing cluster analysis using the 2012 data set, as presented in
Fig. 3.7, the country groups are changing in terms of size and relative distance.
The former EU-15 group has become much smaller. There is now one subgroup of
six countries, again led by Austria and Germany, and another consisting of the three
EMU opt-outs (Denmark, Sweden and the UK). The first subgroup may be regarded
as the ‘core group’ of EU integration as here the relative distances between
the countries are the lowest. Finland, the Netherlands, France and Belgium belong
to this core group, in addition to Austria and Germany. This core group seems to
be fairly homogeneous and also shows some homogeneity with the three
EMU opt-outs.
Further away from the core countries of European integration is a large (and
relatively homogeneous) country group that is dominated by the CEECs (with
92 J. K€
onig
Greece
Portugal
Italy
Spain
Cyprus
Ireland
Malta
Estonia
Hungary
Slovakia
Slovenia
Lithuania
Latvia
Poland
Czech Republic
United Kingdom
Sweden
Denmark
Netherlands
Finland
Austria
Germany
France
Belgium
0 20 40 60
Squared Euclidean Distance
Fig. 3.7 Dendrogram for 2012 (Notes: Cluster analysis based on 25 indicators used to calculate
the EU Index in 2012 (using Ward’s clustering))
Malta). Among the CEECs, Lithuania and Latvia show again the most homoge-
neous integration pattern. With regards to 2004, however, the CEECs were not able
to substantially decrease their relative distance to the core group.
At even larger distances to the core group, a new formation of countries has
emerged in 2012: the GIIPS (plus Cyprus).14 This new formation (of which Greece
is the ultimate outsider) interestingly consists of exactly those EMU members that
needed financial assistance from the European Stability Mechanism (ESM) due to
the eruption of the Global Financial Crisis in 2008. Greece, Italy, Portugal and
Spain belonged to the EU-15 group in 2004, but in 2012, the distance to the core
group is very large and, especially in Greece, much integration efforts are needed to
get back on track.
The cluster analysis of the 2012 data implies that the former ‘two-speed
Europe’ – mainly characterized by the EU-15 and the new member states – has
evolved into a ‘multi-speed Europe’. This multi-speed Europe sees strong and
increasing heterogeneity between the core group of EU integration, the CEECs
and the GIIPS. This tendency is of particular concern as the rising economic
heterogeneity stokes certain fears among the EU citizens which in turn leads to
14
GIIPS (Greece, Ireland, Italy, Portugal, Spain).
3 The EU Index of Integration Effort 93
more radical and nationalist attitudes. It should therefore stand at the forefront of
European policy to reduce this heterogeneity.
3.4 Conclusions
References
Badinger, H., & Breuss, F. (2011). The quantitative effects of European post-war economic
integration. In M. N. Jovanovic (Ed.), International handbook on the economics of integration
(Vol. III, pp. 285–315). Cheltenham: Elgar.
Balassa, B. (1961). The theory of economic integration. Homewood: Richard D. Irwin.
Baldwin, R. (1989). The growth effects of 1992. Economic Policy, 9, 247–281.
Boltho, A., & Eichengreen, B. (2008). The economic impact of European integration, CEPR
Discussion Paper, (6820). London: Centre for Economic Policy Research.
94 J. K€
onig
Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research,
1, 245–276.
Cecchini, P., Catinat, M., & Jacquemin, A. (1988). The European challenge, 1992: The benefits of
a single market. Aldershot: Gower.
Crowley, P. M. (2006). Is there a logical integration sequence after EMU? Journal of Economic
Integration, 21, 1–20.
De Grauwe, P. (2013). The political economy of the Euro. Annual Review of Political Science, 16,
153–170.
European Commission. (2013). Internal market scoreboard, (26). Brussels.
Frankel, J., & Rose, A. (1998). The endogeneity of the optimum currency area criteria. Economic
Journal, 108, 1009–1025.
Hotelling, H. (1933). Analysis of a complex of statistical variables into principal components.
Journal of Educational Psychology, 24, 417–441.
Ilzkovitz, F., Dierx, A., Kovacs, V., & Sousa, N. (2007). Steps towards a deeper economic
integration: The internal market in the 21st century, European Economy – Economic Papers,
(271). Brussels: European Commission.
K€onig, J., & Ohr, R. (2013). Different efforts in European economic integration: Implications of
the EU Index. Journal of Common Market Studies, 51, 1074–1090.
Krugman, P. (1991). Increasing returns and economic geography. Journal of Political Economy,
99, 483–499.
Lockwood, B. (2004). How robust is the Kearney/Foreign policy globalization index? The World
Economy, 27, 507–523.
Lucas, R. (1990). Why doesn’t capital flow from rich to poor countries? American Economic
Review, 80(2), 92–96.
Molle, W. (2006). The economics of European integration. Aldershot: Ashgate.
Mundell, R. (1961). A theory of optimum currency areas. American Economic Review, 51,
657–665.
Ohr, R. (2009). European monetary union at ten: Had the German Maastricht critics been wrong?
(Cege-Discussion Paper, 141).
Organisation for Economic Co-operation and Development (OECD). (2008). Handbook on
constructing composite indicators: Methodology and users guide. Paris: OECD.
Pearson, K. (1901). On lines and places of closest fit to systems of points in space. Philosophical
Magazine, 2(6), 559–575.
Romer, P. M. (1986). Increasing returns and long-run growth. Journal of Political Economy, 94,
1002–1037.
Solow, R. M. (1956). A contribution to the theory of economic growth. Quarterly Journal of
Economics, 70, 65–94.
Spearman, C. (1904). ‘General intelligence’, objectively determined and measured. American
Journal of Psychology, 15, 201–293.
Swan, T. W. (1956). Economic growth and capital accumulation. The Economic Record, 32,
334–361.
Tinbergen, J. (1954). International economic integration. Amsterdam: Elsevier.
Chapter 4
Integration Profiles for Central Europe
and Hungary
Concerning the basic forms of integration, Bela Balassa’s list still serves as a
reference point. Accordingly, he distinguishes the free trade area, the customs
union, the common market and economic and political union as the stages of
integration (though these are not necessarily hierarchical). Balassa does not men-
tion the single market, which, however, can be considered as the common market in
its full and complete implementation.
From a cost-benefit analysis point of view, there is no doubt that customs union
theories still form the basis. With the help of “classical” customs union theories
(trade creation and division, extension and contraction) a cost-benefit analysis can
be quantified with relatively simple equations in terms of efficiency and welfare.
The advantage of this approach is that, based on the calculation of tariff equiva-
lences, the analysis can be extended to the impacts of other trade obstacles
(non-tariff barriers).
The same applies to a monetary union, by the “tariffication” of the conversion
costs or exchange risks. Consequently, the “trade creation” impacts of a monetary
union can be calculated. The limit of the theory is that the comparative advantages
from integration are identified in a static approach, while it is clear that the gains
could be much more substantial. Nevertheless, a static comparative analysis
complemented the stage of “negative integration”.
Later, the emphasis shifted to “dynamic” customs union theories, which
extended their analysis to economies of scale and the impacts of the intensification
of competition and technical progress (innovations). It was clear that, upon this
basis, the possible gains from integration are much larger, while their calculation is
more complicated. Their macro-economic aggregation is much more difficult,
allowing for a larger margin of error. This approach raises the problem of differ-
ences in the impact of micro- and macro-level processes; it is clear that the latter
cannot be based solely on the aggregation of the former, as they are different
qualities. According to dynamic theories, the impacts of integration result in either
more rapid or slower economic growth. But growth is a complex process, and it is
4 Integration Profiles for Central Europe and Hungary 97
difficult or almost impossible to separate or filter out the impacts that integration
has on it. As we are moving towards “positive integration”, new approaches are
needed. In order to balance the advantages and disadvantages of integration, we
also have to include fields such as institutional or public service economics, and
system or growth theories.
Regarding regulation and governance, the major issues constitute their scope,
competence, efficiency and democratic character. A large number of studies have
tried to analyze the extension, intensity and efficiency of regulation. There have
been efforts to quantify these, including the following: applied forms and their
extension; norms and prescriptions; their application and compliance with them;
extension of free trade; institutions and their competences; policy harmonization
and coordination and their depth; the fields of common policies and their efficiency,
etc. The state of political integration and its democratic character are particular
issues of European integration. The analysis of regulatory integration provides first-
rate possibilities for the comparison of the development of different regional
integrations and their relation to the global integration (multi-level governance).
The content of integration needs a complex analysis and can be defined in all
dimensions of the process, particularly in relation to real economy integration and
its institutional frameworks. The real economy approach is based on two schools of
theory of integration, namely international division of labor and interdependence.
Accordingly, integration can be defined as an intensive, complex, durable and
institutionalized cooperation. One of its novel features is that it brings partners
into a new type of relation, which can be referred to as interdependence. These
developments can be measured both as processes and as states of integration.
cultures can influence the knowledge and views of the participants on both sides,
and can contribute to the mutual understanding of different cultures.
One of the main objections against globalization indices is that they are country-
centered, and they overvalue the role of smaller countries. Consequently, important
structural and functional features of the global economy are missed, resulting in
distorted pictures, particularly concerning the global role of larger countries.
According to the 2011 KOF Index, the US ranks in the 35th place, and in terms
of economic globalization it is listed only as 79th.
The US, however, is the number one power of the global economy. It contributes
one quarter of world production and 14% of global trade. It is the biggest global
investor, but external investments abroad are “only” about 20% of its GDP. In the
case of the Netherlands, the same proportion is 68%. In 2010, the stock of American
capital invested abroad was about five times more than that of the Netherlands.
About 40% of world trade is conducted in US dollars. In 2010, 61.5% of interna-
tional currency reserves were held in the American currency (26% in Euro). The US
has military bases in about 30 countries of the world. The role of China is similarly
under-estimated by the different indices.
Integration profiles are our attempt and provide one possibility to measure integra-
tion. It is considered neither better, nor worse than any other method, as most of its
parameters and characteristics are similar to the other ones, including its shortcom-
ings. However, some of its approaches are different.
We use several indicators to create an integration profile. They are similar to
those used, for example, by globalization indices, but the fields covered are
narrower. Mostly economic dimensions are concentrated on, and as far as the
indicators are concerned, stocks are preferred (flow indicators are used only excep-
tionally). In fact, the latter might refine, but not substantially change, the overall
picture.
The analysis is extended in certain directions by the profile. Apart from relations
and their intensity, the structural, performance, dependence, equilibrium and con-
vergence aspects are also measured. Accordingly, we try to go beyond the quanti-
tative characteristics and refer to the quality, and indirectly to the efficiency, of the
integration process.
We do not wish to establish rankings, and, therefore, there is no need for the
aggregation of parameters. Consequently, the picture can be far more telling, as
averaging tends to gloss over important details. Furthermore, by changing the
weights, the rankings can change drastically, which might reduce the value of the
analysis and its usability for economic policy decisions (problems which are
out-averaged).
The aim of the creation of an integration profile is creating a picture composed of
several mosaics. Needless to say, by putting the mosaics together, the final picture
100 T. Palankai and G. Miklos
should be complex and coherent. Therefore, we need information that makes the
picture visible and constructive. The picture should be multi-dimensional, concen-
trating on real integration processes, and should be able to measure the state of
integration (integratedness), participation in the institutional system, the regulative
processes and their quality and efficiency. The profile cannot be drawn without the
analysis of success and the balance of costs and benefits. However, we cannot
provide the detailed measurement of features that are beyond the scope and
possibilities of our work. Thus, as far as the cost-benefit analysis or integration
maturity is concerned, only references will be made to them.
The chapter focuses on the integration profile of (Eastern) Central Europe,
particularly Hungary. Nevertheless, this is not possible without a comparison
with other nations, and these comparisons are relevant only if they concern all the
28 countries of the EU.
The 28 EU members cannot be considered as a monolithic and uniform group. In
fact, based on their sub-regional characteristics (supported by a later analysis), the
28 members can be placed into different groups. First, we have to separate the “old”
and the “new” members. As far as the old members are concerned, a distinction
between Northwestern (Austria, Belgium, Netherlands, Luxemburg and Germany)
and Southwestern Europe (France, Italy, Ireland and UK) is easily justified based on
certain parameters of performance and connectedness. Under certain parameters,
the UK or Ireland can be placed into either group. The performance pattern of the
UK is more similar to South Western Europe and it has historical colonial relations
(Malta and Cyprus) with the region. Ireland has been on the Western periphery of
Europe, but at least in terms of per capita GDP it overtook several developed
partners. In a more strictly defined Northern Europe, the Scandinavian EU (Den-
mark, Sweden and Finland) should be distinguished; in terms of a number of
parameters, Norway can be considered as part of this group. These three groups
form the (“Centrum”) or Northern Centre of the Union.
Today, Northern Western Europe and the Scandinavian EU are one of the most
developed and integrated regions of the world. They mostly correspond to the
objectives of the Lisbon Program (the most competitive regions of global econ-
omy), except for dynamism, which was hoped to be achieved by the whole Union
by 2010. In some respects, they represent a certain sort of “Core” inside the
“Centre”, while Southern Western Europe distinguishes itself with poorer macro-
economic and integration performance.
In the last decades, the Southern EU members (Greece, Portugal and Spain) have
caught up with the other members, but their convergence has been contradictory.
These countries are often considered as the Southern Periphery of the European
Union. The latest crisis has been a demonstration of the North-South rift of the
Union. Lately, South-Western Europe and the Southern Periphery had to face
similar crisis phenomena (e.g. the sovereign debt crisis). Portugal, Italy, Greece
and Spain (PIGS) were in a very similar situation and were saved only by the rescue
operations of the Union.
Based on their performance and integration parameters, four groups can be
distinguished among the recent new members. There are clear differences between
4 Integration Profiles for Central Europe and Hungary 101
Central Europe (the Czech Republic, Hungary, Poland, Slovakia and Slovenia, and
the newly joined Croatia) and the Eastern Balkans (Bulgaria and Romania). The
latter group also claims a Central European status for itself, but on the basis of
historical and cultural traditions (Catholic-Protestant and Orthodox-Islamic), the
distinction remains justifiable (although Romania is a special case). On the basis of
their very different characteristics, the Baltic EU (Estonia, Latvia and Lithuania)
clearly distinguishes itself. These three groups can be considered as the Eastern
Periphery of the Union. The two Mediterranean new members (Cyprus and Malta)
are part of the Southern EU (and periphery). When analyzing these groups, we
would like to refrain from ranking the countries within their groups, even though we
do consider these sub-regional groupings as relevant clusters of analysis.
Although the focus here is on Hungary, special attention and reference will be
paid to the other Central European new members as well. In reality, these countries
should be referred to as “East-Central Europe”, and based on historical, geograph-
ical and economic factors and ties, the Western extension of this sub-region should
not be neglected. Here, West-Central Europe is defined as being composed of
Austria, Germany and Italy, although some place Switzerland also into this group
(Switzerland is a member of the Central European Initiative). In a narrower sense,
only Southern Germany (Bavaria or Baden-Württemberg) and North Italy (Veneto
or Lombardy) belong to this region (parts of the former Habsburg Empire), but the
separation would be too complicated. Historically, Germany often defines itself as
Central Europe (Mittel-Europa).
The main parameters and dimensions of the profile analysis can be summarized
as the following:
Real-economic integration (integratedness)
• Trade integration: flows, stocks, intensities;
• Structure of trade relations;
• Sub-regional concentration and interconnectedness;
• Intra-sector trade, place in value chains;
• Factor integration: flows, stocks, intensities;
• State and characteristics of financial integration;
• Transnationalization of company sectors.
Institutional and regulatory integration:
• Participation in global institutional and policy structures, and regulatory
processes;
• Compliance with EU institutions and policies (acquis communautaire);
Comparative performances:
• Competitiveness of economy;
• State of knowledge based society;
• External economic performances, balancedness of integration (trade bal-
ances, relation of capital export and import);
• Macro-economic performances;
102 T. Palankai and G. Miklos
indicate openness and dependencies. The higher the indices, the higher the
country’s openness and its dependence from external factors and processes.
• Share of regional trade to the total.
Txi/Txt100, or Tmi/Tmt100.
This formula measures regional trade intensity, but at the same time conclu-
sions can be made about the role of external relations. If the regional proportions
are high, it indicates that the country is inward-looking, and the importance of
external relations is smaller, or vice versa. We already referred to the fact that
these proportions are highly dependent on the size and the level of the develop-
ment of the country. Corrections in this respect are difficult, but other parameters
of the profile can refine the picture.
• Comparison of per capita trade, either in terms of the total or regional trade.
Txt/P and Txi/P, or Tmt/P and Tmi/P.
(P ¼ population of the country).
The per capita trade brings the level of development (correlation with per
capita GDP), or to some extent, the size of the country into the picture. The
indicators refer to the depth and quality of integration.
When the intensity of cooperation is analyzed in terms of trade and GDP
relations, we propose the following five different clusters.
Scaling of level of intensity (dependence)
0 10% 30% 50% 70% 100%
I No I Low I Medium I High I Very High I
Table 4.2 Share of trade of goods and services in GDP in the EU (2008, in %)
Export Export of Import Import of Internal Internal
of goods and of goods and export in export in
Country goods services goods services total GDP
Western European core countries
Belgium 73.4 91.0 76.5 92.9 75.9 69.1
Netherlands 60.7 72.7 54.3 64.8 78.9 57.3
Germany 40.7 47.4 33.6 41.4 63.3 30.0
Ireland 43.9 81.2 31.1 71.3 62.8 50.0
France 20.9 26.5 24.0 28.9 63.9 17.0
Italy 23.6 28.9 23.6 29.4 58.9 17.0
Luxemburg 39.7 167.9 51.4 126.8 88.9 147.0
Austria 45.1 60.2 45.2 55.5 67.5 40.7
U. K. 17.4 28.1 23.8 31.4 57.0 16.0
Scandinavian EU
Denmark 33.6 54.7 34.0 52.3 69.8 23.5
Finland 35.5 46.0 32.3 42.0 55.9 19.8
Sweden 38.5 53.5 34.7 46.1 60.1 23.1
Southern EU
Greece 8.2 22.2 26.3 33.3 62.5 13.9
Spain. 17.7 26.6 25.7 32.2 69.6 18.1
Cyprus 7.7 49.8 42.3 62.1 69.3 34.5
Malta 36.1 80.2 57.0 83.9 46.8 37.5
Portugal 23.0 33.8 35.8 42.6 74.4 25.1
Baltic countries
Estonia 53.8 76.1 65.7 80.4 70.1 53.3
Latvia 28.0 41.4 45.0 54.4 68.6 28.4
Lithuania 49.8 60.0 61.4 70.6 60.3 36.2
Central Europe (East)
Hungary 68.3 81.2 68.2 80.3 78.2 63.5
Czech Rep. 66.5 76.7 63.7 75.7 84.9 64.7
Slovakia 73.5 82.4 74.6 84.3 85.4 70.4
Poland 33.2 39.9 37.8 43.5 77.8 34.4
Slovenia 54.0 68.0 61.0 70.2 68.1 46.3
East Balkans
Bulgaria 44.8 70.5 70.4 83.7 60.0 42.3
Romania 24.5 30.9 37.9 43.7 70.5 21.7
EU27 67.5
Japan 38.0 48.0 18.8 28.6 - -
USA 15.3 18.4 14.5 18.0 - -
Source: Eurostat (2009)
106 T. Palankai and G. Miklos
In the last decades there has been a particularly rapid growth of internal trade,
which has brought substantial changes between the proportions of internal and
external trade (Table 4.3).
The share of internal trade of EC in total was about 30% in 1958, increasing to
50% by the early 1970s. This was followed by a slowing down, but by the end of
1990s it reached 62%. The 2000s saw a certain “disintegration”, particularly as far
as the old member countries were concerned. In the 1990s, trade between the
Eastern candidates and the old members was particularly rapid; it can be claimed
that these relations had a galvanizing impact on the integration process itself. In
1990, due to the former East-West division, the EC share of trade with the former
Soviet bloc countries was only around 25–30% of their total trade. Trade relations
in the 1990s increased spectacularly, due to trade liberalizations. In just 10 years,
they approached 70%. After the accession of these countries, the trend turned, and
in the past 10 years the percentage has drastically decreased (by more than 10%
points). Pre-accession trade was boosted by association agreements and by the
gradual overhaul of the regulation of the internal market. Thus, in terms of trade
conditions, the 2004 entry brought no new conditions and possibilities. The only
sphere of trade expansion that remained was the rapid increase of cooperation
among new members.
Trade with non-member countries (such as Norway, Switzerland, Mexico, Korea
or Turkey) is regulated by free trade arrangements. Their share in the EU external
trade is around 15%. The trade pattern with these countries is similar to the internal
ones. Part of the associations is also based on free trade or preferential systems.
Most of the non-European OECD countries, including China, trade on a
non-discriminatory basis (MFN).
The per capita internal export was also used to measure intensity. The population
of Belgium and Hungary are both around 10 million. The Hungarian trade intensity
in % of the GDP is only 5% points lower than that of Belgium (73–68%), but the
internal export of Belgium is almost four times higher, while the difference in per
capita GDP is twice as high as that of Hungary. The Hungarian export intensities in
relation to the GDP are a little higher than the Czech intensities, but in per capita
4 Integration Profiles for Central Europe and Hungary 107
export terms the Czech data are about one third higher. The per capita GDP of
Greece and Portugal is about 40% higher than that of Hungary. The per capita
internal export of Hungary is almost twice that of Portugal and four times more than
that of Greece (Table 4.4).
All in all, Hungary is characterized by a high level of integration and intensity of
trade relations. The share of its internal trade to the total (78%) is also far above the
EU average (65%). The per capita export data also prove the relatively high
intensity of Hungarian integration, both in the global and the European contexts.
large extent on the impact of the CAP. The other sectors correspond to the patterns
of developed countries (Tables 4.5 and 4.6).
There are substantial differences among EU member countries in terms of the
structure of their internal EU trade (export). Concerning the agrarian internal export
(dispatches), Austria, Belgium, Italy, Poland and Portugal are close to the average
(with +/2% differences), and Hungary, Germany and Romania are somewhat
below that. Some countries are far above the average level with shares of 25.6%
(Greece), 20.4% (Denmark), 16.2% (Spain), 15.8% (France) and 14.8% (Bulgaria).
It should be noted that the high proportions of Denmark, France and, to a certain
extent, Spain are different from that of Greece and Bulgaria. The former have a
highly developed agriculture and are the main beneficiaries of the CAP, while the
high Bulgarian share is an expression rather of its structural underdevelopment.
Greece is a special case, thanks to its special, beneficial CAP arrangements. The
low share countries are also quite diverse, with Scandinavian Finland or Sweden, or
the Central European Czech Republic, Slovenia or Slovakia representing different
cases.
Besides agriculture, the shares of manufacturing, and particularly that of
machines and transport equipment, tell a lot about the level of internal integration.
Needless to say, the proportions are relative, and therefore, their comparison might
be misleading. What is striking, however, is that, compared to the close to average
proportions of the Western Core (except Belgium), Central Europe is above the
average, indicating a shift of the European manufacturing bases to the region, that
happened in the last decade. In other manufacturing goods they are also somewhat
above average, with certain trade-offs between the two product groups
(i.e. Hungary).
While Spain is a little above the average in these sectors, the low level of the
Southern EU is striking. This is particularly the case with Greece, concerning
machines and transport equipment (10%). The Greek share of the total internal
4 Integration Profiles for Central Europe and Hungary 109
variations, the other Central European countries, achieved export structures similar
to that of developed countries.
Nevertheless, an important point should be stressed. The bulk of the 30% share
of machines and transport equipment of Hungary in 1989 went to the Soviet bloc
countries, and was far from competitive on the developed markets. The present 60%
goes to markets where the products have to withstand global competition. By now,
Hungary and the other Central European countries have managed to acquire that
capacity.
Hungary, both in its EU-internal and EU-external relations is a net agrarian
exporter. In 2009, the Hungarian agrarian export to the EU was €3.5 bn., while its
import was €2.6 bn., amounting to about a surplus of 25%. In 1988, the Hungarian
agrarian export to the EC was 6.8 times more than the import. Due to the negative
impacts of the CAP, the difference fell to twice as much already by 2002. Full
membership brought an improved market access, but the trends remained. The main
reason – the insufficient competitiveness of the Hungarian agriculture – has not
changed. In 2009, in its external agrarian trade to the rest of the world, Hungary
realized close to half a billion surplus in Euros (Table 4.7).
In 1989, about half of the Central European exports was comprised of “sensitive
products”. Their share (iron and steel; chemicals; textiles and footwear; food and
live animals) was the following: Hungary: 58%, Bulgaria: 50%, Czechoslovakia:
45% and Poland: 44%. (Palankai 1997:117)
In 2009, 88% of Hungarian exports included manufactured goods (SITC 6–8),
and was sold on competitive global markets. The share of the same goods was 72%
in the total import. The same proportions for EU commodity exports were 83% and
63%, respectively.
In 1989, Hungary and other Central European countries had no high-tech exports
to the developed markets. The picture has changed drastically. As early as 1999, the
share of high-tech products in the total Hungarian export reached 25%, and by
2006, it further grew to 28%. These proportions, in 2006, were 17% for old
members, while for the new ones it reached 14% (both figures are average). The
proportions were around 15–17% in case of Estonia, the Czech Republic and
Slovakia. The new members, particularly the Central European ones, in these
respects, have shown remarkable catching-up. (European Economy 1/ 2009)
Besides structural changes, catching up in terms of competitiveness was also
marked by the development of productivity (Table 4.8).
The productivity growth between 1998 and 2008 will be measured here in terms
of per capita outputs and by working hours. Due to the rapid growth over the past
10 years, the Central European countries have caught up in terms of per capita
outputs from 40–50% to 60–70% of the EU27 average.
The rise of Slovakia has been particularly spectacular, going from 56% to 79%
of the EU average, and in light of the Slovenian 84%, it marks a convergence even
among the Central European countries. By 2008, Hungary (with 74%) and the
Czech Republic (with 72%) overtook Portugal (71%), while the Portugal per capita
GDP remained above the Central European rate. It is interesting to note the
differences between the Czech and the Hungarian levels of per capita GDP, trade
4 Integration Profiles for Central Europe and Hungary 111
and productivity. While the Czech per capita data are one third higher, the level of
Hungarian productivity is higher than the Czech one. The picture is somewhat
tainted by the productivity data expressed in output per working hours, which
indicate the differences in activity rates and in the number of working hours.
European countries (Czech Republic, Hungary and Slovakia) had surplus both in
their internal and external trade. At the same time, from the Core, Austria, France,
Italy and UK, plus all the countries of the Southern EU, the Baltics and the Eastern
Balkans produced deficits in their internal and external trade as well. Naturally,
these are the results of several factors, and they go far beyond the countries’
4 Integration Profiles for Central Europe and Hungary 113
Factor market integration is important from the point of view of both efficient
allocation of resources and exploitation of gains offered by integration.
Factor market integration is measured with labor and capital movements. Need-
less to say, both flows and stocks need to be examined. The export and import of
capital (Cx and Cm) indicates the dynamics of integration, and their share in GDP at
any given moment is an indicator of the intensity of integration.
Cx/Y 100 and Cm/Y 100.
The relation of capital export and capital import is also important from the
perspective of the level of development and balancedness: Cx/Cm 100.
In any given country, the high level of foreign investments can express a high
intensity of global or regional integration. However, minimal or zero capital export
implies unilateral dependence and asymmetry in the country’s integration. Gener-
ally, these indicators are balanced in highly developed countries, and this balance is
at a high level of intensity in both dimensions.
In the present world economy, the flow of labor is relatively moderate, and
migration is limited by strict legal regulations in most of the countries. This applies
4 Integration Profiles for Central Europe and Hungary 117
Table 4.11 Stock of FDI in % to the GDP in 2009, and the relation of stock of exported and
imported capital
Invested capital (stock) in Exported capital in Capital export in
Country GDP (%) GDP (%) import (%)
Western European core countries
Germany 24.5 33.5 137
Austria 24.1 22.7 95
Netherlands 70.6 101.1 143
Ireland 68.1 53.6 82
U. K. 44.9 57.2 140
France 35.3 49.1 139
Italy 15.1 19.4 128
Scandinavian EU
Denmark 50 64 128
Finland 30.3 42.7 140
Sweden 49.4 59.8 121
Southern EU
Greece 14.7 8.0 61
Spain. 34.0 39.4 116
Portugal 37.3 25.1 67
New Members
Estonia 72.7 20.7 28
Poland. 34.7 4.5 12.9
Czech Rep. 54.2 3.0 5.5
Hungary. 98.9 40.2 41
Slovakia 60 4 6.6
Slovenia 22.2 11.4 50
Bulgaria 63.2 0.9 1.5
Romania 35.3 0.7 2
EU27 17.7 23.2 131
USA 12.9 17.2 133
Japan 2.3 9.8 470
Sources: Eurostat (2009), OECD (2012)
the EU-average is 20%. In the case of the USA, this intensity indicator is only 17%,
while in the case of Japan it is below 10% (Table 4.11).
When measuring the intensity and level of dependence, we propose using the
same scale as in the case of trade intensities:
Scale of the level of intensity (dependence)
0 10% 30% 50% 70% 100%
I No I Low I Medium I High I Very High I
In terms of capital import, the Netherlands, Estonia and Hungary are all char-
acterized by very high capital integration intensity, with Ireland close behind. In
4 Integration Profiles for Central Europe and Hungary 119
light of actual Hungarian investment data, the Hungarian share seems a little
overestimated, but it is still close to 70% of the GDP. In order to keep the
comparisons consistent, we will rely on Eurostat data. There is no comparable
EU data for Slovakia in 2009, but the OECD data provide a good indication (the
same applies to Denmark). The Czech Republic and Bulgaria qualify for the high
intensity category, with Sweden and UK close behind. The high Central European
proportions are the result of massive investments in the region, particularly from the
middle of the 1990s onwards. The Dutch data reflect the traditional foreign
investor role.
Greece, Italy, Slovenia and Austria fall into the low intensity group. This
indicator also expresses Greece’s low intensity of integration into the global and
European economy. The lower capital import intensity of Germany should be
related to the size of the country, but in terms of capital export, the country belongs
to the medium intensity category. From this perspective, several countries have
been upgraded.
Most of the developed countries are characterized by net capital export positions
(about 30–40% over import). Besides intensity, these figures also reflect the
balancedness of their integration.
In the past decades, the process was influenced by the EU integration process. In
1998, the Spanish capital export was only 60% of its capital import. In 10 years, this
proportion has doubled, and Spain has become one of the most important investors
in South America. This also gives an indication of the country’s global orientation.
The same process can be observed in the case of Portugal, however, it remained a
net capital importer. From among the old members, Austria, Ireland and Greece are
also net importers. Although Ireland is at a high level of intensity, this intensity is
low in the case of Austria, and in case of Greece intensity is practically absent.
The last decade has witnessed the start of capital exports of the new Eastern
members. The pioneers in this process have been Slovenia, Hungary and Estonia. If
the ca. €15 billion Hungarian investments abroad are compared with the €80 billion
FDI in the country, the former amounts to only about 20%. In case of the other two
countries, these investments have just started, and they are around or below 10% in
relation to the FDI, reflecting the asymmetry of their integration. “The capital
export is closely related to level of development, and in case of expanding economy
the increase of that activity is a necessity. At the same time, there is no rule which
can predict how the level of development and the capital export should be related.
Therefore, in absolute terms, one cannot determine whether Hungary is ahead or
behind ‘normal’. This is influenced by the level of development of geographically
close countries (as they are more attractive as a terrain of potential investment), and
their capital absorption capacity.” (Világgazdaság 2004)
120 T. Palankai and G. Miklos
From the 1970s, the start of the capital export of European companies was accom-
panied by their growing transnationalization. From the 1980s, it was strengthened
by rapid integration and the transnationalization of international financial markets.
In this way, the economies of the Core EU countries have become highly
transnationalized, both in terms of their positions in the national economies and
in the global economy. The EU has become one of the main territories of density of
global integration, and the European TNCs play a leading role in this process.
According to estimates, about half of the TNCs have their origins in Europe.
Nevertheless, instead of TNCs, we should rather talk about transnational net-
works, which are combinations of large companies and a number of small- and
medium-sized enterprises (SMEs), covering research and development, production,
financing and various other services. SMEs, in this respect, can be placed into three
different categories. 1) Some are just subsidiaries, daughter companies or direct
suppliers of the larger partners. 2) Some are TNCs in their own right, as they have
high export shares, they export their capital abroad, they build close contractual
relations with foreign partners, and they follow transnational business strategies. A
large number of the SMEs of the core countries follow this pattern. 3) Local SMEs
fall seemingly out of this structure, as they operate on local markets, but they are
dependent on the supply of the TNCs, and most of them face direct competition
(small shops vis-a-vis big supermarket chains). Transnational networks have a
strong transatlantic character; the large European and American TNCs are closely
connected (among others with supply chains or bank backgrounds) and are placed
on each other’s markets. Recently, there has been a multiplication of TNCs of the
emerging countries.
Hungary and the new Eastern EU members, thanks to the high intensity of their
capital integration, are part of this transnationalization process. It is, however,
largely one-sided and asymmetric. This one-sidedness has started to diminish, but
the countries are still at the beginning of this process.
In the last years, several Hungarian (Central European) companies (MOL, OTP
Bank, Trigranit, Matav, Fornetti, etc.) have aspired for transnational positions and
followed transnational strategies. Their expansion, however, is limited mostly to
the neighboring countries.
From the largest 100 Hungarian companies in 2011, based on their export
involvement (i.e., more than 10% of the turnover goes to export) practically
two-thirds qualify for TNC status. Out of the other 36 companies, 19 are foreign-
owned, operating either in the energy (E-On, Shell or GDF Suez) or retail sector
(Tesco, Auchan, Lidl or Metro) and are subsidiaries, fully oriented towards local
markets. The 2011 list was led by MOL; the top ten also included AUDI, GE,
Samsung, Nokia, Philips and TESCO Global. From the first 100 companies there
are only 14 with a more than 33% Hungarian ownership, but most of them are from
the energy, transport and telecommunications sectors, and are typically large, state-
owned companies such as Magyar Posta (mail), Budapesti Gázművek (gas),
4 Integration Profiles for Central Europe and Hungary 121
Table 4.12 The relation structure of EU trade in 2009 (Percentage of internal export)a
Western-
Country Europe Germany Central Europeb Scandinavian EU Southern EU
Northwestern core countries
Belgium 84.9 25.8 4.4 (0.6) 3.7 6.2
Netherlands 81.5 33.3 5.6 (0.7) 4.8 6.7
Germany. 66.8 - 14.6 (1.7) 7.1 8.8
Ireland 86.2 11.4 2.2 (0.7) 2.6 8.4
Austria 69.4 43.1 19.9 (2.5) 2.7 4.2
Luxemburg 79.1 23.2 4.5 (0.6) 5.4 6.3
U. K. 64. 7 20.1 4.4 (0.7) 11.6
Southwestern Europe
France 71.5 25.6 5.9 (1.2) 3.7 16.2
Italy 65.1 22.0 10.6 (7.3) 3.4 16.8
U. K. 64.7 20.1 4.4 (0.7) 6.7 11.6
Scandinavian EU
Denmark 53.3 25.9 6.4 (0.8) 22.5 6.3
Finland 47.6 18.6 8.3 (0.8) 21.1 7.1
Sweden 60.8 17.4 6.7 (0.5) 23.6 6.1
Southern periphery
Greece 66.9 21.3 4.4 (4.7) 3.1 10.2
Spain 74.4 16.1 4.9 (0.9) 2.3 15.2
Cyprus 36.7 13.6 2.3 (2.9) 3.1 37.8
Malta 86.0 30.0 3.5 (0.7) 3.2 3.6
Portugal 55.4 17.4 2.8 (0.8) 3.2 35.7
Baltic countries
Estonia 23.6 8.8 3.8 (0.3) 49.6 1.5
Latvia 28.1 12.1 7.5 (0.6) 11.6 2.9
Lithuania 43.9 15.1 13.6 (3.0) 18.0 15.1
Central Europe
Hungary 66.0 32.3 16.5 (7.8) 2.8 5.8
Czech Rep. 69.3 38.3 20.7 (4.7) 3.4 3.7
Slovakia 56.4 22.6 33.3 (2.7) 3.2 3.9
Poland 69.5 32.8 14.1 (2.19) 6.7 2.2
Slovenia 73.2 28.0 18.4 (4.2) 2.5 2.7
Eastern Balkans
Bulgaria 53.2 17.4 7.3 (6.0) 1.7 21.0
Romania 60.1 23.6 22.2 (3.2) 1.7 5.6
Source: Eurostat (2011)
a
In every case the regional share does not include the given country
b
Central Europe + Eastern Balkans
also on the grounds of the countries’ level of development. Now we will examine
the EU members’ sub-regional connectedness (Table 4.12).
We wish to introduce the notion of “strategic partnership”, which means that the
partner’s share of a country’s or region’s trade is more than 10%. If these
4 Integration Profiles for Central Europe and Hungary 123
Table 4.13 Sub-regional trade structure of Central Europe in 2009 (Percentage of internal
export)a
West Central East Central New Strategic
Country Europe Europe membersb Total partners
West Central Europe
Germany 19.5 14.6 16.3 35.8 E-C-E
Italy 26.1 10.6 17.9 43.0 E-C-E
Austria 54.5 19.9 24.4 78.9 E-C-E
East Central Europe
Hungary 45.4 16.5 24.3 69.7 DE
Czech R. 49.0 20.7 25.0 74.0 DE, SK
Slovakia 36.4 33.3 36.0 72.4 DE, CZ
Poland 43.8 14.1 16.2 60.0 DE
Slovenia 54.8 18.4 22.6 77.4 DE, IT, FR,
AT.
Croatia 58.7 19.8 21.9 79.6 IT, DE, SL
Eastern Balkans
Bulgaria 38.7 7.3 20.6 59.3 DE, EL, IT
Romania 46.1 22.2 25.5 71.6 DE, IT. FR
EU27
Source: Eurostat (2011)
a
In every case the regional share does not include the given country
b
New Eastern Members – Eastern Central Europe and Eastern Balkans
Germany is the first partner for 21 countries. It is a second partner for Cyprus,
Lithuania, Portugal and Spain, and comes only as third for Latvia (12%) and fourth
for Estonia (8.8%).
Between some regions the connectedness is particularly intense. The high level
of connectedness of the new Central European members with the Western
European Core can be further specified. In fact, East Central Europe (which,
since 2013, also includes Croatia) is connected to the West largely through West
Central Europe (Austria, Germany and Italy), and their connectedness is close to
high (44–55%). Slovakia is at a lower level (36%), while Croatia is close to 60%.
On average, the intra-East Central European connectedness is around 20%. Poland
and Hungary are at 14–16%, while the Slovak figure of 33% reflects the intense
Czech and Slovak relations. All in all, the connectedness of Central Europe is very
high (roughly 70–80%; even the Polish level is above 60%).
Hungary is highly connected with the Western European core, with a concen-
tration on Central Europe and Germany (one third of its trade relations are with the
latter). Among its five largest export partners, France, Romania and the UK each
have a ca. 7% share. Hungarian external trade is equally concentrated. One third of
the Hungarian external export goes to Russia, the US and China, while these three
countries give 55% of its external import.
4 Integration Profiles for Central Europe and Hungary 125
Table 4.14 Intensity of interconnectedness for internal export in the main sub-regions of the EU
in 2009
Share export in total Share of population in total Interconnectedness
Region in % in % quotient
EU 27 s 100 100 1
Western 74.8 60.1 1.25
Europe
Scand. EU 6.0 5.0 1.20
East Cent. E. 10.5 10.3 1.02
Baltic c. 0.7 0.8 0.93
Southern EU 6.6 13.8 0.47
Latin EUa 23.9 36.2 0.66
Source: Eurostat (2011)
a
FR, IT, ES, PT
Hungary had €6.3 billion investments in the other 26 EU countries (the total
investments amounted to €6.8 billion). Hungarian investments were less than
€100 million in Germany and Austria, while they were over €2.5 billion in Slovakia
and €1 billion in Croatia. (Endrődi-Kovács et al. 2011: 21)
Table 4.16 Per capita GDP Country 1960 1973 1990 2004 2011
of EU countries between 1960
Belgium 113 112 115 121 118
and 2009 (PPP –
EU27¼100%) Finland 111 115 118 116 116
Greece 44 71 88 94 82
Portugal 41 59 62 77 77
Ireland 63 61 74 143 127
Spain 59 77 76 101 99
Hungary 60a – 41 63 66
1990b 2000 2004 2011
Germany 115 119 117 120
Austria 117 131 127 129
Hungary 41 56 63 66
Czech Rep. 61 69 75 80
Slovenia 74c 79 85 84
Slovakia 61 50 57 73
Poland 33 43 46 65
Sources: Economy (2003), Eurostat (2009)
a
Approximation to the European average, as compared to the
developed Western European countries, and not with the then
EEC
b
The new members’ data are for 1992, which was the peak year of
the transformation crisis
c
1995
of benefits of integration processes. The situation in the South was more contra-
dictory. It should be noted, however, that in spite of the convergence of the general
levels, intra-country regional differences might have remained.
The history of the relations of the Eastern periphery to the Western Centre took a
different path. According to historic data, in 1960, the level of development of
Hungary and Spain was about the same, while the Hungarian level was about 50%
higher than that of Portugal or Greece. But for Hungary and in fact, for the whole
Central and Eastern European region, the next 30 years could be described as lost
decades. Following Soviet policies, the period saw the implementation of “peaceful
competition”, during which the Soviet leadership wished to reach the American
level by 1980. The basis of this complacent conceit was the assumed superiority of
“socialism” over capitalism. This, however, was far from reality, and, on the
contrary, from the 1970s onwards emerging globalization brought to light the
total inaptitude of the bureaucratic and closed central planning systems in face of
global challenges. The result was a humiliating defeat and further peripheria-
lization. The differences were further aggravated by the 1989–1993 transformation
crisis. By the 1990s, the proportions turned around, and Portugal and Greece were
50% above the Hungarian average, while Spain almost doubled its lead over
Hungary (Table 4.16).
4 Integration Profiles for Central Europe and Hungary 129
Following the transformation crisis, from the middle of the 1990s onwards, new
trends have appeared, and it seems that the East has joined the development
processes of other peripheries.
Candidates’ – and later the new EU members’ – economic development was
particularly rapid. On average, these countries produced an about 2% growth
surplus, which would be sufficient for 20–30 years of convergence (depending on
the level of the individual countries). Some countries, in some periods, achieved a
spectacular growth performance (such as the 10% growth of the Baltic countries in
the early 2000s), which accelerated their convergence. As the post-2009 crisis hit
most of the new members seriously, the process of convergence was broken. This
warns us to be cautious about long-term expectations.
The convergence of new members was more marked concerning the economic
structures. Structural changes started and were provoked by the transformation
crisis, and they quickly approached the structural patterns of developed EU
countries.
In Hungary, between 1989 and 2001, the share of agriculture in the GDP fell
from 16% to 4%, while the share of services grew from 42% to 67.5%, which can be
considered as a remarkable convergence to EU averages. Similar changes happened
with the other new members. In fact, if we look at the 2009 data, certain differences
have remained, but they are not of a qualitative character, particularly as far as the
Core and the Central European members are concerned.
Structural convergences are also reflected in trade structures and the intensity of
relations. Needless to say, averages can cover large qualitative differences. The
fine-tuning of structural convergence takes a longer time.
It is generally accepted that the per capita GDP data in themselves are not
enough, and they can even distort the picture. Lately, several institutions
(or banks) have ventured to produce so-called complex convergence indicators,
which can give a little more appropriate and accurate picture about the state of
convergence. The complex approach implies that instead of the unilateral concen-
tration on per capita GDP, it also takes into account several other indicators
(Table 4.17).
Among others, we can mention the analyses of Deutsche Bank Research, World
Economy (Világgazdaság 2004) and the European Centre of ICEG (International
Centre of Economic Growth). The Deutsche Bank’s so-called convergence network
was based on five groups of indicators with 16 variables (growth dynamics, legal,
institutional and regulatory elements, financial and fiscal indicators, etc). According
to this analysis the 15 old EU members were taken as 100. The first group of
candidates was around 75% (Slovenia: 75.6; Czech Republic: 74.6; Hungary: 73.2
and Estonia: 72), the second group was around 66% (Latvia: 67.7; Slovakia: 67.5;
Lithuania and Poland: 65.2), while the Eastern Balkan countries were a little below
60 (Bulgaria: 58.7; Romania: 57.5%), as compared to the level of the old members.
(Deutsche Bank Research, 2001 data– unfortunately later dates have not been
published) (Palankai et al. 2011).
The results show quite a different picture. In the early 2000s, Hungary reached
only half of the EU15 average in terms of per capita GDP and fell more than 20%
130 T. Palankai and G. Miklos
The relation of Hungary to the world economy, before 1990, was characterized by a
special duality. As a small country, it had a structurally open economy, and the
share of export or import reached about 40% of the national product. At the same
time, in institutional and economic policy terms, it was a closed economy. The
Soviet-type centrally planned systems, in their external relations, applied very
strong discriminatory and protectionist regulations. After World War II, these
countries were left out of the process of global opening, both in trade (elimination
of discrimination and substantial liberalization in the framework of GATT), and in
financial-monetary relations (IMF – extension of convertibility).
In some regional integrations, substantial liberalization measures were
implemented, and some of them reached the full, or almost full, liberalization of
their trade relations (EC, EFTA etc.). By the 1970s, these laid the foundations for
the emerging process of global and regional integration. The countries of the Soviet
bloc remained out of these processes, even if the CMEA declared the integration
after 1971. But the Soviet model of “socialist integration” was not able to achieve
real integration. On the contrary, it failed to exploit even the minimum advantages
of the international division of labor. This integration failure played an important
role in the collapse of the Soviet system at the end of 1980s.
As a result of the post-1968 reforms, the Hungarian economy got into a special
situation. The reforms were accompanied by cautious and contradictory opening
measures (decentralization of trade decisions to companies or partial and limited
convertibility, possibility for “joint ventures”), but under the circumstances of low
efficiency and competitiveness, they lead to catastrophic consequences. The most
serious consequence was the heavy indebtedness of the country, and Hungary
became one of the most indebted countries of the world (for example, in terms of
per capita debt). Meanwhile, at the end of the 1980s, the reforms accelerated, and
compared to other Soviet bloc countries, Hungary (and Poland) took the lead in
economic and social transformation.
After 1988, the most important elements of transformation were the liberaliza-
tion of foreign investments, the gradual reduction of the elements of discrimination
and protectionism, and the introduction of convertibility. In the space of a few
years, practically by 1992–93, the institutional and policy opening of the Hungarian
economy was implemented, and these structures became similar to those of the
OECD countries. This meant a rapid and radical “negative” integration (liberaliza-
tion) of the country.
Concerning “positive” integration, joining the EU was an organic part of the
transformation strategy. In the process of the “normalization” of relations with the
EC, Hungary signed a Trade and Cooperation Agreement in 1988, which was
followed by full association (Europe Agreement), signed in 1991. The opening
and the integration process was crowned by the Accession Treaty, followed by full
EU membership in 2004. The other countries of the region followed the same path.
132 T. Palankai and G. Miklos
More than 90% of Hungarian foreign trade is conducted on a free trade basis.
Hungary is fully exposed to the global competition, and participates practically
fully in the Single Market (it complies with its rules as most of the old members).
Hungary is gradually entering the EMU (by meeting the Maastricht criteria, the
European Semester etc.), but it has not joined the Euro-zone yet.
In the last two decades, Hungary has become highly open, integrating its
economy in global, European, structural and institutional terms. Its economy is
characterized with a very high level of integration, which follows the pattern of
highly developed (small) countries. The process was accompanied with structural
convergence and balanced trade relations (surplus with the EU, including Ger-
many). The productivity of the country has increased rapidly, which among others
(such as relatively lower wage increases) has resulted in an improved competitive-
ness. In the global competitiveness ranking, from among the higher stratum of
developing countries (70-80th place) the country managed to reach a lower group of
developed countries. In the 2011 ranking of IMD, it was in the 47th place, but its
best position was in 2001, when it was ranked 26th. Due to faulty economic policies
after 2001, Hungary has lost its advantage and its development has stopped.
Hungarian modernization was largely based on foreign direct investments,
which increased the dualistic character of the economy. Against the highly com-
petitive foreign TNCs, there is a sector of local SMEs with low or absent compet-
itiveness. The main deficit of structural integration is that competitive export
capacities are largely based on import inputs, while the domestic value-added
contains low innovation and knowledge contents. These disproportions characterize
the supply channels and the trade of components. The transnationalization of the
domestic company sector has just started. There is a relatively high sub-regional
concentration of the country’s integration relations. While the average internal and
external trade proportions for the EU are two-thirds to one-third, in the case of
Hungary this proportion is four-fifths to one-fifth. At the same time, the global
exposition of the economy is also high; small fluctuations at the New York stock
exchange are immediately felt in Budapest. These are the main distortions and
deficiencies, which degrade the quality of integration of the country. Similar
tendencies and patterns (albeit with some divergences) characterize the other
Central European countries.
With its very high intensity of relations and level of connectedness to the
Western European core, the Central European periphery has got cloeser to the
Centre. In fact, it can be claimed that the Core has internalized the Central European
periphery. This has a number of consequences, and its characteristics and problems
would need further analysis. In this perspective, Central Europe has an opportunity
to become an organic component of the Centre, but broad and far-reaching changes
are still required. The patterns of periphery in Central Europe still exist, such as
structural dualities, loss of incomes (high debt service and risk premiums, indicated
losses through transfer prices, etc.), not to mention political and cultural factors. It
is clear that the process assumes adjustments and policies, and a lot depends on
proper national programs. These are largely missing. It must be noted that the
4 Integration Profiles for Central Europe and Hungary 133
strategic project of Europe 2020 also fails to address the convergence of the
Southern and Eastern peripheries in an appropriate way.
References
Central Statistical Office of Hungary (KSH). (2001). Material of the Association of Hungarian
Industrialists.
Endrődi-Kovács, V., Hegedűs, K., Szebeni, D. (2011). Magyar tőke aramlasok (Hungarian capital
flows). Tanulm anyok az europai interg
acio témak€ oréből (Studies on European Integration, 3).
Savaria University Press.
European Commission. (2001). Broad economic policy guidelines, European Economy, 72.
Brussels: European Commission.
European Economy. (2003). Statistical Annex.
European Economy. (2009). Five years of an enlarged EU (Economic achievements and chal-
lenges). European Commission, Directorate-General for Economic and Financial Affairs.
Eurostat. (2009). Europe in figures – Eurostat yearbook 2009. Luxembourg: European
Communities.
Eurostat. (2010). External and intra-EU trade – Statistical yearbook data 1958–2009. Luxem-
bourg: European Union.
Eurostat. (2011). Europe in figures – Eurostat yearbook 2011. Luxembourg: European Union.
Heti Világgazdaság (World Economy). (2013, January). Budapest.
Inotai, A. (1998). Magyarorsz ag az Europai Uni o és a 10 tarsult orsz
ag k€oz€otti k€
ulkereskedelem
jellemzői és új ir
anyzatai (Characteristics and New trends in the trade of Hungary, the EU and
the 10 associated countries). Budapest: OMFB.
Kemény, J., & Varga, G. (2011). A visegr adi orsz agok energiastruktúr aja 1990–2007 (Energy
structure of the Visegrad countries) Tanulm anyok az eur opai interg
acio témak€oréből (Studies
on European Integration, 3). Savaria University Press.
KOF Index of Globalization. (2010 and 2011) KOF (Konjunkturforschung). Zürich: Swiss Eco-
nomic Institute, Swiss Federal Institute of Technology.
OECD. (2012). International direct Investment Database. Eurostat, IMF.
Palankai, T. (1997). Integration and transformation of Central and Eastern Europe. Budapest:
Corvina.
Palankai, T. (2005). A magyar gazdaság és társadalom integráció-érettsége, integrációs képessége
és felkészültsége. (Integration maturity, capacities and preparedness of Hungarian economy
and society) Vil aggazdas agi tanszék és Eur opai Tanulm anyi és Oktat asi K€ozpont. Z aro
tanulm anyok. (Concluding studies). Budapest: OM NKFT. Budapesti Corvinus Egyetem. 463.
Palankai, T., Kengyel, Á., Kutasi, G., Benczes, I., & Nagy, S. G. (2011). A glob alis
alis és region
integracio gazdas agtana. Budapest: Akadémiai Kiadó.
Papanek, G. (2010, April). A gyorsan n€ ovekvő magyar kis- és k€ ozépvállalatok a gazdaság
motorjai. (The rapidly growing Hungarian SMEs as the engines of the economy).
K€ ozgazdas agi szemle, LVII.
Szél, K. (2010). Changes in commodity structure of Hungarian foreign trade. ECOSTAT
(Research Institute on Economy and Society).
Világgazdaság (World Economy). (2004, June). Budapest.
Világgazdaság (World Economy). (2007, November). Budapest.
Vujakovic, P. (2009, December). How to measure globalisation. A New Globalisation Index (NGI)
WIFO (Working Papers, 343).
Chapter 5
The EDB System of Indicators of Eurasian
Integration: Eurasian Integration’s Trends
from 1999 to 2012
For almost two decades, regional cooperation and integration has remained one of
the most pertinent issues of economic policy in the post-Soviet countries.1 There
have been hundreds of initiatives and projects that aimed at deeper bilateral and
regional cooperation and integration. In many cases, these initiatives had
overlapping membership and objectives, or they ceased to exist, or were
re-established by the same actors. Agreements with similar content, such as free
trade areas or customs unions, were signed over and over again by the same
countries. This variety of outcomes needed a comprehensive system to monitor
and assess the current processes of economic, political and social interaction
between countries. The CIS region did not possess any of these comprehensive
studies and measurements. Therefore the assessments normally had to be done on
ad-hoc basis. They were limited in terms of the scope of the type of cross-border
1
Throughout the paper, “post-Soviet space” refers to twelve former Soviet Union republics
(Armenia, Azerbaijan, Belarus, Georgia, Russia, Kazakhstan, Kyrgyz Republic, Tajikistan,
Uzbekistan, Ukraine, Turkmenistan and Moldova). Until recently all these countries belonged to
the Commonwealth of Independent States, the largest regional integration organization of post-
Soviet countries in terms of its membership. Georgia left the CIS in 2009, but is considered as part
of the region in the SIEI. Latvia, Lithuania and Estonia are not included in the post-Soviet space in
the sense the word is used in this article.
E. Vinokurov (*) • V. Pereboyev
Centre for Integration Studies, Eurasian Development Bank, St. Petersburg, Russia
e-mail: Vinokurov_EY@eabr.org
A. Libman
Ludwig Maximilian University of Munich, Munich, Germany
interaction covered, and the time span of the analysis. For example, much more
attention is typically given to cross-border trade of post-Soviet countries than to
other aspects of their interaction, such as migration flows or mutual investments.
However, it is questionable whether it is indeed the case that trade should constitute
the area where possible progress of integration across post-Soviet countries should
be observed first (Libman and Vinokurov 2012b).
The deficit became particularly pronounced in recent years, due to the major leap
forward in the development of post-Soviet regionalism associated with the estab-
lishment of the Eurasian Economic Union (EEU) (2015)2 by Belarus, Kazakhstan
and Russia. The EEU is based on the Customs Union (CU) of the Eurasian
Economic Community (EurAsEC)3 (2010) and the Single Economic Space
(2012) of these three countries. The Customs Union, unlike preceding regional
organizations in the post-Soviet space, has a major impact on the regulation of
cross-border trade, both across its members and with third parties (see EDB 2013b;
Astrov et al. 2012; Isakova and Plekhanov 2012; Ushkalova 2012). The EEU is
being endowed with even more ambitious goals. Outside the EEU, there are areas
where post-Soviet countries could potentially exhibit a high level of integration, at
least due to the institutional and infrastructural legacy from the Soviet Union. For
example, the visa-free regime of cross-border movement could be conducive for
integration of labor markets. Available evidence suggests that cross-border migra-
tion flows intensified in the post-Soviet space over the last decade (Golovnin et al.
2013). Common infrastructure and numerous technological complementarities, as
well as cultural similarities such as the role of Russian as lingua franca, could
strengthen interaction across businesses outside the scope of formal integration
organizations created by governments. Increasing cross-border flows of invest-
ments or informal trade communities is an example, and there is some evidence
showing this (Libman and Kheyfets 2011).
The blank space was filled by the large-scale research project initiated and
implemented by the Eurasian Development Bank (EDB)4. The outcome of the
project was the System of Indicators of Eurasian Integration (SIEI). It was com-
pleted at the end of 2009; and was intended to become an instrument to monitor and
2
The Eurasian Economic Union (EEU) is an international organization, formed by the signing of
an agreement between Russia, Kazakhstan and Belarus in Astana on May, 29, 2014, to enter in
force on January 1, 2015. This is the next level of Eurasian economic integration after the Customs
Union (2010) and Single Economic Space (2012). The full text of the Agreement on the Eurasian
Economic Union is available at the website of the Eurasian Economic Commission.
3
EurAsEC as such also includes Kyrgyzstan and Tajikistan, which, as of now, do not participate in
the Customs Union or Single Economic Space. Armenia and Kyrgyzstan are officially in the
process of joining the Eurasian Economic Union, with Armenia likely to join by January 1, 2015
(information as of July 2014).
4
Eurasian Development Bank (EDB) is an intergovernmental development bank committed to the
advancement of integration in the post-Soviet space. The bank was initially established in 2006 by
Russia and Kazakhstan and currently also includes Armenia, Belarus, Kyrgyz Republic and
Tajikistan as its member states.
5 The EDB System of Indicators of Eurasian Integration: Eurasian. . . 137
5
SIEI mutual investment index is based on the data from the permanent EDB “Monitoring of
Mutual Investments of the CIS countries” (EDB Centre for Integration Studies (2012, 2013a)).
138 E. Vinokurov et al.
6
On the de facto and de jure categories in regionalism studies, see e.g. Higgott (1997).
7
For an overview of general (i.e. qualitative and quantitative) monitoring experiences in various
world regions, see De Lombaerdeet al. (2008b).
5 The EDB System of Indicators of Eurasian Integration: Eurasian. . . 139
Comtrade, Eurasian Economic Commission and the Customs Union. The exception
is the SIEI mutual investment index. It is based on the EDB Centre for Integration
Studies’ ongoing long-term project “Monitoring of Mutual Investments in the CIS”
(EDB 2012, 2013a). This is the largest database in the CIS region of mutual
investments including offshore transactions.
A substantial advantage of the post-Soviet countries is that most of them still
maintain a relatively high quality of public statistics. They are at least superior to
that of most developing countries, though certainly less accurate in many instances.
While statistical systems of post-Soviet countries diverged substantially after the
collapse of the USSR, they are still similar to each other in many aspects. This
facilitates the comparative analysis. The Inter-State Statistical Committee was set
up in February 1992, two months after the establishment of the CIS. Since then it
has accumulated a substantial amount of information that is utilized in the SIEI.
Some data used for the computation of the SIEI should not be considered as entirely
accurate: this applies particularly to cross-border migration, which very often is
informal. The SIEI is characterized by a downward bias in estimating the cross-
border migration. This makes the main findings of the first wave of the SIEI
discussed below even more striking.
In what follows we summarize briefly the key elements for computation of the SIEI
– the System of Indicators of Eurasian Integration. As mentioned, it consists of two
sets of indices which correspond to the main aspects of regional integration (see
Fig. 5.1). The first set measures the integration of markets by looking at two general
indices (trade integration; and mutual investments and labor integration) and three
areas of functional integration (electrical power, education and agriculture). Elec-
trical power is chosen as one of the crucial elements of cross-border infrastructure.
It is where post-Soviet countries often strongly depend on each other, and where
substantial potential for cooperation exists. Education (the cross-border movement
of students) is essential in maintaining social integration of the post-Soviet space.
This includes intensive inter-personal contacts, common language and social net-
works, all of which provide background for economic integration. Agriculture, and
specifically the grain trade, represents a very recent phenomenon in the post-Soviet
space; this is unlike power utilities. Some countries have turned into major grain
exporters in the last decade. The second set of indices includes indicators measuring
the convergence of economic systems. In this case, the subject of evaluation is the
convergence of the countries’ main quantitative development characteristics in four
key areas: macroeconomics (growth dynamics), financial policy, fiscal policy, and
monetary policy. In addition, the first version of the SIEI has been accompanied by
an expert survey evaluating institutional cooperation. This is the performance of
countries in formal integration projects within the post-Soviet space, taking into
account the broad range of goals of the respective structures. The expert survey,
5 The EDB System of Indicators of Eurasian Integration: Eurasian. . . 141
Fig. 5.1 Composition of SIEI (Source: based on SIEI II data (Vinokurov 2014))
however, is not part of the system of indicators and is not considered when
calculating the aggregate measures of integration. Thus, the core of the SIEI
includes ten indicators: four for economic convergence and six for cross-border
transactions.
Each of these sets is computed in the following way. First, the SIEI includes a set
of measures of integration of country pairs (dyadic indicator). It characterizes the
extent to which two particular post-Soviet countries are interconnected by means of
cross-border trade or migration, or as a result of convergence of their economic
indices. For the indicators of cross border transactions the values are computed as
the size of cross-border flows (e.g. trade, investments, migration, grain trade etc.)
relative to the size of both economies. For power utilities and agriculture the
measure of size of the economies used is GDP. For migration and movement of
students it is the size of the population of both countries. For trade we use a
somewhat more complex procedure. The final index included in the SIEI is the
average of two sub-indices: the first measures the size of trade flows within the
country pair relative to the GDP of these countries; the second measures the size of
trade flows within the country pair relative to the overall foreign trade of both
countries. The second indicator is more compatible with the standard analysis of
trade integration (although we acknowledge that it represents a rather simple
approach to its analysis, as discussed in other papers of this volume). The first
indicator is more compatible to other indicators used in the SIEI. However, both
components are strongly correlated, and using them separately from each other does
not change the results. It should also be noted that the use of GDP or of population
as a basis for comparison may affect the outcomes of analysis; this problem has
been discussed in Libman and Vinokurov (2012a). For economic convergence the
indicators are computed as the Euclidian distance between individual countries in a
space defined by the metrics used for a particular convergence indicator
(e.g. different measures of macroeconomic development or monetary policy). The
dyadic indicators are obtained for all pairs of post-Soviet countries, if the data is
available. In the final report of the SIEI the analysis includes both pairwise
142 E. Vinokurov et al.
integration indicators for each year and relative change of pairwise indicators over
the period of observation.
The integration of a country and a group of countries (asymmetric indicator)
characterizes the convergence within the post-Soviet region of any of the twelve
post-Soviet states and any of the five large regions. These regions may be of
particular interest from the point of view of practical integration activity, and
each region includes several countries. The reason for using this indicator is
straightforward. Consider, for instance, a case of integration between a very small
and a very large country. Then very often the large country as economic partner is
of crucial importance for the small country, but the small country is by far less
important for the large country. The indicators we have used so far cannot capture
this asymmetric nature of dependence, because the size of trade flows is computed
relatively to the size of both economies. This is a problem of extreme importance
for the post-Soviet space, where countries are characterized by a very strong
economic asymmetry. Thus, another set of indicators is needed. The asymmetric
indicators are computed as follows: for cross-border transactions we compute the
overall size of trade or factor flow between a country and a group of countries, but
compare it only with the country’s economy size or population. For example, while
the dyadic integration index between Russia and Tajikistan would compare the
cross-border trade between these countries (trade flows in both directions) to the
overall GDP of Russia and Tajikistan, the asymmetric indicator compares the trade
between these two countries only to the GDP of Tajikistan. For convergence
indicators the Euclidian distance is computed between a country and the average
for a group of countries.
At this stage it is necessary to notice that the SIEI uses several “groups of
countries” mentioned above for its analysis. This variation is determined by both
a pragmatic need to account for various possible structures of regional integration in
the post-Soviet space and the necessity to analyze the heterogeneity of post-Soviet
countries. Specifically, there are four regional groups considered by the SIEI:
• CIS-12 (post-Soviet countries excluding the Baltics but including Georgia);
• EurAsEC-5 (the five members of EurAsEC: Russia, Kazakhstan, Kyrgyzstan,
Belarus and Tajikistan);
• SES-3 (the three largest EurAsEC countries – Belarus, Kazakhstan, and Russia –
which were moving rapidly towards closer integration at the time when SIEI was
being set up; they had established a full-scale Customs Union by 2011, Single
Economic Space by 2012 and Eurasian Economic Union by 2015); and
• CA-4 (the four Central Asian states participating in integration projects in the
region: Kazakhstan, Kyrgyzstan, Uzbekistan and Tajikistan. Turkmenistan was
excluded as it virtually did not take part in the CIS and Central Asian integration
projects and did not provide any reasonable statistical information on its cross-
border transactions). Thus, for each country the SIEI computes asymmetric
integration indicators with five groups of countries.
Some aspects of integration cannot be mapped onto each other, and connections
between them are not straightforward; therefore, for the purposes of the SIEI, the
5 The EDB System of Indicators of Eurasian Integration: Eurasian. . . 143
focus is clearly on separate indices rather than their aggregates. However, we have
developed two types of consolidated indices that give a wider picture of regional
integration in the post-Soviet space and include all the nine indices: the consoli-
dated index of a country’s integration with CIS-12, and the consolidated index of a
country’s integration within any of the four regions. The overall structure of the
SIEI is shown in Table 5.1.
The indices of cross-border transactions and economic convergence were cal-
culated for 1999–2008 (where possible as some early data is missing). The evalu-
ation of regional cooperation is provided as at the time of this report. We should
also note that higher values of indicators for cross-border transactions correspond to
higher values of integration; and lower values of convergence indicators correspond
to higher value of convergence (as one could expect given the description of
indicators provided above). For the purpose of aggregated values all indicators
have been re-calculated in a way that a higher value corresponds to a higher level of
integration.
In what follows we summarize briefly the main outcomes of the SIEI. To start with,
integration in the post-Soviet space progresses at an uneven pace, both geograph-
ically and structurally. Before 2008, there was a sharp upturn in legal labor
migration and student exchange, whilst integration in the trade, energy and agri-
culture sectors slowed down and the macroeconomic indices of post-Soviet
144 E. Vinokurov et al.
0.8
0.6
0.4
CIS-12
0.2
EurAsEC-5
0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 SES-3
-0.2
CA-4
-0.4
-0.6
-0.8
Fig. 5.2 Consolidated indices of integration of four groups of countries within the post-Soviet
space (2002–2012) (Source: SIEI data)
0.2 2002
0 2008
Tajikistan
Belarus
Uzbekistan
Georgia
Armenia
Turkmenistan
Kazakhstan
Azerbaijan
Moldova
Kyrgyzstan
Russia
Ukraine
-0.2 2012
-0.4
-0.6
-0.8
-1
Ukraine and Moldova continue being rated, and in 2012 they became more
integrated with the CIS region. The key spheres are labor migration (for both), and
trade and investments (for Ukraine). Russia is Ukraine’s main trading partner. The
2014 conflict will result in the dynamics of Ukraine’s economic ties with the CIS
region being negative. Ukraine, Azerbaijan, Moldova, Georgia and Uzbekistan are
participants of the GUUAM organization, which has become informal. They have
always taken a restrained stance towards integration projects within the CIS, and
have consented to very limited or nominal participation. For Ukraine, the limit of its
participation has been the free trade zone.
Kazakhstan, Belarus and Russia, which are the “integration core” of the CU and
the Eurasian Economic Union, complete the 2012 rating. These are large economies
with a comparatively diverse structure of foreign trade, in which economic ties with
the post-Soviet space tend to become less important. These are fairly rich countries
– Kazakhstan and Russia are exporters of fossil fuel. It should not be a surprise that
Russia occupies the last place in this rating. It is the largest post-Soviet economy, it
stands on a par with the rest of the post-Soviet space in terms of population size, and
it accounts for about 75% of GDP.
If we look at individual areas of integration and the integration performance of
various countries, it is not possible to identify any unquestionable leaders in all
aspects of integration among country pairs or groups. Moreover, the structure of
mutual links varies greatly across different CIS markets. To some extent, this is
illustrative of the diversity of interests and resources involved in integration in the
CIS. Belarus, Kyrgyzstan, Tajikistan and Turkmenistan became leaders in various
aspects of integration with CIS-12 region. This shows the large interest that Central
Asian countries have in integration processes on the territory of the CIS. However,
it is partly explained by the relatively small GDP volume and population size of
these countries. Georgia, Kyrgyzstan, Tajikistan and Ukraine are leaders in terms of
the absolute increment of integration indices with CIS-12 in 2012 rating. The
148 E. Vinokurov et al.
countries showing the biggest increase in integration levels in 2012 are Armenia,
Kyrgyzstan and Ukraine (see Table 5.2).
The main volumes of trade flows in the post-Soviet space are focused between
the major countries: Russia, Belarus and Ukraine. The maximum trade integration
level in 2012 was observed in pairs Ukraine-Belarus, Russia-Belarus and
Ukraine-Russia. Belarus is the leader of trade integration with CIS-12, EurAsEC-
5 and SES-3 regions. Kyrgyzstan is ranked second in terms of these indicators.
Kyrgyzstan also sees the maximum level of integration in mutual trade with CA-4
region. The highest increment of trade integration indices for 2009–2012 is
observed in the pair Ukraine-Belarus, and the biggest reduction is in the pair
Ukraine-Turkmenistan. Belarus had the largest increase of integration with
CIS-12, EurAsEC-5 and SES-3, and Moldova had the largest reduction. For the
remaining countries values of integration indicators with these regions have not
changed significantly. The leader in terms of increment of integration with CA-4 is
Kyrgyzstan, and the leader in terms of integration reduction is Moldova.
Tajikistan is leading in labor migration indicator concerning CIS-12 and also
with EurAsEC-5 and SES-3. This can be attributed to the large outflow of labor
resources to Russia in relation to the country’s own population. Tajikistan is
followed by Uzbekistan and Moldova. Kyrgyzstan has the largest level of integra-
tion with CA-4. The pair Russia-Uzbekistan account for the maximum level of
integration in labor migration in 2011, the second place is taken by the pair Russia-
Tajikistan, the third by Russia-Ukraine. Notably, the lowest labor migration index
belongs to Belarus, which otherwise demonstrates excellent integration perfor-
mance in the area of cross-border trade. It is important to notice that for
2009–2011 the formal index of labor migration with CIS-12, EurAsEC-5 and
SES-3 regions was reduced for all countries by over 50% on average. This was
mainly by the reduction of the number of legal workers arriving in Russia registered
by the Federal Migration Service. However, it does not mean a reduction of the
5 The EDB System of Indicators of Eurasian Integration: Eurasian. . . 149
overall labor migration to the country as at the same time, according to expert
appraisals, the volume of illegal migration into Russia increased.
The dynamics of trade in electrical power in the post-Soviet space lags far
behind the growth of CIS economies. In most country pairs, this index shrank
during 2002–2008. The only exception was Ukraine whose integration with
EurAsEC-5 and EurAsEC-3 progressed slightly, whereas its integration with
CIS-12 slowed. This process is also driven by trade in electrical power with
Russia. Our analysis shows that Russia is the main electrical power supplier in
absolute terms, and Belarus is the main recipient. The pair Belarus-Ukraine has the
biggest integration index in this area due to Ukraine selling a large power volume to
Belarus. They are followed by the pairs Armenia-Georgia and Kazakhstan-
Kyrgyzstan. Trade between Kazakhstan and Russia is ranked third in absolute
terms, but it is small compared with the GDP of these countries. 2002–2012 is
characterized by a significant reduction of integration indices of electrical power
trade for all pairs of countries and regions with subsequent stabilization in
2009–2012. The reason for this is both a reduction of trade between Central
Asian countries and the outrunning growth of economies of the countries. The
reform of the electricity sector in Russia did not result in a qualitative growth of
cross-border power flows.
The leader in agriculture integration in the post-Soviet space is Kazakhstan. This
is based on data on cross-border trade in cereals. Kazakhstan is present in all three
leading country pairs: Kazakhstan-Azerbaijan, Kazakhstan-Tajikistan and
Kazakhstan-Kyrgyzstan. In this case, integration of neighboring Central Asian
and Caspian states is presumably based on the export of cereals from Kazakhstan.
Kyrgyzstan and Tajikistan are the leaders in trade integration with CIS-12,
EurAsEC-5 and SES-3, which appears to be caused by the large volume of cereals
export in relation to its economic size. The same is the case with Georgia. Russia
has the lowest levels of integration with CIS-12 and other groups; this is due to its
enormous economy and powerful agriculture sector.
With respect to investments, we can conclude that in 2012 the main “donor” of
investments was Russia, and the main recipient was Ukraine. The pair Azerbaijan-
Georgia is characterized by the largest degree of integration due to the low GDP of
both countries. A high level of investment integration may be highlighted in the
pairs Ukraine-Russia, Russia-Kazakhstan and Russia-Belarus with Russia investing
substantial funds in the economy of partner-countries. Leaders of integration with
CIS-12 region are Armenia, Kyrgyzstan and Georgia. The lowest degree of inte-
gration with CIS, EurAsEC-5 and SES-3 is observed in Turkmenistan, which is
almost not involved in the processes of inter-country capital movement. Russia, due
to its large GDP, has low values of integration indicators with these three regions.
The largest integration degree with CA-4 region is seen in Kyrgyzstan and Georgia.
Azerbaijan and Moldova have no investment cooperation with groups of CA-4
countries.
In the area of academic mobility the main recipient of foreign students in CIS-12
countries is Russia, with most students coming from the main “donors” of the
region: Kazakhstan and Belarus. The third most important “donor” of students is
150 E. Vinokurov et al.
Turkmenistan, with over 20,000 students going to study in Russia, Belarus and
Ukraine according to 2011 data. The academic mobility index is the maximum for
the pair Turkmenistan-Belarus due to the small population of both countries, and
5000 Turkmen students which is a large number in relative terms. Turkmenistan
also has the maximum integration index in education with CIS-12, EurAsEC-5 and
SES-3 followed by Belarus and Kazakhstan. Russia has the lowest integration index
with the regions because it has a larger population. In general, during 2000–2011 all
pair and intra-regional (except for CA-4) integration indices rose progressively. As
a result, the growth of interregional indices was 100 to 140%. There has been an
integration increase in the area of academic mobility in the post-Soviet space.
Unlike the integration of markets, the convergence of post-Soviet economies varies
greatly depending on particular country pairs or country-region pairs. As men-
tioned, convergence is generally not driven by any geographic factors. The key role
belongs to reform strategies selected by particular countries, and macroeconomic
regulation practices that make them become closer. On the whole, we can conclude
that the macroeconomic indices of post-Soviet states were diverging over the last
decade, whereas their monetary policies converged. The main results of our anal-
ysis are summarized in Table 5.3.
To assess economic convergence of individual countries with groups of coun-
tries, we also computed a set of weighted indicators. In this case the SIEI compares
the economic indicators of a particular country with the weighted average of a
country group, and not with the average of a country group, where the weight is
determined by the size of the countries. The logic of this approach is straightfor-
ward. Assume, for example, that a group consists of a set of countries, with some
being relatively large and some relatively small. Then convergence with the largest
countries of this group should matter much more than convergence with smaller
countries in terms of potential for economic integration. In fact, the results without
weighting could be distorted by some very small outliers.
By 2012 the leadership in convergence in CIS-12 belonged to Belarus (fiscal
policy), Armenia (financial policy), Georgia (macroeconomics), and Uzbekistan
(monetary policy). The most integrated pair of countries in 2012 was the pair
Kyrgyzstan-Moldova, and Georgia was the leader of convergence with the CIS
group. Recently the pair Armenia-Uzbekistan has converged, and the leader of
convergence with three groups of countries was Azerbaijan. In terms of macroeco-
nomic convergence, major changes in the level of integration of regions were not
observed for 2009–2012, despite the global economic crisis. The greatest distances
from CIS-12 (in 2009–2012) are demonstrated by Turkmenistan (macroeconom-
ics), Belarus (finance and monetary policies – due to inflation and the drop in the
rate of Belarusian ruble in 2011–2012), and Russia and Moldova (fiscal policy).
Both approaches (weighted and non-weighted indices) have their merits and
demerits. Therefore, economic convergence should be assessed by both methods,
and the results should be treated as complementary.
Table 5.3 The dynamics of convergence of post-Soviet economies (data for non-weighted indices)
Pair-leader Pair-leader Total dynamics of the
(minimum index, (index Leader of convergence with the Leader of integration with convergence index of the CIS
Indicator 2012) reduction) CIS (minimum index, 2012) the CIS (index reduction) region
Macroeconomics Kyrgyzstan – Armenia – Georgia Turkmenistan "
Moldova Turkmenistan
Monetary policy Azerbaijan – Azerbaijan – Uzbekistan Moldova #
Ukraine Moldova
Financial policy Azerbaijan – Armenia – Armenia Armenia "
Armenia Kazakhstan
Fiscal policy Armenia – Armenia – Belarus Russia #
Kazakhstan Russia
Note: increasing the distance (") means lowering the convergence level
Source: SIEI data
5 The EDB System of Indicators of Eurasian Integration: Eurasian. . .
151
152 E. Vinokurov et al.
References
Libman, A., & Vinokurov, E. (2012a). Regional integration and economic convergence in the
Post-Soviet space: Experience of a decade of growth. Journal of Common Market Studies, 50
(1), 112–128.
Libman, A., & Vinokurov, E. (2012b). Holding-together regionalism: Twenty years of the Post-
Soviet Integration. Basingtoke: Palgrave MacMillan.
Malfliet, K., Verpoest, L., & Vinokurov, E. (2007). The CIS, the EU and Russia: Challenges of
integration. Basingstoke: Palgrave MacMillan.
Osterkamp, R. (2008). Measuring regional trade integration in Southern Africa. In A. Boesl,
W. Breytenbach, T. Hartzenberg, C. McCarthy, & K. Schade (Eds.), Monitoring regional
integration in Southern Africa Yearbook (Vol. 8, pp. 342–355). Stellenbosch: TRALAC.
Shin, K., & Wang, Y. (2003). Trade integration and business cycle synchronization in East Asia.
Asian Economic Papers, 2(3), 1–20.
Ushkalova, D. (2012). Formirovanie Tamozhennogo Soyuza i Edinogo Ekonomicheskogo
Prostranstva Rossii, Belarusi i Kazakhstana: Problemy i Predvaritel’nye Itogi. Moscow:
IERAS.
Vinokurov, E. (Ed.). (2010). The EDB system of indicators of Eurasian Integration. Almaty:
Eurasian Development Bank.
Vinokurov, E. (Ed.). (2014). The EDB system of indicators of Eurasian Integration II (Centre for
Integration Studies, Report no. 22). Saint-Petersburg: Eurasian Development Bank. The
20-page English summary of the report is available at: http://www.eabr.org/general//upload/
CII%20-%20izdania/SIEI-2014/EDB%20Centre_Report%2022_SIEI%20II_Analytical%
20resume_Eng_1.pdf
Vinokurov, E., & Libman, A. (2011). Regional integration trends in the Post-Soviet Space: Results
of quantitative analysis. Problems of Economic Transition, 53(12), 43–58.
Part II
The Americas and the Caribbean
Chapter 6
Measuring Integration Achievement
in the Americas
Gaspare M. Genna
6.1 Introduction
This chapter will describe one method to measure the complex set of regional
integration achievements from their early beginnings. The Integration Achievement
Score (IAS) was first developed by Hufbauer and Schott (1994), but was later
refined and expanded (temporally and geographically) by Genna (2002). The
measure taps the complexity of integration by disaggregating the phenomenon
into its components, assigning values to these components, and then calculating a
final index. By systematically measuring integration, researchers can perform side-
by-side comparisons, especially large N econometric analysis. I will illustrate the
application by assessing the integration achievement in the Americas. Like any
method to measure complex phenomena, the method does have some limitations
which need to be addressed. The paramount of these is the assessment of treaty
implementation and the consistency of the implementation among the member
states of the RIO. Lastly, another issue I will address is the analytical application
of the IAS when one is faced with the problem of multiple RIO memberships.
1
Systematic measures that can travel across all regions should not be viewed as a way to
normatively judge (praise or criticize) RIOs. Scientific analysis is not interested in promoting
integration nor is it interested in suppressing it. In other words, assessing regional integration is not
an exercise akin to judging a beauty contest. The aim is to develop techniques to compare apples
with apples so that we have generalizable findings.
6 Measuring Integration Achievement in the Americas 161
fulfill the specific RIO obligations. I verify the implementation of RIO obligations
among member states using information contained in various years of the Europa
World Year Book (EWYB) and cross-referenced with other specialized sources.
The first category is trade in goods and services. This category is the foundation
of regional integration and was for a time the only operational definition (see
Balassa 1961). A RIO could, theoretically, have no provisions for trade so a zero
value is possible, although unlikely. The next value up is allocated if countries
develop a preferential trade agreement. Such an agreement allows for reduced
duties or regulations on trade, but does not eliminate them. A partial free trade
area is in place when some categories of goods and/or services are allowed to flow
without tariffs. A full free trade area is an agreement that allows for free trade on all
categories. A customs union is in place when countries have a uniformed tariff for
goods coming into any RIO member from a non-member. It is possible for customs
union to be in place while having a partial free trade area. In this situation, the RIO
is assigned a 3.5 in this category. The highest value is given when RIO members
remove all barriers (tariff and non-tariff) between each other.
The second category is free movement of capital. Liberalization in this category
refers to direct investment in partner countries with the associated ability to
162 G.M. Genna
withdraw investment. It is important to note that values in this category are due to
formal agreements among the RIO members and due to a country’s unilateral
liberalization of investment flows. At the first level, foreign direct investment
(FDI) allowed in limited form. FDI at this level can be restricted by being tied to
conditions of percent ownership by a citizen, partnerships with domestic firms, or
other methods that would limit the independent exercise of the investment. The next
level occurs when countries agree on a method that will allow for complete
withdrawal of investment without penalties. At the next level foreign investment
is allowed, as well as capital withdrawal, but not in areas involved in national
government procurement. Often these areas are restricted due to national security
concerns, but there are some minor exceptions to this. This next level allows for full
capital mobility except for large scale mergers and acquisitions. This subcategory
does not apply if the national government regulation requires approval for smaller
mergers and acquisitions if the investor is a citizen of a RIO member country but
not for its own citizens. Restrictions on government procurement areas must be
removed for a value of 4 to be assigned. The highest value is given when there is
full capital mobility (both in and out of the country) without restriction.
Following liberalization of capital is labor mobility. The unrestricted ability of
labor to seek higher wages or other employment opportunities in the partner
countries signifies that a single labor market is present. This single labor market
is a strong indicator of an integrated economy because an important factor of
production can be efficiently distributed. Right of movement refers to the automatic
permission of entry for employment given by one member state to a citizen of
another member state. At the first level, this right is granted for select professions,
but at the subsequent levels, it is given to all categories of workers. Having the right
of movement can be restricted by other means. Such restrictions include the
transferability of professional qualifications. Agreements that standardize or at
least recognize university degrees or other professional certifications warrant a
value of three. Individuals could also be constrained from moving if their pensions
or other retirement devices are restricted in some manner. This can include com-
plete or partial forfeiture. The highest value is assigned when countries adopt
agreements for the freedom of movement without the restrictions mentioned.
The next category is that of supranational institutions. The member states’
collective deliberations are at the core of all RIO decision making. However,
many RIOs also have regional institutions that participate in decision making at
varying degrees. At the highest level, supranational institutions are central decision
making actors in various areas. In order to score the maximum value in this
category (that of 5), the supranational institutions need not command all authority
within the RIO. For example in the case of a federal arrangement, the center does
not possess all decision making power. The individual sub-national units do hold
power and sovereignty in many areas. The values in this category assess the degree
of importance in the decision making process. At the first level, a RIO establishes
nominal institutions. These offices, often termed secretariats have no mandate other
than to prepare and perhaps host meetings. At the next level, these secretariats are
mandated to gather information and provide advice to the member states for their
6 Measuring Integration Achievement in the Americas 163
collective decision making. At the next level, RIO institutions begin to have a more
direct inclusion in the decision making process by having the ability to amend
proposals. However, like reports and advice, amendments can be ignored because
the member states can veto the amendment by overriding them with their own vote.
The ability to veto proposals now aligns supranational institutions with their
intergovernmental counterpart. The highest value is assigned to those supranational
institutions that operate as primary decision node. This means that member states
legally cannot, individually or collectively, reverse or block a decision made by a
RIO institution.
The last two categories involve the monetary and fiscal policies of the RIOs. The
category of monetary coordination refers to the progress in the establishment of
common policies that adjust exchange rates among the member countries. At the
first level, member states simply consult with each other regarding policy. Although
consultation is mandatory at this level, commitments are not required. However, the
transparency involved in these consultations can assist in informal coordination of
individual policies. At the next level, member states commit to maintain parity
among their individual currencies. How they will maintain parity is up to the
individual member states. However, if they decide to coordinate interventions in
their currencies in order to maintain parity, then the RIO moves up to the next level
of this category. Next, the RIO members can become more committed to a single
monetary policy by establishing a regional central bank. This bank would oversee
all coordination by providing strategic planning, but also operates as an indepen-
dent data gatherer and monitor of member states’ activities. The highest value is
assigned to RIOs that have a single currency used for all transactions under the
governance of a regional central bank.
The final category is fiscal coordination, which refers to the establishment of
spending criteria for the member states. Integration in this area helps to maintain
stability that can be harmed should some governments develop excessive govern-
ment debt, promote subsides that could harm trade patterns, and/or develop uneven
tax codes. At a minimum, governments can officially consult each other regarding
fiscal policy in order to promote transparency and dialog. Next would be the
establishment of commitments regarding deficit spending and taxation. Commit-
ments would be strengthened at the next level when credible sanctions would be the
norm to address member states that break commitments. The addition of a uniform
tax code would provide a balanced environment for investment. The highest value
is assigned to those RIOs that have a single budget that would finance programs and
other initiatives outside RIO institutional maintenance.
In this section, I briefly outline the integration measurement for five RIOs in the
Americas: Andean Community (CAN), Central American Common Market (also
known as the Central American Integration System; SICA), The Caribbean
164 G.M. Genna
CAN was established in 1969 by Bolivia, Chile, Colombia, Ecuador, and Peru
through the Cartagena Agreement. Venezuela joined in 1973. Under the original
Cartagena Agreement, these countries were referred to as either the Andean Group
or the Andean Pact. However, since the installation of the 1996 Reform Protocol of
the Cartagena Agreement, the group adopted their current formal name and also
goes by title of the Andean Community.2 The original objectives of CAN were to
create a common market with a harmonization of social and economic policies
(Ocampo and Esguerra 1994). Chile withdrew in 1976 because of the wide diver-
gence of its domestic policies vis-a-vis the other members. Its withdrawal was a
particularly strong blow to integration given the strong complementary nature of
the Chilean economy with Colombia and Venezuela (Ocampo and Esguerra 1994).
Intra-regional political issues also hurt the Andean Group. A series of military
coups in Bolivia between 1978 and 1980 led to a test of members’ resolve to keep
the sub-region democratic. Bolivia’s military government was not recognized by
the other members and it threatened to withdraw in 1980. Similarly, the group did
not recognize President Alberto Fujimori’s suspension of the constitution and
subsequent autocratic rule. This led to the brief suspension of Peru’s membership
in 1992. The suspension was also due to Peru’s incompatible preferences regarding
the new negotiations of the common external tariff (CET). However, it is difficult to
disentangle the two seemingly interrelated reasons. Peru was readmitted in 1994,
but it did not resume full participation until 1997 (Commission Decision 414). One
final political issue was Ecuador’s border dispute with Peru in 1981 that erupted in
to a full border war in 1995 and was not resolved until 1998. Venezuela left in 2006
in part due to ideological disputes with other member governments and in part due
to Venezuela’s desire to join MERCOSUR.
Toward the end of the lost decade, the Andean Group began its efforts to revive
regional integration. One of the first actions was the Quito Modifying Protocol
(1987), which recommitted the members to a CET and began the gradual effort to
rescind Decision 24 (Bulmer-Thomas 1994). In 1988, the members established the
Andean Inter-municipal Bank in order to finance public works (EWYB 2001).
These revival efforts culminated in the 1996 Trujillo and 1997 Sucre reform
2
The Spanish name is Comunidad Andina or CAN.
6 Measuring Integration Achievement in the Americas 165
protocols of the Cartagena Agreement which formally established the new Andean
Community.
The new incarnation of CAN includes a set of ambitious economic and political
objects. This is to include the realization of a CET, a common market (including
labor mobility), a common agricultural policy, a common foreign policy, macro-
economic coordination, and strengthening of regional institutions. Of these objects,
only the CET, the common market, and the strengthening of regional institutions
has seen concrete action; although some steps have been employed for the others.
Commission Decision 535 established a CET on approximately 62% of all imports
into the region.
Decision 563 officially codified the text of the Andean Sub-regional Integration
Agreement (the new Cartagena Agreement of 1997). In doing so it installed
Chapter II of the Agreement which establishes the Andean Integration System
(AIS). The AIS is simply the official institutions that operate as decision-making,
advisory, and administrative bodies. The Andean Presidential Council represents
the highest decision making body of the AIS. It is made up of all the presidents of
the member states and convenes annually. The council has a rotating chair who
holds that position for 1 year. The Commission of the Andean Community consists
of representatives from each member and is the main policy-making body. This
power is shared with the Andean Council of Foreign Ministers and begins as
initiatives from the Presidential Council, the member countries, or the General
Secretariat of the Andean Community. These policies are titled “Decisions,” such
as the one found above. It also has the responsibility to implement and evaluate
policy. The Council of Foreign Ministers is the grouping of national foreign
ministers who meet at least once a year for the purpose to develop common external
policy and to coordinate the process of integration. It meets prior to the Presidential
Council in order to also prepare for common positions and declarations that come
out of that summit. The Council of Ministers also has the power to elect, remove,
and evaluate the Secretary-General of the General Secretariat. The General Secre-
tariat implements all the decisions of the decision making bodies listed above
through functional departments. The Secretary-General is elected for a 5-year
term. The Andean Parliament is a weak deliberative body of the AIS. Currently
the members of the Parliament are representatives from the national congresses
(generally members of committees associated with ANCOM). The representatives
of Colombia, Ecuador, and Peru are now being elected directly and it is expected
that all members will eventually have direct elections in place. It deliberates over
decisions and adopts legislation than puts those decisions in forces. The final
institution is the Andean Community Court of Justice. The Court began operation
in 1984, comprising five judges one from each of the member countries for
renewable 6 year terms. The Presidency of court operates on a one year rotating
basis among the five. It has jurisdiction over legal ruling concerning CAN law and
also operates as an arbiter over disputes. The Courts’ powers have expanded in the
new modification protocol to the founding treaty. This includes new jurisdiction in
labor disputes, and appeals of inaction.
166 G.M. Genna
The CAN IAS for 2009 is 2.17. Given the establishment of a full free trade area
and a customs union, the trade in goods and services score is 4. Individuals and
firms have full access for foreign investment and withdrawal except for areas of
national government procurement. This gives it a capital mobility score of 3. The
labor mobility score is 1 since right of movement is reserved for select professions.
Supranational institutions are important since they have the ability to amend pro-
posals (a score of 3). There is some commitment to maintain currency values among
the members (a score of 2), but no fiscal policy coordination is in place.
Current difficulties may lead to a lowering of CAN’s IAS value. When Colombia
and Peru’s free trade agreement with the Unites States (US) (2006) goes into full
force, it will put into question the CAN customs union. In addition, Colombia and
Peru’s focus is increasingly drawn to the newly formed Pacific Alliance, while
Bolivia is focusing in the other direction with its MERCOSUR application. Lastly,
Ecuador’s economic foreign policy is becoming more in line with Venezuela.
Unless the traditional protectionist bent of the customs union changes, it will be
difficult for some of the member states, especially Peru, to continue to abide by
CAN’s requirements. This would mean a lack of treaty implementation and the
subsequent lowering of CAN’s institutionalized integration.
Central America was at one time united as a federation from independence through
1838 (Bulmer-Thomas 1994). So when we speak of Central American integration,
it would be more appropriate to speak of its re-integration. The effort of
re-integration began in 1951 with the establishment of the Organization of Central
American States (OCAS). On December 13, 1960, with the signing of the General
Treaty of Central American Economic Integration, the member countries
established the Central American Common Market (CACM),3 which was ratified
by all the members by the end of 1963. These agreements represent one of the
earliest cases of regional integration, one nearly as old as the European project.
However, the level of integration attained by the countries of Belize, Costa Rica, El
Salvador, Guatemala, Honduras, Nicaragua and Panama is small. In fact the CACM
functionally ceased to exist by 1969 as a result of conflict among its members.
Successful efforts to restart the project did not begin until the mid-eighties. In 1986,
the new CACM tariff and customs agreement went into effect. The agreement
developed the limited CET, eliminated intra-regional non-tariff barriers, and pro-
moted agricultural trade liberalization. Newer initiatives produced the Protocol of
Tegucigalpa to the agreement establishing the CACM and in doing so inaugurated
3
In Spanish, it is referred to as the Mercado Común Centroamericano.
6 Measuring Integration Achievement in the Americas 167
the Central American Integration System or SICA, using the Spanish acronym.4
SICA formally went into effect in 1993.
The CACM is now a subcomponent of the larger SICA project, although a very
significant portion. The role of SICA is to coordinate the activities of the four
subsystems of integration: political, economic (CACM), social, and environmental.
To these ends, SICA is a network of supranational institutions and intergovern-
mental arrangements that facilitates decision making. The top of the decision
making hierarchy is the summit meetings of the presidents of the member states.
Decisions, accords, protocols, treaties, and initiatives are finalized during these
meetings by consensus. The chair of these meetings is appointed on a rotating basis,
every 6 months. Ideas, however, are first introduced into the system at lower levels
of decision making. One such intergovernmental arrangement that first discusses
ideas is the Council of Ministers. The Council is made-up of the various foreign
affairs ministers of the member states. Other sectoral and intersectoral ministers
gather in separate meetings. Decisions at the Council meetings are made by
consensus, although the majority vote is permissible under certain circumstances.
The Consultative Committee includes representatives from various social organi-
zations, such as business organizations, trade unions, and academic institutions.
The Committee provides input into the process by assisting the Secretary-General
of the SICA General Secretariat. They do not hold veto power in the process. The
General Secretariat was established as a true SICA supranational institution with
the Protocol of Tegucigalpa. Through its divisions, which are headed by the
Secretary-General, it forms the bureaucracy of SICA and coordinates the overall
integration process. The Secretary-General is appointed during the presidential
summits. Other technical and functional secretariats also exist under the General
Secretariat, but are more autonomous than the individual divisions. One specialized
secretariat of importance is the one that oversees the implementation and offers
evaluation of the CACM (namely the SICA). In 1989, a parliament was established
within the framework of the CACM. Each member country receives 22 representa-
tives through direct elections. The Court of Justice includes one magistrate from
each member and is the final authority over disputes related to the integration
system.
SICA’s IAS value for 2009 is 1.17. Given the uneven application and limited
tariff agreements, the trade in goods and services score is 2. FDI is only allowed in
limited form among individuals and firms from the member countries (capital
mobility score is 1). Labor mobility is also limited to a select set of professions
(score of 1). Supranational institutions do play some role, given their information
gathering and advisory roles (score of 2). Monetary policy coordination is limited to
consultation, although there has been some talk of establishing a common currency
(score of 1). There is currently no fiscal policy coordination.
4
SICA stands for the Sistema de la Integraci
on Centroamericana.
168 G.M. Genna
CARICOM actually represents two integration efforts, the Community and the
Common Market, with a great deal of membership overlap. Like the Central
American case, the origins of CARICOM begin with the collapse of their federa-
tion. The West Indies Federation (WIF) was an initiative of the British government
and began in 1958. The WIF fell apart in 1962 as first Jamaica and Trinidad and
then the other members declared independence from Britain and did not wish to
maintain membership in the WIF. Antigua, Barbados, and Guyana formed the
Caribbean Free Trade Association (CARIFTA) on December 1, 1965, but it did
not go into effect until May 1, 1968 (Boxill 1997). The CARIFTA was delayed so as
to give opportunities to the other states of the Caribbean basin to join. While only
the country of Trinidad and Tobago was among the founding members (along with
the three just mentioned), Dominica, Grenada, St. Kitts-Nevis-Anguilla, Saint
Lucia and St. Vincent became members in July 1968; Jamaica and Montserrat in
August 1968; and British Honduras (Belize) in May 1971.
CARICOM was established in 1973 in the Treaty of Chaguaramas by Barbados,
Jamaica, Guyana and Trinidad & Tobago. The Georgetown Accord introduced
eight others, all British territories, into CARICOM: Antigua, British Honduras
(Belize), Dominica, Grenada, Saint Lucia, Montserrat, St. Kitts/Nevis/Anguilla
and St. Vincent. All eight territories signed the Accord to become full members
by May 1, 1974. The Bahamas became a member of the Community in 1983 but
never joined the Common Market; Suriname became a member in 1995; Haiti is the
latest member when it joined in 2002.
There are two areas of interest for CARICOM: political and functional aims and
economic integration. The former aims are guided by Community action. This
focuses on coordination of foreign policies and functional cooperation that includes
services that benefit the people, development of greater understanding among the
peoples, and the advancement of social, cultural, and technological development.
Of these, coordination of foreign policies among members is of great importance
because it unites small and mini-states in a coalition vis-a-vis third parties. The idea
is to unite in a regional organization so that members are able to negotiate with one
voice, especially with regard to free trade agreements. In 1997 the members
formalized joint action by establishing the Regional Negotiating Machinery
(RNM) body (EWYB 2001). The RNM consists of the Chief Negotiator, Chief
Coordinator, technical advisory groups, and negotiating working groups. The
negotiating working groups have areas of expertise that comprise of issues dealing
with the members’ association in the Lomé Convention with the European Union
and other EU related negotiations, Western Hemispheric related issues like the
impact of NAFTA, General Agreement on Tariffs and Trade (GATT) rounds and
the World Trade Organization (WTO), and other non-economic related issues in the
Americas.
The second area of interest, regional integration, which is guided by Common
Market action, focuses on trade relations, balanced economic development,
6 Measuring Integration Achievement in the Americas 169
5
Given that many of the members are also members of the British Commonwealth or are still
territories of Britain, their head of state is the monarch. Therefore, heads of government meet
instead of heads of state.
170 G.M. Genna
MERCOSUR6 is a relatively new integration project, but one with old roots and
large implications for the South American integration. The core of MERCOSUR
has been the relationship of Argentina and Brazil which comprise the largest and
among the most developed economies in South America. Argentina and Brazil first
began the efforts towards economic cooperation under the Pinedo Plan (named after
Federico Pinedo, Argentinean Minister of Finance) in 1940, which was also wel-
comed by Osvaldo Aranha, the Brazilian Finance Minister (Bulmer-Thomas 1994).
Early efforts did not produce success due to the instability of democracy and poor
economic growth that produced rival military regimes during the 1970s.
The initial negotiations of MERCOSUR began with Brazil and Argentina. In the
midst of their consolidation efforts, Presidents Alfonsı́n of Argentina and Sarney of
Brazil signed a cooperation agreement in 1986 (Fritsch and Tombini 1994). This
cooperation agreement produced 20 protocols on such things as the elimination of
trade barriers on a common list of products, quotas on wheat exports from Argen-
tina, and energy supply cooperation (Bulmer-Thomas 1994). In November 1988
they singed the Argentina-Brazil Treaty of Integration, Development, and Cooper-
ation, which set a timetable for the elimination of all trade barriers. In July 1990,
both signed the Buenos Aires Act that puts together a commitment to a CET and an
elimination of all inter-regional trade barriers by 1995. Other than the implied
objectives of these actions, the agreements were also part of the respective admin-
istrations’ plans to modernize their economies and curb inflation. The impact of
these agreements sparked fear of trade and investment away from Paraguay and
Uruguay leading to the signing of the Treaty of Asunción in March 1991 and
formally establishing MERCOSUR among the four countries. The objectives of
the founding treaty are the liberalization of intraregional trade, a common external
tariff, harmonization of laws and regulations concerning rules of origin, and the
mutual consultation on macroeconomic policies (Pereira 1999). The original time-
table for the CET was not changed, which was met, but with exceptions: only 80%
of extra-regional imports are covered (Bulmer-Thomas 1994). The goal is to
eliminate all exceptions by 2006 (EWYB 2001). The free trade area also went
into effect in 1995, but only on 85% of intra-regional trade (EWYB 2001). The
liberalization of all intra-regional trade is effectively an open ended goal.
Other South American countries have associate membership in MERCOSUR and
others are in various stages in seeking full membership. Chile, Colombia, Ecuador,
and Peru are currently associate members. Guyana and Suriname are also on course
to become associate members as they await treaty ratification. Associate member-
ship means that states do not have full voting rights and are not members of the
customs union. They do, however, have some access to the full members’ markets
and therefore form a partial free trade area. Recognizing Chile’s geographic
6
MERCOSUR is the Spanish acronym for Mercado Común del Sur. MERCOSUR is also known in
Brazil as MERCOSUL which is the Portuguese acronym for Mercado Comum do Sul.
172 G.M. Genna
advantage7 and dynamic economy, the full members of MERCOSUR toyed with the
idea of extending full voting rights to it. However the idea died when Chile first
negotiated and then signed a free trade agreement with the US. The full members
perceived this action as having “broken ranks”. Bolivia, which has been an associate
member since 1996, is seeking full membership. In 2012, it achieved the status of
acceding member and will become a full member after it implements the common
external tariff. There has been some talk of Ecuador joining, but no definitive action
has been taken. The latest country to fully join is Venezuela, which did so in 2012.
The Brazilian move to devalue its currency in 1999 without consulting the other
members created major economic problems in the region, particularly with Argen-
tina. Coordination was not formally in place and there was no obligation to inform
any partner, let alone consult with them, regarding macroeconomic policy. The
result was retaliatory action by Argentina on Brazilian products, which in turn
prompted Brazilian action on Argentinean products. When the dust settled, the two
agreed informally not to continue their unilateral actions and Brazil agreed to
Argentinean tariffs in order to help its crippled economy.
The formal result of this episode was the agreement in Florianopolis (2000)
which made the first steps towards macroeconomic policy coordination. Targets for
inflation were chosen: no more than 5% in 2002–2005, 4% in 2006, and 3% from
2007 (EWYB 2001). The agreement also included a public debt reduction to 40% of
GDP by 2010 and fiscal deficits to no more than 3% of GDP in 2002 (EWYB 2001).
The data from members would be reported to the Macroeconomic Monitoring
Group, who would also establish harmonized methods of calculating the data. In
addition, the idea of developing a common currency was entertained. The Monetary
Institute for Mercosur was created to look into its feasibility.
MERCOSUR is unique among the integration projects in the Americas in that it has
the power to sign international agreements on behalf of its members. This status as a
legal entity was given to it by the members as a result of the Ouro Preto Protocol (1994)
to the founding treaty. As a result, MERCOSUR has tried to develop an international
presence, with some success. The successes involve the ability to negotiate a free trade
agreement with CAN (2005), Israel (2007), Egypt (2010), and the Palestinian Authority
(2011). It is also undergoing protracted talks with the EU that began in 2000.
The Ouro Preto Protocol also established the MERCOSUL institutions. The
institution with the final say in decision making is the Common Market Council.
At the top of the Council are the members’ presidents and below them are the
ministers of foreign affairs and economy. The Council’s responsible for the polit-
ical direction of the integration process. The meetings of the Council follow the
European model. The presidents meet twice a year (June and December) to discuss
and sign accords based on work of two other MERCOSUR institutions, namely the
Common Market Group and the Trade Commission. Ministerial working groups
meet 2 days prior to the Council summit in order to negotiate agenda items and the
7
Chile borders the Pacific Ocean and therefore gives MERCOSUR a key link to the markets of the
Far East.
6 Measuring Integration Achievement in the Americas 173
final communiqué. The summit ends with the submission of the final communiqué,
which lists the progress and new initiatives for integration, and possibly the signing
of a new protocol to the founding treaty. The ministers also meet during the year in
order to facilitate the integration. They issue decisions that achieve the goals set out
by the treaty and its protocols.
The Common Market Group is the executive body that is responsible for the
implementation of the measures. It performs this task through specialized working
groups8 whose members are made up of representatives of public entities of the
national governments. They meet various times a year and issue resolutions on how
to best implement the decisions of the Council. The Trade Commission is also an
intergovernmental institution chiefly responsible for the monitoring compliance of
the CET. They do this by issuing directives which report on findings. They also
suggest integration initiatives to the Group.
Less powerful but significant institutions were also established by the Ouro Preto
Protocol. The Joint Parliamentary Commission (JPC) is made up of parliamentar-
ians from the member states. The JPC was established in order to coordinate the
actions of the members’ legislatures and add a democratic dimension to the
integration process. Each member country has a parliamentary committee that
votes with MERCOSUR legislation before it is voted on by the full congress. No
act, treaty, protocol, or other related MERCOSUR agreement can be acted upon
unless it is approved by the members’ legislatures using their constitutionally
mandated processes. The members of these individual committees meet as the
JPC to discuss new and old initiatives of integration. They reach decisions by
consensus. While they meet as the JPC, the only power they possess is that of
consultation. Their real power, along with their parliamentary colleagues, therefore
is in voting on MERCOSUR legislation while performing their domestic mandates.
However, this power is latent because the voting record indicates that the individual
MERCOSUR legislative committees almost always indorse legislation and the
members’ legislatures almost always approves. There are two explanations for
the very large percentage of MERCOSUR related legislation approved. First, the
Council and the Group are sensitive to the objections of a member of a country’s
legislature. If there is a considerable obstacle, the Council and the Group will
negotiate so that the legislation passes. This is done informally during the formu-
lation stage. Another reason is that legislators often do not display interest in
international affairs. They are mostly concerned with the narrow interests of their
constituents and will enter the discussion when it involves those interests. Therefore
legislators will approve the decisions of the MERCOSUR committees and any other
committee involved in the legislative process. At the June 2003 presidential sum-
mit, one of the items on the communiqué was the establishment of the true
MERCOSUR parliament that would have legislative power and be directly elected.
8
The working groups are split up into the following areas: communications, mining, technical
regulations, financial matters, transportation and infrastructure, industry, agriculture, energy, labor
relations, employment, and social security.
174 G.M. Genna
NAFTA represents the newest and least developed integration effort. The economic
ties between the three members, Canada, Mexico, and the US, had been strong
before agreement went into effect. For example, by 1945, 83.5% of all Mexican
exports went to the US, and capital flowed south as labor moved north (Bulmer-
Thomas 1994). However, the arrangement maintains a central role for the US in the
three way relationship given the weaker ties between Canada and Mexico. Also the
linkages among the three, from the outset, were to be purely economic without any
discussion involving deepening the arrangement beyond its free trade character.
NAFTA began as a bilateral free trade agreement between Canada and the US
signed in 1988, went into effect in 1989, and was officially known as the Canada-
USA Free Trade Agreement (CUSFTA). This agreement (as well as NAFTA) was a
major departure from the US preference of exclusively negotiating trade agree-
ments multilaterally through GATT. Canada and the US originally invited Mexico
to participate, but Mexico opted not to participate. The chief US motivation to
include its southern neighbor was to improve overall US-Mexico relations. By
doing so, the US would be able to extend this relationship to other Latin American
countries in a future hemispheric-wide trade deal.
Mexico became a member of the new NAFTA after the accord went into effect
in 1994. Entering into an economic partnership with the US also marked a major
preference shift for Mexico. Having had a large portion of its territory taken from it
by the US as a result of the Mexican-American War, Mexican governments have
had an uneasy relationship with their northern neighbor. However, the new Partido
Revolucionario Institucional leadership’s abandonment of older and ineffective
economic policies for neoliberal thinking shifted the Mexican government’s view
towards its northern neighbor. The great need for capital on Mexico’s part and the
security required by Canadian and US investors prompted a stronger continuation
of the economic liberalization. Efforts begun under the de la Madrid administration
(Lustig 1993). One way that all sides could maintain liberalization in Mexico was
through NAFTA. During the negotiations, Mexico maintained a focus on capital
mobility in the form of direct investment as an imperative part of the final treaty
(Ros 1992).
The agreement was also a change in preference on behalf of Canada, but not one
that was as great as those of Mexico and the US. Instead it was part of a continuing
176 G.M. Genna
shift in preferences starting with the Tokyo Round (1973–1979) of trade talks of the
GATT. During this round Canada began to adopt a liberal trade policy. Also, it was
viewed that the existing US-Canada interdependence needed a legal framework to
improve confidence in the long-term continuation of the partnership (Weintraub
1997). This added greater certainty for capital and would encourage greater FDI
into Canada.
The provisions of NAFTA include only targets for free trade implementation and
dispute settlement. While it does have two side agreements, one for environmental,
the other for labor issues, NAFTA itself does not address notions of political
collaboration. After a 10 year phase out period, all goods and services under the
approved economic sectors are now traded tariff-free. Some economic sectors were
viewed as vital and were not touched by NAFTA. Petroleum in Mexico, culture in
Canada, and airlines and radio communications in the US are sectors that fall under
this provision (EWYB 2001). Recent unilateral legal changes in Mexico, however,
will soon liberalize the petroleum sector.
NAFTA does not have provisions to directly provide public goods such as
resource transfer from the US and Canada to Mexico. It was under the threat of
non-approval that such provisions were left out, including labor mobility (although
capital mobility was included). For example, the creation of North American
Development Bank in order to finance environmental and infrastructure projects
along the Mexico-US border required equal contributions from Mexico and the
US. We can look at the Mexican peso crisis during the 1994–1995 as an indication
of crisis management under NAFTA. In order to stabilize the falling value of the
peso, Mexico was in need of foreign reserves. Although the US Congress opposed
aid to Mexico unless onerous strings were attached, President Clinton used the
Exchange Stabilization Fund, which was under the Treasury Department and did
not require congressional approval, to provide credit to the Mexican central bank. If
a common economic resource was available through NAFTA, then domestic
politics were not to come into play under crisis situations. In its current framework,
special needs of the members will always involve domestic politics.
NAFTA also included a langue that would allow the free flow of goods across
the continent by liberalizing the transportation. According to the agreement, a cargo
truck carrying tariff-free goods from Mexico to Canada or the US should be allowed
to travel across the US. The provision, however, has not be enacted. Today, the
truck stops within a few miles inside the US and the cargo must be transferred onto
a truck that is registered in the US. The inability to enact the provision is due to
differences in Mexican and US trucking regulations that have not been homoge-
nized due to heavy lobbying by the US trucking interests.
Institutionally, NAFTA is weak because the negotiators wished to minimize its
political content. Given the asymmetrical economic power distribution among the
three and the low level of agreement the three have on international issues, the weak
NAFTA institutions safeguard against US domination over political matters. For
example, greater US dominance in a CET would mean dealing with the issue of
Cuba, a country that Canada and Mexico have normal relations with, but the US
does not.
6 Measuring Integration Achievement in the Americas 177
The only two institutions mentioned in the agreement are the Commission and
the Secretariat. Both serve to implement the agreement and to settle disputes
associated with it. The Commission is made up of relevant secretaries and ministers
of the three countries. Under the agreement, the Commission is chiefly responsible
for disputes, but the Secretariat actually performs the tasks of dispute resolution.
However it may be more accurate to refer to it as “Secretariats” because no single
administration exists. Instead the Secretariat is divided into three sections, one for
each country. They do not have an integrated structure9 and are not financed by a
common NAFTA fund. Each country is responsible to maintain their own NAFTA
Secretariat office, in their own countries, using funds allocated from their individual
national budgets. The work conducted in each of these offices is to mirror one
another. This means that each country is individually responsible for administering
the NAFTA provisions.
The chief task of these offices is to administer the dispute settlement found in
Chapters 11, 14, 19, 20 of the agreement. Disputes can arise between private
individuals or firms, but all disputes are handled by the governments of the three
on their behalf. The first attempt to settle a dispute is by intergovernmental
consultation. If it cannot be settled in this manner, then the dispute is given to the
Commission which is then handed off directly to the Advisory Committee on
Private Commercial Disputes. If the Committee cannot resolve the dispute, then a
panel of experts in the relevant field is chosen from a predetermined roster. It is up
to the relevant Secretariat office(s) to administer the dispute through these panels.
This panel now has the final authority to adjudicate the dispute.
NAFTA holds an IAS value of 1.67 for 2009. It is considered a full free trade
area since all non-tariff barriers have been removed. This gives it a score of 3 under
the trade in goods and services category. Full access is provided for foreign
investment and capital withdrawal, except for national government procurement
(a score of 3). Labor mobility is restricted to only a limited set of professions
(a score of 1). Supranational institutions are limited to information gathering
(a score of 2). Monetary policy is at the consultation stage (a score of 1). The
NAFTA partners have not discussed fiscal policy coordination.
Figures 6.1 and 6.2 illustrate the IAS values for the five RIOs. Figure 6.1 plots
the values over time. The RIOs with the longest and steady increases in integration
achievement is CAN and CARICOM. The least integrated, but among the oldest, is
CACM (SICA). Figure 6.2 displays the IAS values and the individual subcompo-
nents for 2009. As one would expect, the trade in goods and services category has
the highest values among the RIOs. Those RIOs with customs unions (CAN,
CARICOM, and MERCOSUR) have the highest values. The second highest value
is the degree of capital mobility category, except for SICA and MERCOSUR. The
degree of labor mobility is identically low in each RIO. Supranational institutional
importance is also low in each of the RIOs except for CAN. CAN is also the
exception for the relatively low levels of monetary policy coordination. Last, none
9
But they do have a single website.
Fig. 6.1 Integration Achievement Score Levels Over Time among Various Latin American RIOs
(1963–2009)
Fig. 6.2 Integration Achievement Score and Subcomponents by Latin American Regional Inte-
gration Organization (2009)
6 Measuring Integration Achievement in the Americas 179
of the RIOs registers any degree of fiscal policy coordination. Overall, the RIOs of
the Americas display low levels of integration achievement with an emphasis on
trade and investment. Labor mobility has not been achieved. The low level of
influence of supranational institutions, the low level of monetary policy coordina-
tion, and the lack of fiscal policy coordination indicates that decision-making
sovereignty is still firmly in the hands of the member states.
Thus far, this chapter has focused on measuring integration using the IAS and the
application of the measurement among the RIOs of the Americas. Next, I will focus
on two problems associated with using the measurement in analyzing regional
integration and their solutions. The first problem is the issue of overlapping
memberships. Although this chapter describes the integration achievement of five
RIOs, the Americas is home to eight integration projects (or more depending on the
definition of regional integration). As Fig. 6.3 shows, many states are members of
more than one RIO. The problem of overlapping memberships arises when the unit
of analysis is the individual country. Given multiple memberships, it would be
difficult to disentangle the spaghetti bowl in order to explain which RIO impacts the
country and to what degree. For example, if we are interested in explaining RIO
impact on the economic performance of Paraguay, it would be difficult to say if
integration due to the MERCOSUR agreements had any more or less of an impact
than say trade with the associate members of MERCOSUR (through the FTA), the
Union of South American Nations (UNASUR), or the Latin American Integration
Association (ALADI). Or if there was an indirect influence coming in from
CARICOM or NAFTA.
One solution is to aggregate the data so that the unit of analysis is not a single
country but pairs of countries (country dyads). This way, one could match the RIO
membership that includes both countries. For example, a Paraguay-Brazil pairing
would include the MERCOSUR value while the Paraguay-Bolivia pairing would
include the IAS value for the MERCOSUR-FTA. The problem persists if the pairs
are members of two or more RIOs. Paraguay and Brazil are members of both
MERCOSUR and UNASUR. Another solution would be to choose the higher
IAS value. In the case of Paraguay, MERCOSUR’s IAS value would be higher
than either UNASUR or ALADI. This solution rests on the assumption that greater
integration achievement translates to a greater impact on the country. One could
conduct robustness checks by swapping out IAS values or possibly including
multiple entries. The overlapping membership problem, however, disappears
when the unit of analysis is the RIO or if one is interested in counting the number
of memberships as the variable of interest.
The second problem involves implementation, and consistency of implementa-
tion, of RIO agreements. Signing and ratifying an agreement does not necessarily
mean that agreements are implemented. When coding agreements, the researcher
180 G.M. Genna
defection from agreements over time. The true value of integration should go down
if one or more member states begin to break agreements.
There are two related solutions to the problems of implementation and consis-
tency. The first is constant monitoring of RIO activities using field reports and
experts’ documentation. Field reports are often published by international organi-
zations or found in reputable yearbook entries. The reports often catalog problems
associated with implementation, including the degree of the problems. These
reports can be validated by experts who hold specific knowledge of particular
RIOs. The point is to use multiple sources in order to produce accurate measure-
ments. One can apply a conservative approach and not assume that an agreement
has been implemented (or implemented fully) until it can be validated by two or
more independent sources.
The second solution involves waiting. Since it may take some time before
information can be validated, one can institute a minimum delay period before
using the coded data. For example, the current version of the IAS only includes data
up to 2009 and therefore includes a 5 year delay. In other words, I go back through
5 years’ worth of information (field reports and experts’ documentation) to be sure
that the level of IAS in 2009 is accurate. The use of a 5 year delay is based on
personal experience. Using a longer delay may produce more accurate results but
limit number of cases in the analysis. Using a shorter delay may be less accurate but
increase the number of observations.
6.5 Conclusions
References
7.1 Introduction
B. Byiers (*)
European Centre for Development Policy Management (ECDPM), Maastricht,
The Netherlands
e-mail: bby@ecdpm.org
Q. de Roquefeuil
European Centre for Development Policy Management (ECDPM), Maastricht,
The Netherlands
Saana Consulting, London, UK
an EPA policy change, with indicators defined for each stage of the chain. Other
options range from a more descriptive and less complex approach focusing on
compliance, to more costly, analysis-intensive and in-depth impact evaluation
implement. Ideally the system should build on existing structures, linkages,
information and data-sharing systems and link to where there is an existing
demand and interest for such information.
7.2 Context
and at subsequent five-yearly intervals, in order to determine the impact of the Agreement,
including the costs and consequences of implementation and we undertake to amend its
provisions and adjust their application as necessary. (Official Journal of the European
Union, 2008, L 289/II/1955)
The EU-CARIFORUM EPA resulted from the end of the Lomé preferences that
had granted preferential market access to some of Europe’s ex-colonies up until
2007. These preferences provided the CARIFORUM countries with quasi duty free
market access on 97% on their goods, although a preferential quota system
remained in place on some of the region’s agricultural exports to the EU. For the
Caribbean, these preferences were particularly important for a set of key agricul-
tural goods, for example sugar, rum and bananas. The Caribbean Regional Nego-
tiating Machinery (CRNM), the body in charge of negotiating the EPA at the time,
estimated that, were Caribbean exports to lose Lomé preferences and face tariff
rates provided in the EU’s Generalized Scheme of Preferences (GSP), the next best
option available to them, sugar exporters would have to pay an annual $202 million
in additional levies on sugar, $64 million for Bananas, and $11 million for
Alumina.1
While preferences for agricultural products were clearly a motivation for signing
the EPA, the CARIFORUM region also considered other issues such as Services,
Investment and Intellectual property as beneficial for its own development strategy,
something that sets it apart from other EPA negotiating groupings that have been
more reluctant to engage on service liberalization and so-called Singapore issues.
This is due, to some extent, to the importance of sectors like Tourism for the region,
and the overall higher share of services in the region’s Gross Domestic Product
(GDP). Several studies and commentaries have already analyzed the agreement in
detail (see for example Meyn et al. 2009; Sauvé and Ward 2009).
1
http://www.normangirvan.info/wp-content/uploads/2008/11/eclac-assessmenrt-of-the-cf-eu-epa.
pdf
7 Monitoring Regional Integration in Practice: Reflections from the EU. . . 187
CARIFORUM
CARICOM-DR FTA
CUBA
DOMINICAN
REPUBLIC CARICOM
BAHAMAS HAITI
TRINIDAD &
SURINAME TOBAGO CSME
BARBADOS BELIZE GUYANA
JAMAICA
CARIFORUM-EU EPA
Fig. 7.1 Caribbean country membership of regional trade agreements (Source: ECDPM 2006)
It should also be noted that the EPA framework has been promoted with the
explicit aim of fostering regional integration. As the figure below illustrates, the
Caribbean is home to a host of regional groupings, amongst them the Organization
of Eastern Caribbean States (OECS), the Caribbean Community (CARICOM), and
the Association of Caribbean States (ACS). In order to accommodate the Domin-
ican Republic, who is not formally a member of CARICOM, but who used to
benefit from Lomé preferences, the EPA was negotiated with the Forum of the
Caribbean Group of African, Caribbean and Pacific (ACP) States (CARIFORUM).
All CARIFORUM Members are also members of CARICOM, except for the
Dominican Republic. CARIFORUM deals with EU related affairs – the EPA and
EU aid disbursed to the region – and is housed within the CARICOM secretariat
(Fig. 7.1).
At the time of writing, the implementation of the agreement was uneven across
the region. Indicatively, by late 2013 the treaty had been ratified in only 5 out of
15 CARIFORUM countries. Although in some countries, such as Barbados, the
EPA can be provisionally implemented without ratification, in others, legislation
requires an act of parliament even to lower tariffs, meaning that even this cannot
take place without ratification. Trinidad and Tobago is a case in point where the
agreement was not yet being applied even provisionally until very recently.
188 B. Byiers and Q. de Roquefeuil
Although existing regional integration efforts in the Caribbean region need not
impact on the EU- CARIFORUM EPA, in reality the history of integration in the
region plays an important role in defining relations across countries and between the
national and regional institutions.
Ever since the short-lived West Indies Federation from 1958 to 1962, and indeed
prior to this, regional integration has been seen in the region as a logical necessity
but political liability, with many practical difficulties. The statement of the Wood
Report from 1922 still rings true today:
however much it would be to the evident advantage of these colonies to secure machinery
for greater unity and cooperation, there are practical and political objections which, for the
present at any rate, make this impossible. (Payne 2008: xviii)
The geographical nature of the region clearly is a part of this. The CARIFORUM
regional grouping is made up of 15 countries scattered across the Caribbean, with
Belize City in the West lying more than 3500 km from Georgetown Guyana, with
13 island state signatories in between.2 Further, they vary in size from almost
10 million inhabitants in Haiti to 67,000 in Dominica; and in incomes from GDP
per capita of PPP$1171 in Haiti to PPP$31,978 in the Bahamas in 2011 (World
Bank 2013). A historical reading shows precisely that small size and lower incomes
have been a large part of the push factor towards greater integration to survive in the
world as small island states, while resistance has largely stemmed from the unwill-
ingness of larger countries in the region to finance smaller, poorer members (see
Payne 2008, for example). The enormous geographical spread and isolation of the
region’s countries is also seen in Fig. 7.2.
In addition to these more structural aspects that impinge on the enthusiasm for
regionalism in the Caribbean and therefore on progress in monitoring agreements, a
number of more current aspects also pertain.
Slowness in establishing EPA monitoring mechanisms relates in many ways to
the slow and uneven progress in implementing the EPA agreement itself. “Why
2
Based on online “Distancefromto” tool: http://www.distancefromto.net/distance-from/Guyana/
to/Belize
7 Monitoring Regional Integration in Practice: Reflections from the EU. . . 189
Fig. 7.2 The geographical spread of the Caribbean region (Source: ESF 2009)
monitor if there is nothing to monitor”, some say. Part of the reason for this slow
implementation is the priority that has been given to other contextual factors since
the signing of the agreement. The most salient of these include the following:
The financial crisis hit several of the Caribbean economies hard, particularly
through a fall in tourism receipts but also declining Foreign Direct Investment
(FDI) inflows, remittances and worsening fiscal balances. Countries less reliant on
the tourism sector, such as Guyana and Suriname, weathered the crisis better
although even here their main export sector, mining, has also been negatively
affected.3 The Dominican Republic and Belize have been least affected by the
financial crisis. While this has potentially distracted CARIFORUM governments
from EPA concerns, the same may also be the case for the EU, currently embroiled
in its own crisis. This is not to say that both parties would not ultimately like to see
the EPA as a success.
Further, several of the Caribbean islands were hit by natural disasters during
2009 and 2010. In October 2010, Hurricane Tomas struck some of the already
weakened economies such as Dominica and St. Vincent and the Grenadines, thus
leading to further declines in output and growth, and thus also occupying political
attention.
Partly resulting from the above, there is a perception that a general slowdown of
the regional integration process in combination with changes in governments may
3
Based on figures from IMF (2010, 2011a, b, c, d, e, f, g, h, i, 2012).
190 B. Byiers and Q. de Roquefeuil
have allowed regional integration and EPA implementation to slip down the agenda
in a number of countries.
On top of this, the EU is not the principal trading partner for CARIFORUM
countries, and for a number of them the importance of the EU is declining relative
to other partners.4 For most, the EU represents the third or the fourth most important
trade partner. As such, the EPA is perhaps less important economically than trade
relations with other trade partners.5
Notwithstanding the importance, the EPA negotiations were controversial in the
region. There remains some debate surrounding the worth and potential impact of
the EPA in the region that may influence some governments’ impetus to implement
the agreement (see for example Bishop et al. 2012).
As the EU has unilaterally implemented most of its commitments, delays in
implementation on the CARIFORUM side do not face risks of slow-downs on the
EU-side, perhaps again lowering the priority given to the exercise, particularly
given the current and historical background.
In several countries, particularly the smaller members, government capacity to
implement the agreement in a timely manner is a constraint, with very few qualified
staff available, for example, for legal drafting, while the EPA agreement requires
engagement and coordination across a large number of institutions6. Given this
constraint, again monitoring may not be considered a priority.
Considerable fruitless efforts have already been made to establish a monitoring
mechanism through civil society at the time the EPA was signed but petered out due
to the lack of resources for regular meetings and the lack of progress in actual
implementation of the agreement. These failures may make it harder to garner
support for renewed attempts.
As this contextual summary suggests, both historical and current conditions
hinder implementation of the CARIFORUM EPA agreement thus providing an
unfavorable context for establishing a monitoring system. This is fundamental to
keep in mind in discussing the form a monitoring system might take.
4
Average trade levels between the EU and the Caribbean ACP countries have decreased during the
period 2006–2011 with the Caribbean share of total EU exports decreasing from 14.9 to 9.2 %,
while the share of total EU imports has decreased from 20.5 to 14.5%. Based on EC figures: http://
ec.europa.eu/trade/creating-opportunities/bilateral-relations/statistics/
5
Looking at EU-Caribbean trade relations through time, trade with Haiti, St. Kitts and Nevis,
Trinidad and Tobago and Surinam (despite a temporary drop in 2010) has increased between 2006
and 2010, while flows between the EU and Antigua and Barbuda, Bahamas, Barbados, Grenada,
Jamaica, St. Lucia and St. Vincent and the Grenadines have decreased during the same period.
Particularly Jamaica, Grenada and St. Lucia have experienced a sharp decline in their trade with
the EU.
6
Humphrey and Cossy (2011) cite the case of Barbados where 27 bodies are counted as “Principal
Implementing Bodies”.
7 Monitoring Regional Integration in Practice: Reflections from the EU. . . 191
As pointed to above, two broad, overarching reasons stand out for monitoring the
EPA. One is relatively straightforward: monitoring compliance and implementation
of the agreed measures. This primary function of ensuring that both parties imple-
ment the required changes in national legislation to comply with the EPA is far
from straightforward in a regional context mired by institutional weaknesses, and in
which decisions taken at the regional level are often not applied in national
legislation (Stoneman et al. 2012).
The second, more innovative function of the monitoring mechanism proposed in
the agreement is “to determine the impact of the Agreement, including the costs and
consequences of implementation” in five year reviews. The text of the joint
declaration is quite explicit in stating that the outcome of this review, informed
by the monitoring system, could lead to “amend its provisions and adjust their
application as necessary” as cited above. In this case the monitoring provisions of
the EPA combine both the function of a compliance mechanism and a safety valve,
injecting a measure of flexibility in the agreement.
Compliance is easier to monitor than development impact, with clear require-
ments that must be enacted. Here, the act of monitoring amounts to finding out
whether or not a country is in compliance or not. This nonetheless raises some
practical considerations, such as how to confirm that measures agreed in principle
are applied in practice. This may imply a need for “proof of compliance” of specific
elements of the agreement, for example from the customs system to show that the
required tariff is indeed being applied at the border.
With regards to outcomes, given the explicit focus on development impact of the
agreement, Meyn et al. (2009) list three effects that could be monitored: the export
effect, the adjustment effect, and the revenue effect.7 The export effect articulates
the idea that in order to benefit from increased market access to the EU, Caribbean
exports might need assistance in the form of Aid for Trade or other financial flows.
The adjustment effect refers to the increased competition facing Caribbean industry
after the liberalization of EU imports. Finally, the revenue effect concerns the loss
in tax revenue resulting from the lowering of tariffs. It is important to note that in
the case of the domestic adjustment and revenue loss, the EPA already contains
significant flexibilities in the form of safeguards and provisions allowing Caribbean
countries to reinstate previous MFN tariff should they run into “serious
difficulties”.
7
http://www.odi.org.uk/sites/odi.org.uk/files/odi-assets/publications-opinion-files/4205.pdf
192 B. Byiers and Q. de Roquefeuil
The general idea of combining compliance and impact monitoring is, as Brüntrup
et al. (2008) put it “to ensure that results feed back into the design and implemen-
tation of the agreement or accompanying measures”. Ideally, monitoring would be
accompanied by regular and continuous consultation with stakeholders to identify
problems with trade policy and its implementation, to identify unforeseen impacts,
and allow adaptation of policy and implementation strategies to take account of
newly uncovered issues.
Monitoring has been occurring in the Caribbean to some extent but only in an
ad-hoc manner. Some, but not all CARIFORUM governments have set up national
implementation units, and a Regional EPA Implementation Unit, housed in the
CARICOM headquarters, provides technical assistance to the national nodes.
However, the systems of coordination and communication among the national,
sub-regional, regional, and extra-regional bodies are reportedly ad-hoc, with the
division of labor unclear between the national and regional levels, and with regards
the CARIFORUM directorate mandate vis-a-vis national EPA units on a number of
issues with both parties awaiting and expecting directions or demand from the
other.
A more systematic approach to monitoring would require systematized infor-
mation and data collection and analysis, institutionalized public-private sector
dialogue to provide information on the principal problems to be addressed and to
arrive at appropriate remedies. From a policy perspective, it is useful to conceptu-
alize monitoring efforts along four lines: objectives, scope, institutional actors and
methodology.
At the time of writing, the roles, purpose and mandate of different EPA implemen-
tation units vary across countries in the Caribbean region. This is problematic for a
regional agreement where CARIFORUM countries are expected to speak with one
voice and apply some measure of uniformity. For example, some national units
have taken an explicit export promotion role with a view to promoting the utiliza-
tion of the trade agreement, while others are focused on compliance. Some have
their mandate restricted to monitoring implementation, and not development
impact. A regional system to monitor a regional agreement such as the EPA
requires at least some level of agreement on the overall purpose of monitoring to
allow aggregation across the region and comparability, while allowing for flexibil-
ity in terms of the depth of analyses and sectors chosen at the national level.
The objectives of the EPA and what should be monitored are defined in article
1, and are as follows: (i) poverty eradication, (ii) promotion of regional integration,
(iii) integration of CARIFORUM states in the world economy, (iv) improving trade
policymaking capacity of CARIFORUM states, (v) increasing investment,
7 Monitoring Regional Integration in Practice: Reflections from the EU. . . 193
(vi) private sector initiative, supply capacity, competitiveness and economic growth
in the CARIFORUM region and finally (vii) strengthening EU-CARIFORUM
relations on the basis of solidarity and mutual interest. Article 3 further defines
sustainable development as ‘(...) the human, cultural, economic, social, health and
environmental best interests of their respective population and of future genera-
tions’, while ‘(...) decision-taking methods shall embrace the fundamental princi-
ples of ownership, participation and dialogue’.
Not all these objectives may initially be covered by a national and regional
monitoring mechanism. Given the emphasis on poverty eradication, sustainable
development and economic growth, the need to prioritize, and the somewhat more
direct impact the EPA may have on these objectives, this chapter focuses on the
necessary indicators to measure the impact the EPA might have on these
dimensions.
Closely related to the question of the objectives of monitoring efforts, the focus of a
monitoring system must also be addressed early on. In the Caribbean, these
questions revolve around whether the regional and national monitoring systems
should seek to go beyond the EPA per sé to monitor broader trade policy in general,
or integrate with monitoring of the Caricom Single Market and Economy (CSME).
Beyond choosing the exact policy areas to monitor, it is also important to distin-
guish between monitoring of compliance with the agreement, and monitoring the
outcomes or impacts, as explained above.
An effective monitoring system would ideally mainstream EPA monitoring into
the trade policy cycle of the Caribbean countries. This would take EPA implemen-
tation and its effects as a subset of broader trade policy implementation and
impacts. Regular updates would then be used to consult with stakeholders on how
policies should adjust (or not) to the information that emerges. This would not
necessarily require vast financial resources, but would require political support to
ensure cooperation among Ministries and regional bodies to ensure regular and
systematic communication and information flows, overseen by a National Moni-
toring Committee or a similar such body. This option then implies the need to
embrace trade policy monitoring at a broader scale, and as a tool to improve and
inform economic policy in all the countries in the CARIFORUM region, something
that however may be over-ambitious given the constraints faced.
Based on interviews, it may also be useful to distinguish between impacts from
the agreement and actual “use” of the agreements – a form of intermediate impact,
which can then guide export promotion efforts or technical assistance to firms
struggling to meet sanitary and phytosanitary measures (SPS) or overcome Tech-
nical Barriers to Trade (TBT). This implies carefully gathering data relating to more
administrative aspects of trade. In the Caribbean issues that were mentioned
relatively often were, for example, trade with the French overseas territories,
EPA-related trade certificates emitted for trade in goods in the region, or numbers
194 B. Byiers and Q. de Roquefeuil
Given the need for broad discussion and ownership of any monitoring project, there
needs to be an early discussion of who will contribute to monitoring efforts and to
establishing a monitoring system, and how these actors will interact. Monitoring
and the use of monitoring is inevitably political, and there is a risk of “capture”. But
this should not be seen as a reason not to monitor, nor should it be seen as a reason
not to involve diverse segments of society in these efforts.
For the purpose of objectivity and inclusiveness, civil-society actor participation
is key in designing the system and following the process. This should include
representatives with an interest in social and environmental welfare, but also
private sector bodies and associations and other interested groups. Indeed, private
sector coalitions and chambers of commerce are key actors for the monitoring
process and are taking an increasingly active role even in engaging on issues of data
collection and monitoring.
A useful proposal might be a National Monitoring Committee. This would be
formed of representatives from government, civil society and the private sector,
including the Ministries of trade and finance, with particular directorates involved
in trade policy, and bodies outside government such as universities with a role in
gathering data on and analyzing the economy. The precise role of each of the
members of such a body or committee naturally needs to be clarified early on, but
this might be done according to what information they can offer, what constructive
inputs can be brought, or a desire simply to follow the process as an observer. The
role of such a committee would be to oversee and guide the monitoring process,
potentially overseeing other national level sector-specific monitoring committees,
but also liaise with regional structures.
Having agreed on the actors to involve, it is also important to agree on the
institutional arrangements for their engagement in terms of the organizational
structures and responsibilities. Ultimately, this must also reflect the practicalities
of different forms of interaction, particularly between the national and regional
7 Monitoring Regional Integration in Practice: Reflections from the EU. . . 195
The scope to be covered clearly determines how areas are to be monitored, with a
range of possible methodological approaches. The choice of approach depends on
the responses to the issues raised above, as well as available data and capacity for
analysis. For the monitoring system to be sustainable one would expect that it
should build as far as possible on existing structures, linkages, information and
data-sharing systems.
In relation to monitoring compliance, although potentially a technocratic exer-
cise of checking that elements of the agreement have been formally incorporated
into national legislation, some degree of systematic control is required to ensure
that practice follows form. The basics for setting up such a system exist in the
Caribbean with systematized matrices of country-level requirements requiring only
to be completed and compiled at the regional level. Nonetheless, compilation itself
apparently poses problems given the unclear mandates at different levels of the
regional machinery.
A fundamental constraint for monitoring impact is data quality and availability.
This should be kept in mind when coming up with a methodological framework. In
the Caribbean, this was cited as a major constraint to previous monitoring efforts
and was regularly cited as a constraint, although there are clearly differences across
countries.
In terms of establishing impacts, potential approaches include a “Results-chain
approach”. In this approach, for specific sectors one maps out the direct and indirect
potential effects of an EPA-induced policy change, with indicators defined for each
of the relevant stages of the chain. While not scientifically rigorous, this is seen as a
feasible option where more scientific approaches are impractical. Even so, this
approach is not feasible for all potential EPA impacts, again requiring some
answers to the above questions on scope before precise methodologies can be
defined.
Mapping out potential impacts in a result chain logic is particularly difficult for a
modern free trade agreement like the EPA. Mapping out the adjustment effects of
tariff dismantlement and revenue loss is feasible, and has been done for three
Caribbean countries in the past (see Meyn et al. 2009). For services and Singapore
issues, it is significantly harder, but not impossible. In many cases it will require
detailed information about the market structure of a given industry in a country, and
of the local regulations in place. Naturally, not all transmission channels are
relevant for all sectors, and some might be more directly affected than others for
a given policy change introduced by the EPA. The channels may also affect each
other, or run in contrary directions regarding development impact.
196 B. Byiers and Q. de Roquefeuil
Two cross cutting concerns emerge from the review of the monitoring efforts of the
EPA going on in the Caribbean. These are likely to be of value for other policy
oriented attempts at monitoring regional integration, hinting as they do at tensions
between the aims and principles of a monitoring mechanism (inclusiveness, feed-
back loops, etc.).
The first issue is the inherently political nature of monitoring, and the political
economy dynamics behind a monitoring system. This can be read as the “tradi-
tional” political economy of trade liberalization, in which groups affected by
adjustment costs are much more likely to organize and lobby against free trade. It
is important to keep in mind that to such groups, attempts at monitoring provide an
access point into policy making, bringing both benefits and risks. It is therefore
important to incorporate this sort of thinking when thinking about the actors to
involve in a monitoring system, and when devising the institutional relations
between different parts of a monitoring system, although to date this has been
raised more as a concern by the European Commission than actors in the Caribbean.
This does not mean that groups affected by economic adjustment should be
systematically kept out of a monitoring system. On the contrary, a well-designed
monitoring system should be able to identify such groups and provide them with
financial assistance that will help them to improve their competitiveness or shift
their activities to a new sector. It does mean, however, that monitoring “impact” of
the agreement is likely to arouse interest from groups who might not be represen-
tative of the rest of society, and that there is a real risk of “capture” if this is not
recognized from the outset.
Relatedly, and as mentioned, the Caribbean context is marked by a certain
degree of scepticism about the benefits of the EPA. This scepticism cannot, to a
large extent, be tied back to protectionist interests. Rather, it is often expressed in
terms of choice of development model for the Caribbean, the balance between
regional integration in CARICOM against the signing of agreements with external
partners, and the place of the EU in the group’s external ties. A monitoring system
should not seek to sidestep these broad societal debates, but, on the contrary, to
provide a platform for them.
Secondly, there is a tension between, on the one hand, a trade agreement’s purpose
of providing a stable and predictable environment for private sector operators and
7 Monitoring Regional Integration in Practice: Reflections from the EU. . . 197
investors, and, on the other, the flexibility introduced by a monitoring system that
can modify some of the agreement’s provisions. Indeed, a key function of a trade
agreement is to act as a “lock-in” instrument for trade policy and domestic regula-
tion that falls under its scope. If these can be rolled back too easily because of a
monitoring provision, or if it is perceived to be so by investors and operators, the
agreement’s worth is likely to diminish greatly.
The two are valuable objectives, but they are to some degree conflicting.
Institutional provisions can deal with these tensions. In the EU-CARIFORUM
EPA for example, there is an elaborate institutional machinery regulating the
governance of the agreement. This greatly diminishes the chances of seeing one
side coming back on its commitments unilaterally because of the monitoring
provision.
Nevertheless, introducing flexibility, combined with the capture risk highlighted
above, does raise some questions. For example, in 2012 some governments asked
that some of the tariff lines be taken out of the region’s liberalization commitments
in view of the EPA’s five year review, because the industries behind them would
face serious competition from Europe.
7.4 Conclusion
References
Bishop M. L., Heron, T., & Payne, A. (2012). Caribbean development alternatives and the
CARIFORUM-European Union economic partnership agreement. Journal of International
Relations and Development, 1–29.
Brüntrup, M., Jerosch, F., Keijzer, N., Loquai, C., Rampa, F., & Reichert, T. (2008). Monitoring
economic partnership agreements inputs to the negotiations and beyond. Bonn: DIE and
ECDPM.
ECDPM. (2006). Overview of the regional EPA negotiations: Caribbean-EU economic partner-
ship agreement (ECDPM InBrief 14C). Maastricht: ECDPM.
ESF. (2009). http://www.esf.be/new/wp-content/uploads/2009/09/carifourm-states-map.png
Humphrey, E., & Cossy, M. (2011). Implementing the economic partnership agreement: Chal-
lenges and bottlenecks in the CARIFORUM region (ECDPM Discussion Paper 117). Maas-
tricht: ECDPM. www.ecdpm.org/dp117
IMF. (2010). Antigua and Barbuda: 2010 Article IV Consultation and Request for Stand-By
Arrangement – Staff Report; Supplements; Public Information Notice; Press Release; and
ED’s Statement. IMF Country Report No. 10/279, September 2010.
IMF. (2011a). Dominican Republic: Fourth Review Under the Stand-By Arrangement and
Request for Waiver of Nonobservance of Performance Criterion-Staff Report; Informational
Annex. Press Release on the Executive Board Discussion. IMF Country Report No. 11/177,
July 2011.
IMF. (2011b). Trinidad and Tobago: 2010 Article IV Consultation-Staff Report, Supplement, Staff
Statement, Public Information Notice on the Executive Board Discussion, and Statement by the
Executive Director for Trinidad and Tobago. IMF Country Report No. 11/73, March 2011.
IMF. (2011c). Belize–Article IV Consultation. IMF Country Report No 11/340, November 2011.
IMF. (2011d). The Bahamas – 2011 Article IV Consultation. Country Report No. 11/338,
December 2011.
IMF. (2011e). St. Vincent and the Grenadines, 2011 Article IV Consultation. Country Report
11/343, December 2011.
IMF. (2011f). St. Lucia – Request for Disbursement under the Rapid Credit Facility and Emer-
gency atural Disaster Assistance – Staff Report, Staff Supplement, Press Release and State-
ment by the Executive Director for St. Lucia. IMF Country Report No. 11/278, September
2011.
IMF. (2011g). Suriname: 2011 Article IV Consultation—Staff Report, Informational Annex,
Public Information notice on the Executive Board Discussion and Statement by the Executive
Director for Suriname. IMF Country Report No. 11/256, August 2011.
IMF. (2011h). Guyana: 2010 Article IV Consultation—Staff Report, Supplement, Public Infor-
mation notice on the Executive Board Discussion and Statement by the Executive Director for
Guyana. IMF Country Report No. 11/152, June 2011.
IMF. (2011i). Jamaica: Third Review Under the Stand-By Arrangement—Staff Report; Informa-
tional Annex; Staff Supplement; and the Press Release on the Executive Board Discussion.
IMF Country Report No. 11/49, February 2011.
IMF. (2012). Dominica – Request for Disbursement under the Rapid Credit Facility—Staff Report
and Press Release. IMF Country Report No. 12/47, February 2012.
Meyn, M., Stevens, C., Kennan, J., Highton, N., Bilal, S., Braun-Munzinger, C., Lui, D., van
Seters, J., Campbell, C., Rapley, J. (2009). The CARIFORUM–EU Economic Partnership
Agreement (EPA): The Development Component, Study by ODI, ECDPM and CaPRI.
European Parliament, DG for External Policies of the Union.
Payne, A. J. (2008). The political history of CARICOM. Kingston: Ian Randle Publishers.
Sauvé, P., & Ward, N. (2009). The EC-CARIFORUM Economic Partnership Agreement: The
Outcome on Services and Investment. ECIPE.
7 Monitoring Regional Integration in Practice: Reflections from the EU. . . 199
Stoneman, R., Pollard, J. D., & Inniss, H. (2012). Turning around CARICOM: Proposals to
restructure the secretariat. Report prepared and submitted to the CARICOM Secretariat by
Landel Mills Consultants, UK.
World Bank. (2013). World development indicators. http://data.worldbank.org/indicator/NY.
GDP.PCAP.PP.CD. Accessed 20 May 2013.
Chapter 8
Comparing Integration in Europe and Latin
America: Wishful Thinking, Self-Perception
and Reality – A Comment
Joaquı́n Roy
During the first decade of the new century, the way in which European and Latin
American integration has evolved has produced surprising parallels and differ-
ences, depending on the prism applied. There are three different but complementary
angles of analysis, highlighting this intriguing contrast. In the first place, evidence
shows the survival of historical schemes and different frameworks, which are in the
company of more ambitious and legally documented projects, all in the company of
well recognizable frustrations in regional integration experiments.
In the second place, one can easily detect the appearance given by a considerable
degree of self-serving evaluation of governmental and institutional origins that has
persisted on both sides of the Atlantic. In the third place, there is the flat reality,
subject to more objective inspection and commentary. The result of the combina-
tion of these three dimensions does not have to be considered as negative or
positive, but only as realistic.
Latin American integration is a glass both half-full and half-empty. For example,
the ambitious integrationist would confidently state that the current balance of
accomplishments is disappointing. However, more optimistic observers might
argue that the picture could be worse, taking into account the dual problem of
high expectations combined with a lack of alternatives. But even in this case the
option of analytical silence cannot be justified regarding the limitations to Latin
American integration and its root causes.
J. Roy (*)
University of Miami European Union Center of Excellence, University of Miami,
Coral Gables, FL, USA
e-mail: jroy@miami.edu
entities. Not only did they differ in membership (CELAC has been depicted as OAS
without Canada and United States), but in essence. CELAC has no treaties, insti-
tutions, headquarters or budget. CELAC, he added, is not an “organization”, but a
“community”. In spite of that, the EU, always respectful of Latin American and
Caribbean claims of own personality, has accepted the transfer of the role previ-
ously played by the Rio Group to the new body. Meanwhile, one has to remember
that the institutional limitations of the Central American system, the winner of the
recent EU moves in signing Association agreements, are also reflected in the
identification of the signatories of the Agreement. The documents show that the
partners are the European Commission and five separate Central American states
with signing subject to individual ratification. The Central American Integration
System (SICA) does not have supranational sovereign authority.
Latin America has been seized between its ambivalent path towards regional
integration and the attraction of the influence of the European model. However,
the so-called “new regionalism” and the still present "third wave" of integration
have a variety of options for insertion into international markets, which can be used
simultaneously and not exclusively with mutual benefits. Dissatisfaction with the
existing blocks can lead to different scenarios. The first is to continue to be guided
by inertia; the second is the perennial temptation to opt for a tabula rasa.
There would be a third way, based on learning from the experiences to be applied
to new entities, which should not be considered as examples of strictly “integra-
tion”. This is the case, for example, of UNASUR. This entity is still undefined as a
consultation forum, security framework or entity with real integration potential. In
sum, instead of abandoning all the accumulated experience, priority should focus
on regional integration aspects such as stability and the creation of a native brand,
combined with the acceptance of the guideline which has already prevailed in
Europe itself. This is the so-called “variable geometry”. At all times, there should
be a need to accept the centrality of the agreements to be implemented and
respected, with a willingness to change and adapt them to new circumstances.
It might be wondered at this point about the causes of the slow and frustrating
regional integration path in Latin America. In that sense, it is possible to point to
some explanations. Among them, one should pay attention to the innate character-
istics of the nature of the huge Latin American territory, producing a geographical
slavery. In contrast to the easy land community of most of Europe, where in the
course of a day you can travel by rail between distant countries and where road
transport is a regular means of trade and tourism, in Latin America the distances
make these internal movements arduous if not impossible. Thus, this factor rein-
forces the nature of territoriality, imposing customs procedures and legal barriers
against the free movement of citizens.
204 J. Roy
Secondly, historical evidence reveals that war, the root of European integration,
has hardly had relevance in Latin America, with the exception of specific inter-
governmental conflicts, which are still a cause of quarrels and disputes between
states by claims of geographical limits. These, ironically, very often do not have a
natural geographical origin, but are relics of colonial administrative limits. Inter-
nally, regional integration is shown as problematic when the full national integra-
tion itself is nonexistent or very weak in many countries. The majority of the
population is not inserted, it is discriminated against and it is beaten up not only
by poverty but also by inequality, the worst on the planet. An inclusive nation made
by choice, not by inheritance, is then a chimera.
While Latin American history reveals the absence of a Jean Monnet who could
persuade the power about the benefits of integration, the strength of presidentialism
that circles as the prevailing political system since the time of the founding fathers
is an imposing obstacle for the successful implementation of integration ventures.
The successive waves and transformations of populism and caudillismo have
converted the prospect of sharing sovereignty into an unusual doubtful task.
Above all Latin American shortcomings, what stands out is the misunderstanding
of the concept of supra-nationality or its explicit rejection in terms of the establish-
ment of institutions, independent and equipped with budgets able to finance inte-
gration projects. Weakness (or simply absence) of the various general secretariats in
the mode of the EU Commission make the effective functioning of integration a
mission impossible, where any decision depends on the schemes agreed by presi-
dential summits. Hence, the rise of the convincing arguments generating an alter-
native ’integration’ model of free trade agreements modeled from the United States
and adopted as a supplementary remedy of the European Union.
Despite all the difficulties, the EU model is still valid in this context. In the last
decade, the survival of the original and essential European project has been based
on learning from the mistakes of the past and in the adaptation of new risky
frameworks to new circumstances. The most recent of these self-help corrections
has been the reform of the institutions by means of successive treaties leading to the
Lisbon compromise. In this case, Europe has not fallen in the double trap of trying
to start from scratch or leave the job in the hands of inertia. The best argument in
response to the questioning and doubt about the European integration process, both
for its justification in times of crisis as by the permanence of its role model or
universal benchmark, especially to Latin America, is the present balance, not
speculation or predictions without basis about its future.
First, the forecast (or desire expressed by some) of the death of the EU is
answered with another assertion: if it dies, it will have been of an illness called
“success”. The EU has met each and every one of the missions that were imposed
on it, from the fundamental and foundational mandate of bringing an end to
European wars. Secondly, the combination of international cooperation and internal
effort has generated the fact that never in European history have more people in
more countries lived better in nearly half a century. Three generations living today
remember intimately, or by family memories that have lived beyond their parents
and grandparents.
8 Comparing Integration in Europe and Latin America: Wishful Thinking. . . 205
Daniel Tanoe
9.1 Context
Regional integration and the creation of an African common market has been the
vision of African leaders since the early years of independence. There are several
critical reasons for this.
• First, a common market combining Africa’s 54 mostly small and fragmented
economies will lead to economies of scale that make countries competitive.
• Second, it would provide access to a wider trading and investment environment,
inducing backward and forward linkages, and promote exports to regional
markets, building experience to enter global markets.
• And third, it would provide a framework for African countries to cooperate in
developing common policies and services for trade, finance, transport and
communications and other sectors.
For one principal reason: to regain control of its own economic destiny and to
cope with a rapidly globalizing world by increasing its bargaining power and
asserting its own priorities and preferences in diversifying its production structure.
At the sub-regional levels, various integration arrangements have been launched
by the Regional Economic Communities (RECs) and other sub-regional integration
organizations. The RECs aim to create a Free Trade Area (FTA), a Customs Union
(CU) and eventually a Common Market (CM) and an Economic Union (EU)/
African Economic Community (AEC). The strategy for achieving these objectives
is to stabilize and gradually remove tariffs and non-tariff barriers (to lead to FTA),
D. Tanoe (*)
Investment Policy Section, Regional Integration and Trade Division, United Nations Economic
Commission for Africa (UNECA), Addis Ababa, Ethiopia
e-mail: dtanoe@uneca.org
adopt a common external tariff with respect to trade with third countries (to lead to
CU), harmonize and implement common policies (macro-economic policies and
sector policies notably in trade, money and finance, transport/communications,
agriculture and industry) and promote free circulation of all factors of production
including people (to lead to CM and EU).
In light of what is stated above, it is possible to group the integration objectives
within the sub-regions and the continent into the following three core categories:
• Short-term objectives: Achievement of FTAs and CUs
• Medium-term Objectives: Achievement of CM, including monetary union
• Long-term objective: Achievement of the EU/AEC.
The details, sequencing, modalities, and tempo for achieving the above core
categories of objectives within the sub-regions obviously vary from REC to REC.
However, differences in terms of substance, schedules and expected end-results
among the majority of the RECs/integration institutions are fairly marginal.
Assessing Regional Integration in Africa (ARIA) was initiated by the Economic
Commission for Africa (ECA) as a series beginning with a landmark report in 2004
(UNECA 2004). ARIA thoroughly assesses the status of Africa’s integration
process; the performance of Member States and RECs; and draws lessons for the
future. Although there has been an extensive array of reports, meetings and
conferences on the subject of integration in Africa, ARIA was the first coherent
and comprehensive analysis on a national, sub-regional, regional and sectoral basis.
The report makes recommendations on what needs to be done to enhance progress
and overall effectiveness and efficiency of the process.
ARIA’s primary audience comprises African Heads of State and Government.
Its main justification was to provide them with a management tool to monitor the
progress being made on integration commitments that they, and their governments/
countries, have made. Thus, African leaders and governments would systematically
know about how they had performed on their undertakings. The idea was to
evaluate progress made and things left undone on a biennial basis, and thereby
help spur African governments to revitalize their integration efforts.
The report’s secondary audiences include:
• African ministers, public and parastatal institutions across all sectors and
portfolios;
• The regional African plurilateral community: i.e. African Union; African Devel-
opment Bank; the Regional Economic Communities (RECs); other regional and
sub-regional institutions;
• The African private community that should become involved deeply in integra-
tion matters – i.e. the financial sector, the business community, the professions,
the academic community, the private voluntary and Non-Governmental Orga-
nization (NGO) sector; and
• African civil society and the public at large: i.e. those who most need to be
convinced of the need for African integration and involved in the process of
bringing it about.
9 Assessing Regional Integration in Africa (ARIA): Indicators of Integration. . . 211
In addition to its value for Africa, ECA was of the view that ARIA would have
considerable value, utility and appeal to other stakeholders outside Africa such as
institutions in the United Nations (UN) family and a variety of sector-oriented
development agencies, the international academic, research and consulting indus-
try, and the international business community with a presence in, or potential
investment interest in Africa.
It was crucial that if the ARIA assessments were to add value to previous reports
on Africa’s integration and be meaningful and relevant, key indices of integration
needed to be included, and appropriate methodology and measurement scales
developed, and continually refined. Such indices/measures needed to be based on
an as objective and hard quantitative base as possible in order to avert/avoid or at
least minimize arguments about “subjectivity” or “bias” in portraying the perfor-
mance (i.e. integration effort) of one country or another, or of one REC or another.
Performance indicators were indeed an innovation of the ARIA exercise and
were developed as a basis on which comparison of results and performance against
common denominators were analyzed in each of the following clusters. As this was
the first time ECA was embarking on this idea of constructing indicators on regional
integration to include in the ARIA report, we reasoned that any attempt to be
comprehensive, exhaustive or definitive was not only ambitious, but probably,
also unfeasible for a number of reasons including data limitations. We therefore
wanted to focus on what is affordable, doable and practicable for the first ARIA.
A number of data sources were used to build the data-base and to generate the
time series data for the various indicators of the sectors covered by the analysis in
ARIA I. The greater body of the data was collected through a structured question-
naire specifically designed to collect both quantitative and qualitative data and
information on the indicators at the country and REC levels in Africa. Responses
for the questionnaire were obtained through field missions to 51 countries and to all
14 RECs. This was supplemented by secondary but very valuable data sources from
a number of sources such as the UN organizations including United Nations
Conference on Trade and Development (UNCTAD) and United Nations Industrial
Development Organization (UNIDO), as well as from the World Bank, Interna-
tional Monetary Fund (IMF), United States of America (USA), statistical sources,
specialized sectoral institutions, research bodies, various websites and other
published and unpublished data sources.
weighting became important. Weights were assigned for all indicators reflecting
each indicator’s “relative importance” in measuring performance within the cluster.
Further weights were assigned to each cluster, which reflected the relative impor-
tance of the cluster for regional integration vis-a-vis the other clusters. These
weights then allowed for the construction of an overall Integration Index for a
given country or REC. Assigning weights, however, was not a straightforward
procedure. It required knowledge and understanding of the particular indicators/
clusters and their respective importance to enhance regional integration. Thus,
weights were obtained through consultations and discussions among cluster
experts.
In a nutshell, the main objectives of constructing indices in the context of ARIA
were:
• To assess each country’s performance and relate it to the goals and objectives of
each REC and that of Africa as a whole, as well as the performance of each REC
to that of Africa;
• To compare the contributions that each member country in a REC has made
towards the realization of such goals and objectives, in addition to the contri-
butions that each REC has made to that of the continent at large;
• To monitor the performance of each country, REC and the continent as a whole
with respect to regional integration efforts over time; and
• To enhance the quality of the analysis of ARIA by providing indices upon which
scores and ranking are made at country, REC and Africa level.
Where:
Xij,0 ¼ the value of Xij at time t ¼ 0 (the base year value; in this case is 1994)
i ¼ 1,2, . . .,N indicators
j ¼ 1,2, . . .,J countries
t ¼ 0,1,2,. . .,T years
The indices as defined in (9.1) were generated for all countries within a REC or
within Africa. These indices measure the relative changes, as compared to the base
year, of a particular indicator.
9 Assessing Regional Integration in Africa (ARIA): Indicators of Integration. . . 215
Where I*ir,t is an index for the ith indicator for the rth REC at time t. X*ir,t is the
aggregate value of the ith indicator (aggregated over all countries in a REC) for the
rth REC at time t., and X∗ ir , 0 refers to the aggregate value of the same indicator at
base year (t ¼ 0). Thus I*ir,t is used to measure changes over time at REC level.
Case 2
• The average of the best performers of the REC:
– If a REC has more than 6 members, the average of the top 4 performers of the
REC (based on indices) is taken as the yardstick
– If a REC has less than 6 members, the average of the top 2 performers of the
REC (based on indices) is taken as the yardstick
– For continental level comparisons, the average of the top 6 African per-
formers is taken as the yardstick
Case 1 is straightforward. However, most indicators (for example, exports,
imports, etc.) do not have a pre-determined target that can be used as a norm or
yardstick. In such a situation, the approaches stated under case 2 were used to
generate a yardstick for each indicator. In this case, calculating the average of the
top 4 (or top 2 or top 6) performers was done according to the following steps:
Step 1 Calculate a simple average index for each country within a REC (each
country within Africa or each REC within Africa). Let this average index be
defined by
X
T
I ij, t
I ij ¼ ð9:3Þ
t¼1
T
Where Īij is the average index for indicator i, country j over t ¼ 1,2,. . .,T time
periods. That is, a single value is calculated for each country for a particular
indicator over the time periods (years) where data has been completed.
Step 2 The second step involved ordering/sorting the average indices for all
countries within a REC in a descending or ascending order. Suppose the average
indices of countries are ordered in a descending (decreasing) order. And suppose
a REC has more than six members.
That is, let the average indices be Ī1, Ī2,. . .,Īc, where c > 6.
Given these ordered indices, calculating the average of the top performers
(in this case average of top 4 performers) depends on the particular indicator’s
contribution towards Regional Integration.
If an increase in an indicator contributes positively to Regional Integration (for
example, exports and imports), then the norm/yardstick is given by the average of
the first four indices in this particular case. That is,
X
4
Ii
b¼ ð9:4Þ
i¼1
4
Xc
Ij
b¼ ð9:5Þ
j¼c4
4
It should be noted that the key role of these yardsticks is to help establish
intervals around them upon which scores can be assigned to countries within a
REC to measure their relative performance. The construction of such intervals and
hence scoring and ranking performance are accomplished as follows. It should be
noted that the concept of using the average of the top (2 or 4) performers is also
important in terms of minimizing some extreme values or outliers in the table.
Once the b-value is calculated, the standard deviation of the indices of countries
within a REC (or countries within Africa and RECs within Africa) is generated in
order to construct intervals or borders around the given b-value. The standard
deviation measures the spread of the performances of countries or REC around
the average performance within a given set of indices. The standard deviation,
which is the square root of variance, is defined as:
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 2
u n
uX I k I
s¼t
k¼1
n1
Where Ī is the mean of all indices within a REC (within Africa or for all RECs), Ik is
the kth index, and n represents the total number of indices within a REC.
Note that different subscripts are used to simplify the burden of having too many
subscripts.
It is important to note that, in a standard statistical sense, the standard deviation
is used to construct an interval around the mean/average. In other words,
constructing intervals around the mean, in this case, would mean comparing the
performance of countries against the average performers. This, however, is incon-
sistent with our definition of “best performance”. Thus, instead of using the mean,
this comparison employs the b-value as a point of reference for best performance,
and therefore intervals/borders are constructed around this value.
Thus, boundaries/intervals are defined around the b-value with a radius of 1/5 X
standard deviation (1/5 X s), meaning the length of the interval will be 2/5 X
standard deviation (2/5 X s). A total number of 11 distinct intervals are constructed
for each indicator within a REC for which scores are assigned depending on the
interval.
218 D. Tanoe
Scores are assigned for each index for all time periods, depending on where that
particular index lies within the given intervals. A maximum score of 10 is assigned
for the best performance and zero for the last (lowest) performer(s).
Then, these scores are averaged over the given time period (1994–1999) to
obtain an average score for each country (or for each REC). The first ARIA assessed
progress between 1994 and 1999. Subsequent ARIA editions discontinued.
Finally, a ranking is given to each country or REC based on the average scores,
where a rank of number 1 is given to the best performer(s) and the least performer
(s) become the last in the rank.
On the basis of the scores obtained for each REC over time, averages of scores for
each year using the REC level indices over the number of RECs were calculated to
obtain the average score/performance of the RECs in the various years. An index of
the scores (with base year 1994 ¼ 100) is then calculated on a year-by-year basis
showing the change in performance over the given period of time for the RECs,
which is known as the Composite Integration Index (CII).
The Weighted Composite Integration Index (WCII) is the total sum of the average
REC index multiplied by the corresponding GDP weight of that REC. The com-
posite integration index measures relative performance of a REC or RECs within
the continent, but does not take into account the size of that particular REC in
relation to the others. This calls for some sort of weighting procedure in order to
reflect the REC’s overall performance, in which the GDP weighted composite
integration index has been obtained.
The regional integration indicators were constructed on the basis of data collected
from member states, the secretariats of the RECs and other regional and interna-
tional organizations. Detailed questionnaires covered the eight clusters indicated
1
As revealed through the indicators at time of release of ARIA 1 in 2004.
9 Assessing Regional Integration in Africa (ARIA): Indicators of Integration. . . 219
above. The questionnaires were designed to solicit both quantitative and qualitative
information, and they were supplemented by data collection missions to the RECs
and a selected number of countries. Sectoral integration indices were calculated as
weighted composites of sectoral indicators—chosen to reflect the intensity or
impact of regional integration in each sector. The eight sectoral indices were used
to obtain composite integration indices for the regional economic communities and
for all of Africa. Progress by the RECs during the period covered (1994–1999) was
estimated as a weighted measure of performance in the eight clusters. Where a
sectoral indicator was constructed from several other indicators, the trend in that
sector was calculated as a weighted average of the subindicators. For example, the
money and finance indicator was a weighted average of inflation rate, external debt,
investment, and budget deficit. Thus, a single weighted composite index for the
RECs was developed as a single time series, with the base year value taken to be
100. The composite index for Africa was an average (weighted by GDP) of the
integration indices for the RECs. It measured the continent’s total integration effort,
assessing progress towards the integration goals of the RECs, the Abuja Treaty, and
other continental and regional integration initiatives and policies.
In some cases the base year levels and scores were low (in trade, for example),
which tended to amplify progress in the following years. Thus the indices often
showed an initial spurt in performance. Annual changes in the index measured
progress or retrogression and allowed comparison among the RECs. Tables 9.1 and
9.2 show the general trends of progress at the REC and sectoral levels respectively.
Table 9.1 Integration indices for Africa’s regional economic communities, 1995–1999 (Index
1994 ¼ 100)
Regional economic community 1995 1996 1997 1998 1999
CEMAC 129.7 135.7 136.0 134.8 128.4
CEN-SADa 122.9 130.8 133.7 121.2 121.0
CEPGL 90.6 89.5 93.7 91.2 86.6
COMESA 110.1 123.0 125.2 127.2 119.4
EAC 114.7 120.3 118.5 120.5 119.2
ECCAS 124.6 128.1 132.0 126.8 121.7
ECOWAS 117.2 130.8 130.3 136.6 133.9
IGAD 113.0 114.1 120.8 119.8 119.7
IOC 116.2 126.2 118.3 123.8 109.6
MRU 90.2 96.4 119.3 109.3 117.1
SADC 115.6 131.5 131.0 137.2 136.9
UEMOA 117.4 132.3 133.4 138.6 137.1
UMA 101.4 100.4 101.3 99.5 100.4
Simple average 112.6 119.9 122.6 122.0 119.3
Weighted average 114.9 124.7 126.1 125.5 123.6
Source: Economic Commission for Africa, from official sources
Note: Given the significant component of the trade sector in the calculation of the indices, SACU
was excluded from this table. SACU’s published trade data were usually aggregated and could not
be used for the calculations
a
CEN-SAD had recently been established, hence its results reflected primarily actions of members
participating in overlapping RECs
220 D. Tanoe
Through the first edition of ARIA published in 2004, we wanted to start a process
that would evolve over the years just as the indicators used in the United Nations
Development Programme (UNDP)’s Human Development Report (HDR) and the
World Bank’s World Development Report (WDR). The ARIA assessment of
progress in integration using both the quantitative and qualitative analysis focused
on the treaties, goals and the objectives of the RECs. Areas of particular signifi-
cance to regional integration were encapsulated in the 8 cluster areas referred to
earlier. Data to buttress the assessment both on a quantitative and qualitative basis
was fundamental. However, in the African context, some of this data is not easy to
come by, either through questionnaire or from conventional sources. Nonetheless,
these indicators provide an important basis for refinements over time. The concept
of regional integration indicators is not an easy one to grapple with. But these small
steps, we hope, will one day cumulate into an authoritative body of knowledge for
use by stakeholders in this complex field of regional integration indicators.
The 6th Joint Annual Meetings of the ECA Conference of African Ministers of
Finance, Planning and Economic Development and the African Union Conference
of Ministers of Economy and Finance, held in Abidjan, Côte d’Ivoire, in April
2013, and the 6th Conference of African Ministers in charge of Integration, held in
Port Louis, Mauritius, in May 2013, have called for enhanced monitoring and
evaluation of Africa’s integration. Consequently, ECA and the African Union
Commission are currently in the process of developing an African regional inte-
gration index, a proposal on which was presented to and endorsed by the 7th Joint
Annual Meetings of the ECA Conference of African Ministers of Finance, Planning
and Economic Development and AU Conference of Ministers of Economy and
222 D. Tanoe
Finance held in March 2014 in Abuja, Nigeria. The Conference urged the two
institutions to continue working on the project. Drawing from the ARIA experience,
work is currently underway to finalize the methodology for this renewed African
Integration Index initiative, and proceed with the collection of data. Stay tuned.
Reference
UNECA. (2004). Assessing Regional Integration in Africa: ECA Policy Research Report, Addis-
Ababa.
Chapter 10
Monitoring Regional Integration
in the African, Caribbean and Pacific (ACP)
Regions
Jean-Michel Salmon
The ACP Monitoring Regional Integration (MRI) Project, managed by the ACP
Secretariat under a European Development Fund (EDF) grant, was launched in late
2008 for a period of 20 months (November 2008–June 2010).
The main purpose of this project was to kick-start a collaborative initiative
amongst a series of ACP Regional Integration Organizations (RIOs). This initiative
aimed at developing a common system of regional integration (RI) indicators with a
view to making a common tracking of progress made against the RIOs’ objectives.
This aim should be seen in the context of the mutual accountability principle
derived from the 2005 Paris Declaration on Aid Effectiveness.
The present chapter gives an overview of the project institutional context, the
project governance, conceptual choices made, the system technical contents and,
finally, the project results.
Disclaimer: the views expressed in this chapter are those of the author and do not necessarily reflect
neither the views of the ACP Secretariat nor the views of the ACP RIOs cited in the text. The analysis
is based on the several reports from the ACP MRI downloadable from the project website (http://
mri.acp.int), inter alia the Baseline Study Report (final version), the Field Missions Overall Report
and the Second Seminar Final Report. Hence the information contained strictly relates itself to
the Project period and does not include any further development possibly made by any ACP RIO since
mid 2010.
J.-M. Salmon (*)
Faculty of Law and Economics, Université des Antilles, Schoelcher, Martinique
STRADEVCO, Fort de France, Martinique (French West Indies), France
e-mail: stradevco@wanadoo.fr
The Cotonou Agreement between the ACP Countries and the EU, signed in June
2000 for a period of 20 years, emphasizes regional economic integration and
specifically the role of ACP RIOs.1
Article 9 of Annex IV reads
« At the beginning of the period covered by the Financial Protocol, each region shall
receive from the Community an indication of the volume of resources from which it may
benefit during a five-year period. The indicative resource allocation shall be based on an
estimate of need and the progress and prospects in the process of regional cooperation and
integration. »
In the early 2000s, data on RI in ACP countries was very limited in terms of
indicators, especially quantitative ones. A possible exception to this was trade data.
Therefore most assessment/monitoring exercises of RI processes in ACP regions
were related to case study reports, based on some rather qualitative analysis/
judgments made, together with the little quantitative data available. One could
also find synthesis papers prepared by ACP RI experts (e.g. Dinka and Kennes
2007). Within most publications, authors tracked fundamentals of RIOs such as
their membership, objectives, instruments and achievements (see for example
Lyakurwa et al. 1997). Thus, they were focusing more on the qualitative aspect
of particular RI processes, rather than on the state or degree of RI, which in any case
was never measured.
Against this backdrop, the European Commission (DG Development) resolutely
decided to support the development of systematic RI indicators for the ACP regions
1
Articles 28–30 of the Agreement and articles 6–14 of Annex IV.
10 Monitoring Regional Integration in the African, Caribbean and Pacific (ACP). . . 225
2
Earlier on, the UNECA (UNECA 2004) had launched a pioneering system of RI indicators in
Africa (see previous chapter of the present book). This was not without its own critics in some
ACP RIOs, notably as expressed during the above-mentioned workshop. In particular the
COMESA Secretariat regretted the lack of consideration of RI policy effort in the system proposed
by UNECA, with actually 17 selected indicators (among 19) being outcome or impact indicators
(as opposed to input or output indicators), a majority of which corresponding to purely national
data. COMESA also emphasized the need to take into account diverging RIOs agendas, and the
difficulties with the RIOs ranking process as proposed by UNECA.
3
These are trade liberalization, trade facilitation, trade in services, transit facilitation, monetary
convergence, domestic payments and settlements systems, fiscal environment, government inter-
vention in the economy, capital flows and foreign investment, governance issues, regulatory
environment, licensing requirements.
226 J.-M. Salmon
During 2007 the ACP Secretariat took the initiative to restart a consultative process
amongst ACP RIOs for the building of a RI indicators system. The Secretariat
received the financial support of the EU to launch the ‘ACP Monitoring Regional
Integration’ project for that purpose, becoming the project contracting authority.
It is worth noting that under the ACP Secretariat organizational chart, the project
was reinitiated and managed not by the Sustainable Economic Development and
Trade Department, but by the Department of Political Affairs and Human Devel-
opment, in charge, inter alia, of relationships with international organizations.4 The
project was finally developed with the collaboration of a Consultancy and two
senior experts.
Twelve ACP RIOs finally involved themselves in the project (see Table 10.1).
Several international organizations (African Development Bank, African Union,
Table 10.1 Regional organizations involved and the number of related member states
ACP region Regional organizations Number of MSa
Sahel-Saharan Africa CEN-SAD 25 ACPb
Southern Africa SADC 15 (including 7 COMESA)
Central Africa CEMAC 6
ECCAS 10 (including 6 CEMAC)
Western Africa ECOWAS 15 (including 8 UEMOA)
UEMOA 8
Eastern and Southern Africa COMESA 19 (including 7 SADC)
EAC 5
IGAD 7
IOC 4 ACPc
Caribbean CARICOM 15 plus Dominican Republic
Pacific Pacific Forum 14 CPd
Source: MRI project documents (as quoted in Disclaimer), namely Baseline Study Report
a
The number of MS was calculated from a visit on the website of each of these organizations, as
from November 2008
b
Libya, Morocco and Tunisia are also MS of CEN-SAD, but we have only included the ACP MS in
the group
c
France is also a member of the IOC on behalf of Reunion Island, but again we have only counted
ACP MS in this group
d
Australia and New Zealand are also members of the Pacific Forum, but we have only included the
ACP MS, which are also part of the regional PICTA free-trade agreement
4
Therefore the project thematic scope would become relatively large and not limited to trade and
economic issues, i.e. to regional economic integration only.
10 Monitoring Regional Integration in the African, Caribbean and Pacific (ACP). . . 227
The main purpose of the ACP MRI project was to harness the recent proliferation of
approaches to RI indicators so as to develop and implement a coordinated system of
RI indicators for ACP regions.
From the project Terms of Reference (ToRs) and the inception phase, it was
indeed clear that the RI indicators system to be built should be considering RI as a
multidimensional process, so that the system would not be focused only on eco-
nomic and trade integration. The expected ACP MRI system would rather include
as many thematic domains as necessary, following a ‘holistic’ approach.
The system would also be able to measure both the degree (or state) of RI and the
RI policy effort: this led to the choice of using the results chain analysis categories.
Again as requested by the project ToRs, the system shall also allow both for
inter-regional comparisons of ACP RI experiences (comparative approach) and for
the monitoring of each individual ACP RI experience vis-a-vis its own agenda
(reflexive approach).
The issue of the future involvement of the ACP RIOs involved in the project and
their MS was a fundamental one: the implementation of the system would ulti-
mately have to be backed by them, both at the political and at the technical level
(including data production and collection).
Of particular importance was therefore the ongoing identification process by
these RIOs and their MS of the interests they would find and serve through the use
of this monitoring system, as well as the support they would gather in order to
supply its needs, inter alia, in terms of data requirements and data storage.5
In other words, the end-use of the project and of the monitoring system itself was
one of decisive importance. At the project inception, some ACP regions were
considering the project as a possible example of an overly ‘top down’ exercise
“coming from Brussels”, which “we did not ask for” and which will “provide a
surveillance tool for our main donors”, in particular in the context of the hotly-
negotiated EPA agreements.
5
This holds especially true in face of the existing gap between RIOs’ ambitious policy agendas and
the limited means they can access to implement these, a gap actually growing with the so-called
“deepening” of the RI process, notwithstanding the donors’ increasing will to further support this
process.
228 J.-M. Salmon
However, with some close dialogue at the field level with many RIO senior
administrators, it was then considered that, in the context where all ACP RIOs were
committed to a process of building their own monitoring tool, the project outputs
could be helping a lot, including in terms of exchanges of best practices. So the
support earned by the project kept growing from its inception phase.
Finally, it was largely understood that the end-use of the expected system was
simply to monitor RI, and not to evaluate RIOs and their projects/programs or even
less to pinpoint the least effective ACP institutions or actors. In other words, this
ACP MRI system was being regarded as an operational tool with a view to
providing for a more evidence-based decision-making process at many levels,
i.e. national, regional, continental, and even for the whole ACP, while remaining
as depoliticized as possible. With this tool made readily accessible, any stakeholder
at any level would be able to find easily the required standardized information on
any ACP RI aspect and process he/she is interested in.
Given its need for comprehensive, holistic and thematic coverage, it was further
proposed and agreed on that the ACP MRI system shall:
(a) be embedded in the Cotonou Partnership Agreement vision of RI,
(b) put into practice the recommendations of the ACP-EU Joint Parliamentary
Assembly – especially the ones included in its ‘Resolution on experiences
10 Monitoring Regional Integration in the African, Caribbean and Pacific (ACP). . . 229
From the reviewing exercises as well as from the six essential methodological
principles set forth above, an initial proposal for the ACP MRI system architecture
was developed under a Three-Level Tree Structure, distinguishing the system
“dimensions” (at the upper level) from the system “domains” (at the intermediate
level) and the system “areas” (at the lower level). Individual indicators would be
positioned (clustered) in their respective area. This structure remained unchanged
throughout the large consultation process, even if its contents were fine tuned.
The dimensions finally proposed are Regional Governance, Economic Integra-
tion, Functional Cooperation and Social Integration. They were selected in accor-
dance to their relevance and the methodological principles set forth above, after
having considered the different choices made by other experts with similar objec-
tives. Also taken into account were the issues, objectives and priorities earmarked
by the European Commission in its Communication on RI for Development in ACP
Countries.
Each dimension was then disaggregated in two or three domains, according to
identified analytical needs and to official jargons, while keeping in mind the
objective of a balanced architecture. The same applies also to each domain,
which was in turn disaggregated into two or three areas.
To illustrate, the “Economic integration” dimension further disaggregates itself
into three domains, namely “Trade integration”, “Monetary and Financial integra-
tion” and “Production and economic cohesion”.
10 Monitoring Regional Integration in the African, Caribbean and Pacific (ACP). . . 231
Then, the “trade integration“ domain further develops into three areas, namely
“goods trade”, “services trade” and “trade-related areas”.6
As for evaluating both the RI degree and the RI policy effort, it was proposed and
decided to follow the categories of the results chain analysis, namely the input-
output-outcome-impact typology.7 This acknowledges that the RI policy effort is
captured by input and output indicators, while the RI degree is more related to
outcome and impact indicators.
The next step was systematically to identify individual indicators with a view,
whenever possible, to representing fully the results chain typology at each level of
the architecture. This allows for tracking the RI policy efforts undertaken by ACP
RIOs vis-a-vis their respective own policy agenda, as well as the results to which
they arrive, in a reflexive and/or a comparative manner.
The envisaged CSRI includes a rather limited number of core indicators, while still
allowing for a relevant representation of the different levels of the system archi-
tecture. These two opposite requests lead to a CSRI made of a range of 50 core
indicators, meaning simple averages (but not a strict constraint) of two indicators
per area and a dozen per dimension.
As for the selection of these 50 core indicators, in addition to striking a balance
in terms of both the system architecture and the results chain typology (see Annex
A.10.3), the following criteria were applied to around 200 candidate indicators:
Relevance, Quantifiability, Simplicity, Availability & Cost.
After consultation of stakeholders, the number of core indicators was raised to
75, i.e. an average of 3 per area (see Annex A.10.2 for a full list).
6
Following the official jargon of trade agreements the latter corresponds to issues such as trade
facilitation, competition, procurement, TRIPS, etc. See Annex A.10.1 for the full list of domains
and areas under each dimension.
7
Which are the key terms in evaluation and results according to the OECD/DAC (see OECD
2002).
232 J.-M. Salmon
Regional
Governance
100
80
60
40
20
Economic
Social Integration 0 Integration
Region A
Functional
Region B Cooperation
Chart 10.1 The Regional Integration “Diamond” (by RI dimensions) (Source: MRI project
documents (as quoted in Disclaimer), namely Baseline Study Report)
The next conceptual and analytical steps were to develop synthesis and comparative
tools of ACP RI from the core indicators of the CSRI. This would lead to the
designing of RI composite indexes and related tables and graphs that could be later
on included in a regular report on ACP regional integration.
The envisaged composite indexes were to be elaborated at the dimension level,
i.e. representing each of the four dimensions (Regional Governance, Economic
Integration, Functional Cooperation, Social integration and regional stability),
leading to the designing of four composite indexes to be graphically represented
through a RI ‘Diamond’ following Ruiz Estrada (2004), as illustrated by Chart 10.1.
Each ‘dimensional’ composite index could be computed from a sample of core
indicators belonging to the dimension in question or alternatively from composite
indexes designed at the domain level, as an intermediary figure: the indexes at the
dimension level would then be computed from the indexes developed at the domain
level (e.g. as weighted averages). The latter approach presents the advantage of
further detailing the comparative analysis of ACP RI experiences, with some more
disaggregated representations as in the RI ‘Radar’ illustrated in Chart 10.2.
The final touch of this proposed approach based on composite indexes would be
to design an overall RI composite index from the four ‘dimensional’
composite ones.
However, making available a single RI composite index is challenging, for
several reasons.
10 Monitoring Regional Integration in the African, Caribbean and Pacific (ACP). . . 233
Community
resources
100 Institutional
Regionhood coordination and
80 organisation
60
Human Collective security
Development 40 and democracy
20
Other domains of 0
Trade integration
cooperation
Financial and
Communication
Monetary
and Energy
Production and Integration
Transport economic cohesion
Region A
Region B
Chart 10.2 The Regional Integration ‘Radar’ (by RI domains) (Source: MRI project documents
(as quoted in Disclaimer), namely Baseline Study Report)
8
This is the approach followed within the UN system, when it comes to the assessment of the MS
economic vulnerability, measured through the Economic Vulnerability Index (EVI), which is a
composite index used by the ECOSOC Committee of Development Policies as a criteria among
others (in relation to benchmarks) to establish every five years the proposed next list of LDCs to
the UNGA. But before a final decision is made with respect to a MS inclusion/graduation as a
result of these quantitative criteria, a qualitative profile of its vulnerability, prepared by UNCTAD,
is consulted. To illustrate the relevance of this approach, just recall that this is how the Maldives
avoided graduation in the early 2000s, not without strong criticisms by then, before having to face
the consequences of the Tsunami in 2004.
234 J.-M. Salmon
As for weighting procedures to build the composite indexes at all levels, they
were yet to be specified in the ACP MRI project.
The ACP MRI project in the end could not provide for the implementation of the
CSRI at field level, mainly through lack of time – i.e. too short a project duration.
While the initial project duration of 24 months had to be reduced to 20 months in
relation to EDF management constraints, it was initially considered that this
duration could provide for both the indicators system conceptualization (the base-
line phase) and its implementation (the second phase).
However the large number of ACP RIOs to coordinate in the project proved very
time-consuming, as did the close consultations at field level which nevertheless
were indispensable to ensure participation and ownership of the system by
these RIOs.
A subsequent phase would have been necessary for the system to materialize
with the related data collection, treatment and storage, but it could not yet get off
the ground.
10.7 Conclusion
Whatever the merits of the ACP MRI project, and they were many, its experience
demonstrates once again that, when it comes to monitoring indicators systems more
generally, the system size (and cost) should always be considered given owner’s
capacity, the system end-use and its ownership by main stakeholders. For only
positive answers to these crucial aspects will ensure some materialization and
above all sustainability to the desired system.
10 Monitoring Regional Integration in the African, Caribbean and Pacific (ACP). . . 235
Annexes
Number Number
Number Number of of
Number Number Number of of specific general
of Number of of input output outcome Impact Impact
Dimensions domains of Areas indicators indicator indicator indicator indicator indicator
Regional 3 7 20 4 6 8 2
governance
Economic 3 7 21 0 8 3 5 5
integration
Functional 3 9 20 1 12 0 4 3
cooperation
Social 2 4 14 0 8 1 2 3
integration
TOTAL 11 27 75 5 34 12 11 13
Source: Document (CSRI Methodological Guide)
References
11.1 Introduction
This chapter presents and assesses the EAC Common Market Scorecard (CMS)
which was published by the World Bank Group (WBG) in 2014, in collaboration
with the EAC Secretariat (World Bank/EAC 2014).
The East African Community (henceforth EAC) itself was established in 1999
by Kenya, Tanzania and Uganda after signing the EAC Treaty.1 Its objectives are to
deepen cooperation among member states in political, economic, and social fields.
They include the ambitious goals of establishing a customs union (2005), a com-
mon market (2010), a monetary union and ultimately even a political federation of
East African Partner States. Burundi and Rwanda joined the EAC later than the
others (in 2007), and joined the customs union in 2009.
A fairly classical economic integration process is thus followed. ‘Deep integra-
tion’ is pursued, requiring trade liberalisation, followed by trade facilitation through
1
More precisely, the EAC was revived after it had already existed between 1967 and 1977. The
treaty was signed in 1999 and entered into force in 2000.
A. Ombudo K’Ombudo
EAC Common Market Scorecard 2014 Project, Nairobi, Kenya
Advocate of the High Court of Kenya, Nairobi, Kenya
US International University, Nairobi, Kenya
P. De Lombaerde (*)
NEOMA Business School, Rouen, France
UNU-CRIS, Bruges, Belgium
e-mail: philippe.de-lombaerde@neoma-bs.fr
M. Borda
UNU-CRIS, Bruges, Belgium
2
On the implementation challenges for the regional harmonization of commercial laws, see also
Agaba (2011).
3
An EAC Metrology Bill was mooted in 2009.
11 The East African Community Common Market Scorecard 241
based tracking of the progress towards achieving a common market and contributes
to maintaining attention and resources towards the key success factors necessary for
achievement of a better investment climate in a common market arrangement.
Taking into account the mandate of the WB Investment Climate Advisory
Services (now the WB Trade and Competitiveness Practice), the focus of the
monitoring exercise was on business-related policy implementation issues. The
11 The East African Community Common Market Scorecard 243
purpose and scope of the EAC Common Market Scorecard are thus more precise
and limited than most of the proposed indicator systems so far. The most relevant
reference points seem to be the European Commission’s Internal Market Score-
board4 and ASEAN’s Economic Community Scorecard.5
4
The endorsement of the Action Plan for the Single Market by the European Council of Amster-
dam of 17 June 1997 led to the creation of the ‘Single Market Scoreboard’, first published in
November 1997. The European Council emphasized “the crucial importance of timely and correct
transposition of all agreed legislation into national law; the need to fully inform citizens and
business about the Single Market and the need for active enforcement of Single Market rules in the
Member States” (European Commission, 1997: 1). DG Internal Market and Services developed
quantitative and qualitative methodologies to assess (i) the transposition of Internal Market
directives into national law, and (ii) the number of infringement proceedings initiated by the
Commission against the member States. The Internal Market Scoreboard (IMS) is published twice
a year since then. See: http://ec.europa.eu/internal_market/score/index_en.htm. Interestingly, the
methodology of the IMS has also been adapted to be used by the EFTA member states that signed
the EEA Agreement. With the ‘Internal Market Scoreboard – EFTA States’ (IMS-EFTA) the
EFTA Surveillance Authority aims at measuring the effectiveness of the Internal Market rules that
are part of the EEA Agreement and encouraging the transposition of the Internal Market directives
in a timely manner. In addition, the IMS-EFTA contains information on the infringement pro-
ceedings commenced by the EFTA Surveillance Authority against the EFTA States with the
objective to ensure correct enforcement of the rules. The IMS-EFTA, together with the Interim
Report on Transposition Status of Directives, has been published since May 1998 and is published
twice a year (Costea et al., 2008). For further details and updated information, see Chap. 1 in
this book.
5
ASEAN adopted the ASEAN Economic Community Blueprint in Nov. 2007, outlining the
measures to be taken (and the time frame for implementation) in order to establish a competitive
single market by 2015. The first Scorecard was published in March 2010 and covered the period
2008–2009. See: http://www.asean.org/publications/AEC%20Scorecard.pdf. It should be
observed, however, that the scope of the AEC Scorecard goes beyond the ‘single market’; it also
covers other dimensions: ‘competitive economic region’, ‘equitable economic development’, and
‘integration into the global economy’. See also, Chap. 13 in this book.
244 A. Ombudo K’Ombudo et al.
6
Ideally, EAC Partner States should notify the EAC Secretariat and other EAC Partner States
about potential measures that are under consideration and that are likely to affect the enjoyment of
rights under the protocol by other Partner States. However, currently there exists no infrastructure
or clear policy for a system of notifications. The scorecard partially addressed this problem, by
outlining key measures that the Partner States had taken that were inconsistent to their obligations
to the EAC Common Market Protocol. However, this only provided information as regards
existing measures, but not (perhaps the more important) ex-ante measures i.e. legal, regulatory,
administrative and policy proposals that have not yet been promulgated. The latter are more useful,
given that they are designed to allow other EAC Partner States to exercise their right of advance
comment on potential common market-limiting measures.
7
UNECA has relevant expertise in monitoring category III variables, built-up in the framework of
its ARIA Reports and indicator system (UNECA 2002, 2003, 2004). See, Chap. 9 in this book.
Also of relevance is the recent ACP Monitoring Regional Integration project, involving the design
of an indicator system for monitoring the regional integration processes in the various ACP
subregions. See: http://mri.acp.int/spip.php?page¼indexen. See, Chap. 10 in this book.
11 The East African Community Common Market Scorecard 245
The ambition of the monitoring exercise was consciously limited to these four
categories, which are all directly measurable in some way or another. It was
considered as being beyond the scope of a monitoring exercise to establish the
causal link between policy and target variables, i.e. to establish ‘results’ and
‘impacts’. Results and impacts (category V) cannot be directly measured but
require statistical analysis and econometric estimation, which is better kept outside
the monitoring system.8 The role of monitoring is, precisely, to provide all stake-
holders with relevant data in an impartial and transparent manner. The interpreta-
tion and further (statistical) analysis of these data are tasks for the interested
stakeholders. This is also the approach used in the Internal Market Scoreboard
(see above) and the AEC Scorecard. The latter is a compliance tool, not an
instrument for impact assessment (Rillo 2011).
A step-wise approach was followed by concentrating first on Category I and II
(where the mandates of EAC Secretariat and WBG coincide), which can then be
expanded by the Secretariat in the future to include category III. WBG could further
play a role with respect to category IV by adapting its business climate question-
naires for this purpose.9
A consensus view emerged among the stakeholders that the monitoring system
should not be limited to the regulations and measures not implemented by the
deadlines that can be found in the Protocol and its Annexes (i.e. following the
European model), but that it should rather be designed to accompany the whole
Common Market implementation process, in line with art. 50. As seen in other
regions, monitoring does not necessarily end after the provisions of an agreement
have been practically fully implemented: the concerns of private sector actors
trading and investing under the terms of the agreement will still need to be attended
and any further commitments that the partner countries may have assumed during
the implementation process will have to be monitored, including directives by EAC
Council of Ministers, on an ongoing basis (De Lombaerde et al. 2008b). The
ambition was therefore to develop an instrument which continues to be useful
beyond the specific short-term deadlines.
8
See e.g. De Lombaerde and Van Langenhove (2010: 2–29).
9
The World Bank Group conducts a number of business questionnaires on business climate issues.
Enterprise Surveys (www.enterprisesurveys.org) are firm-level surveys of a representative sample
of an economy’s private sector. The surveys cover a broad range of business environment topics
including access to finance, corruption, infrastructure, crime, competition, and performance
measures. Doing Business Reports (www.doingbusiness.org) provide measures of business regu-
lations for local firms in (currently) 183 economies and selected cities. The Investing Across
Borders Report (www.iab.worldbank.org) is a World Bank Group initiative comparing regulation
of foreign direct investment around the world. It presents quantitative indicators on economies’
laws, regulations, and practices affecting how foreign companies invest across sectors, start
businesses, access industrial land, and arbitrate commercial disputes. According to the 2016
edition of the Ease of Doing Business Report, performance across the EAC economies varies:
Rwanda is ranked 62nd globally, followed by Tanzania (131st), Kenya (136th), Uganda (150th),
and Burundi (152nd), on a total of 183 economies covered by the Report.
246 A. Ombudo K’Ombudo et al.
10
An on-line reporting tool has been designed with support from TradeMark and GIZ.
11 The East African Community Common Market Scorecard 247
11
In the case of the ASEAN Economic Community, benchmarks are found in the AEC Blueprint
and in the AEC Strategic Schedule (Rillo 2011).
12
http://www.doingbusiness.org/data/distance-to-frontier
248 A. Ombudo K’Ombudo et al.
Table 11.1 Original implementation schedule of regulations related to free movement of workers
(number of occupational titles covered by regulations to be implemented per year)
2010 2011 2012 2013 2014 2015
Burundi 7
Rwanda 5
Tanzania 5 2 2 5
Uganda 4
Source: The East African Community Common Market (Free Movement of Workers) Regulations
– Annex II, EAC Secretariat, 2009
• Data representing the reference “community output,” which include the EAC
Common Market Protocol, and legal acts collected at community source.
• Data representing action taken at EAC Partner State level in pursuit of imple-
mentation of the community objectives: it comprises information about national
laws transposing community law or decisions or approximating national law to
EAC law. Most of this information is mined through desk research, expert panels
and through a verification process coordinated through the MEACs.
The scorecard team initially expressed the intention to construct a series of
indices on the basis of the following information:
(a) Index on Completeness of the EAC Common Market Protocol. This index
would measure the extent to which the EAC Common Market offers action-
able guidance for the implementation of the EAC Common Market Protocol. It
would measure which parts of the protocol have schedules and annexes, and
where the protocol provides that further instructions will be offered through
agreement by partner states, whether there is evidence that such agreements
250 A. Ombudo K’Ombudo et al.
have been reached. It would aggregate information on the four freedoms and
two rights.
(b) Index on Directives of the Council of Ministers. This index would measure the
extent to which decisions of the EAC Council of Ministers are focused on
delivering the provisions of the Common Market Protocol. In other words, this
is a measurement of the work rate of the Council of Ministers towards
delivering actionable instructions to the EAC Partner States and the EAC
Secretariat to deliver on the EAC Common Market Protocol.
(c) Index on Acts of the East African Legislative Assembly (EALA). This index
would measure the extent to which decisions of the East African Legislative
Assembly are focused on delivering the provisions of the Common Market
Protocol. In other words, this is a measurement of the work rate of EALA
towards delivering actionable instructions to the EAC Partner States and the
EAC Secretariat to deliver on the EAC Common Market Protocol.
The rest of the indices would measure the level of compliance by each EAC
Partner State towards commitments made to allow the free movement of goods in
the EAC. A composite score for the EAC for each index would also be provided. All
agreements and commitments that provide actionable instructions from the Com-
mon Market Protocol, EAC Council of Ministers, EALA, and EACJ would be
covered.
(d) Index on Free Movement of Goods. This index would cover issues contained in
the Protocol on the Establishment of the East African Community Customs
Union.
(e) Index on Free Movement of Persons and Labour. This index would cover: the
free movement of persons; standard identification systems13; travel documents;
free movement of workers; harmonization and mutual recognition of academic
and professional qualifications; and harmonization of labour policies, laws and
programs.
(f) Index on Rights of Establishment and Residence. This index would cover the
right of establishment; right of residence; and access to and use of land and
premises.
(g) Index on Free Movement of Services. This index would cover free movement of
services; national treatment; and the most favoured nation principle.
(h) Index on Free Movement of Capital. This index would cover: elimination of
restrictions on operations related to: equity and portfolio investments; bonds
and other debt instruments; money market instruments; collective investment
schemes; bank transactions; direct investments; repatriation of proceeds from
sale of assets; and other transfers and payments relating to investment flows.
13
The EAC has a harmonized passport. In an effort to encourage ease of movement, Kenya and
Rwanda made it acceptable to cross their mutual borders with only an identification card, and in
2010 the two countries entered into a reciprocal agreement waiving the work permit fee.
11 The East African Community Common Market Scorecard 251
For the first edition of the EAC CMS, these indices were finally not withheld.
The focus was restricted to de jure compliance by EAC member states, i.e. not de
facto implementation, and no aggregate indices were calculated.
The analysis for the 2014 Scorecard was based on the assessment of 683 regu-
lations and laws with a direct link to the implementation of the common market, as
well as other reports, legal notices, and also trade statistics. Of the 683 regulations
and laws, 14 are in the area of goods, 545 in the area of services, and 124 in the area
of capital (WB/EAC 2014: 34).
In the area of trade, focus was placed on compliance with four obligations:
(a) elimination of tariffs and equivalent measures on intra-regional trade,
(b) elimination of non-tariff barriers, (c) implementation of a common external
tariff, and (d) harmonization and mutual recognition of SPS standards and technical
standards (WB/EAC 2014: 38). In the area of services, the focus was on the key
obligations as listed in Part F of the Protocol (articles 16.1–7). Sectors covered
include: business services, communication services, distribution, education, finan-
cial services, tourism, and transport. In the area of capital, the 20 capital market
operations listed in the Protocol Schedule on the Removal of Restrictions have been
targeted.
Data sources included: (a) national laws and regulations, (b) implementation
reports of national authorities, (c) national official statistics, (d) EAC reports,
(e) international treaties, and (f) reports of international organizations (such as
IMF, World Bank Group and UNECA). The review process was conducted via
desk research in combination with expert opinion and stakeholder consultation
(including ministries, regulatory agencies, and professional associations). Expert
opinion was gathered by means of surveys addressed to lawyers, other professional
service providers, academics, business associations and other expert respondents
based in the five EAC Partner States, and selected on account of their expertise on
areas covered by the protocol.
All original data were treated confidentially. The pool of potential respondents
was based primarily on the following sources of information: (a) members of the
East African Business Council; (b) EAC professional associations focusing on
professional services liberalized for trade under the protocol; (c) trade and invest-
ment lawyers; and (d) firms that trade and invest in the EAC.
Conflicting responses were cured through counter-checks with desk research,
use of “best respondents” to clarify conflicting issues, and through verification/right
of reply missions to public authorities, coordinated by Ministries of the EAC. Being
categorised as “best respondent” requires that the individual is qualified specifically
in the legal issue contested, not only by having attained the academic qualification
but by being a practitioner in the field.
Confidential meetings were convened with Ministries of EAC after completing
the desk research and compiling the first draft of the EAC CMS. The objective was
to work with the ministries as a coordinating mechanism to check through the
information in the draft report with relevant line ministries.
The approach which has been followed in this monitoring process presents three
types of limitations: (a) substantive, focusing on the content and coverage of the
252 A. Ombudo K’Ombudo et al.
indicators; (b) methodological, concerned with the objectives of the desk research,
the respondent questionnaire design and data collection; and (c) limits to the
implications of the indicators, addressing their potential interpretation, uses, and
relationships with other scorecards or evaluation material covering the EAC com-
mon market process. These limitations should be kept in mind when interpreting
the data.
Substantive Limitations:
(a) Where the EAC CMS measures legal and regulatory compliance to the EAC
CMP, it does not take into account bilateral agreements between EAC Partner
States; for example, the agreement between Kenya and Rwanda that allows
citizens to travel with only national identity cards.
(b) EAC CMS does not consider specially regulated economic zones in determin-
ing compliance to the protocol. Therefore legal regimes for special economic
zones, export processing zones (EPZs), and other areas governed by special
legal frameworks are excluded from the scope of the project.
(c) If a right granted under the CMP is extended generally as part of the external
trade policy of the country to foreign nations, then this is counted as compli-
ance to the CMP i.e. the right or freedom does not have to be exclusive to EAC
countries.
Methodological Limitations:
(a) EAC CMS is not a survey of perceptions,14 but of legal facts and expert
responses collected through standardized questionnaires completed by a
small number of specialists in each EAC country.
(b) The data does not require to be based on a statistically significant sample of
respondents in each EAC Partner State. This survey is complemented with
thorough desk research, and an intensive consultation process, including a
verification process with authorities in the EAC countries.
(c) The EAC CMS does not seek to measure the ‘de facto’ application of laws and
regulations that facilitate the implementation of the protocol. The possibility
remains of a discrepancy between the passed laws and regulations and the
actual practice (non-tariff barriers). Thus, countries could score higher than
they would if a compliance process was in place.
Main Limitations of Interpreting the EAC CMS:
(a) EAC CMS does not provide an indication of the actual level of enjoyments of
rights and volumes of transactions in the common market, but only an indica-
tion of the legal and regulatory environment allowing these transactions to
14
Where checklists were shared with entrepreneurs and business people, the input sought was not
their perceptions, but they were rather contacted as a source of information, specifically as regards
administrative letters from public authorities requesting their compliance with various industry
regulations.
11 The East African Community Common Market Scorecard 253
happen. It should therefore not be assumed that increased indicator scores will
automatically lead to higher levels of trade and investment in the EAC.
(b) EAC CMS data should not be used as a proxy for government business reforms
in general. It only measures legal and regulatory compliance to the protocol,
and in so doing, facilitate policy dialogue by identifying good practices, track
reforms, facilitate sharing of reform experiences, and enable research and
analysis.
A Reference Group was proposed to provide stakeholder input into the process of
designing the scorecard. Working closely with the World Bank Group, the refer-
ence group:
• Advised on the scope of coverage and feedback mechanism for the EAC CMS.
• Reviewed and provided inputs into drafts generated by World Bank Group in
operational plan areas such as: (a) objectives, (b) methodology, (c) reporting
strategy, (d) communications, and (e) ownership and governance.
The reference group comprised persons drawn from EAC stakeholder institu-
tions, professionals with experience in development planning monitoring, experi-
enced practitioners in regional integration and other regional integration
communities that have done scorecards before. The stakeholder institutions
included World Bank Group, Ministries of EAC, EAC Secretariat, East African
Business Council, national institutions already implementing scorecards (nomi-
nated by MEACs where necessary), research and civil society institutions, private
sector bodies, United Nations Economic Commission for Africa (UNECA), and
United Nations University (UNU-CRIS). A secretariat function for the reference
group was offered by the World Bank Group.
The Scorecard was published by the World Bank Group, and was handed over to
the EAC Secretary General and staff, and to Ministers responsible for East African
Community Affairs in each of the EAC Partner States, much in the style of the EAC
Doing Business Report. The EAC Partner States have adopted the results of the
scorecard as part of their official monitoring tools and the EAC Council of Minis-
ters have directed Partner States to implement the recommendations of the
scorecard.
In the elaboration process of the first edition of the Scorecard it was also
observed that the obligation for each member state to regularly inform the EAC
Council of any new laws and administrative guidelines which affect trade in
services and free movement of capital had not been commonly respected.
Establishing an effective reporting mechanism would add transparency and facil-
itate monitoring of the member states’ progress towards liberalization.
254 A. Ombudo K’Ombudo et al.
The experience with other existing scorecards has shown the potential of this type
of monitoring instrument. For example, the regularly published transposition def-
icits in the Internal Market Scoreboard have become often cited figures in national
parliaments and the mass media. The IMS is effectively contributing to pressuring
member states to implement timely and correctly the regionally decided rules.
The most delicate issue for reporting and communicating the EAC CMS is not so
much technical, but political, as it concerns the question of the comparative perfor-
mances of the member states. There are different approaches that other scorecards
have used, and are a reflection of the regional integration environment and of cultural
differences. For example, the naming and shaming technique of the EC is not seen to
cause major embarrassment, because the emphasis was on emulation, and the failures
were both very technical and the gaps were expressed in a few percentage points. This
technique may pose some challenges to a much more recent economic union, where
variances may be markedly larger. The ASEAN scorecard emphasizes collective
responsibility and does not publish any list of good and bad performers.
It was the view of stakeholders that the EAC Common Market Scorecard should
express country-related information and draw comparisons from this information. It
was recommended, however, that the terminology adopted should be ‘soft’ when
describing national performance, for example the Lisbon process approach in the
EU emphasized successes, with lists, for each priority area, of “leaders” (or another
suitable designation), for partner states who have completed the transposition or the
implementation in a timely and accurate manner; it also lists “points to watch” and
“country recommendations”, to indicate partner state situations where results are
not yet up to expectations.
The 2014 scorecard, the first of its kind in EAC, covers the liberalization of trade in
goods, services, and capital flows, which are the three pillars of the Common
Market project (World Bank/EAC Secretariat 2014). Progress was registered by
all partner states, especially in goods. Nevertheless, much work remains in
guaranteeing the free movement of capital. In the following paragraphs a brief
summary of the key findings is presented.
The removal of restriction and prohibition from erecting new ones is governed by
the Customs Law of the Community (as specified in Article 39 of the Protocol on
11 The East African Community Common Market Scorecard 255
the Establishment of the East African Community Customs Union). Under Article 5
(2)(a) of the Protocol, Partner States are committed to eliminating tariff and
non-tariff barriers to trade, establishing a common external tariff, and harmonizing
and mutually recognizing certain trade standards. The review examined legal
obligations arising from the four commitments above, and entailed a review of
laws, regulations, legal notices and trade statistics relevant to the movement of
goods in the EAC.
Formally, all member states have eliminated tariffs on intraregional trade, but
measures with equivalent effects remain and constitute barriers to trade in all
member states. Non-tariff barriers (NTBs) reported in the region amount to
51 and are mostly related with sanitary and phytosanitary measures, rules of origin,
charges of equivalent effect to tariffs, and technical barriers to trade.
Though most member states, except Burundi, are in formal legal compliance
with the implementation of the Common External Tariff (CET), they all belong to
other free trade areas and, thus, apply different tariffs to extra-regional trade
partners, which conflicts with the spirit of the CET. The perforation of EAC’s
CET remains a serious structural problem, and an obstacle to achieving free
circulation of goods. This is because perforation discourages removal of internal
borders and complicates implementation of a common trade policy. Article 37 of
the EAC Customs Union Protocol recognizes the existence of other free trade
obligations of member states, but requires them to formulate a mechanism to
make these other free trade arrangements compatible with the protocol.
The 2014 edition of the Scorecard reviewed more than 500 key sectoral laws and
regulations of the EAC Partner States identifying at least 63 measures inconsistent
to commitments to liberalize services trade within the EAC (nonconforming mea-
sures, NCMs). The term “nonconforming measure” refers to specific provisions in
laws and regulations that negatively affect the liberalization commitments under-
taken by each member state in Annex V of the CMP. The goal of the analysis was to
identify specific legal measures that are inconsistent with commitments incorpo-
rated in the protocol. The review focused on professional services (legal, account-
ing, architectural, and engineering), road transport, distribution (retail and
wholesale), and telecommunications legislation.
The report showed that none of the Partner States has fully liberalized trade in
services as agreed under the CMP.
Professional services account for nearly three-fourths (73%) of the 63 identified
measures. Telecommunications and retail were the only studied sectors without
measures identified as inconsistent to the protocol. The measures are most common
in Tanzania (17) and Kenya (16), followed by Rwanda (11), Uganda (10), and
Burundi (9). Burundi’s strong performance on the scorecard is partly due to the fact
that some of its sectors are not yet regulated through sectoral legislation. Across the
256 A. Ombudo K’Ombudo et al.
Annex VI of the protocol identified 20 operations that had to be free from legal and
regulatory impediments including securities, credit, direct investment and personal
capital operations.
Kenya, complying with 17 out of the 20 unrestricted operations, makes it easiest
to move capital across the EAC. Tanzania and Burundi, complying with only 4 out
of the 20, make it hardest. All EAC partner states have restrictions that affect
inward investment from other EAC economies. And among the member states,
Burundi and Tanzania also impose restrictions on outward direct investment.
Four EAC Partner States—Burundi being the only exception—have introduced
exemptions to the protocol without following notification requirements to other
member states or the EAC Secretariat. At least nine such exemptions are in place,
justified by concerns about prudential supervision, public policy, money launder-
ing, financial sanctions agreed to by Partner States, and financial disturbances.
Despite signing the protocol in 2010, and contrary to the requirements of Article
24, the adoption of new restrictions on the movement of capital can be observed in
some laws.
Article 24 (c) of the protocol requires member states not to introduce new
restrictions on the movement of capital and payments connected with such move-
ment. Member states have applied measures inconsistent to this regulation.
The EAC Secretariat and EAC member states are urged to enforce the notifica-
tion mechanism in order to create a transparent and credible system for monitoring
intra-regional free movement of capital. And in the exceptional case that member
states deemed necessary to impose restrictions on the movement of capital, these
should be intended to be temporary. Exemptions allowed under Article 25 (1) of the
protocol should be applied only when essential.
Capital controls restrict investors from Burundi and Tanzania to invest in other
EAC markets. And for Tanzania, its regulatory framework restricts access to its
own stock exchange for investors from the EAC region.
Article 25 (1) of the protocol allows member states to restrict the freedom of
movement of capital for reasons of prudential supervision, public policy, money
laundering, and financial sanctions agreed to by member states. But Article
25 (2) of the protocol requires the states that adopt any of the restrictions stipulated
in Article 25 (1) to notify the EAC Secretariat and other Partner States and furnish
proof that a restriction was reasonable and justified. This approach was designed to
allow for discussion of the proposed actions, taking into account the views of state
and non-state actors that might be affected by them, and for monitoring to ensure
11 The East African Community Common Market Scorecard 257
that such restrictions last only as long as needed. Yet no EAC member state has
complied with these notification requirements. Uganda has 3 exemptions, Tanzania,
Kenya and Rwanda each 2 and Burundi none.
11.7 Conclusions
This chapter assessed the EAC CMS, as published for the first time in 2014 by the
World Bank Group, in collaboration with the EAC Secretariat. The Scorecard is
part of the implementation process of the EAC Common Market Protocol, which
came into force on July 1st 2010, and aims at monitoring the measures that are
taken (or not taken) by the EAC member states to review their domestic regulatory
environments and ensure compliance with the Protocol, and it recommends reform
measures if needed.
The Scorecard assesses member states’ compliance with the agreed commit-
ments on the liberalization of intra-regional flows of goods, services and capital
over a period of 18 months. The 2014 edition of the Scorecard identified at least
51 non-tariff barriers affecting trade in goods and 63 non-conforming measures in
the trade of services. With respect to the liberalization of capital flows, only 2 of the
20 operations covered by the CMP were free of restrictions in all of the EAC
member states.
From the experience with the first edition of the EAC CMS, a number of lessons
can be drawn:
– Even if a monitoring system is limited to assessing de jure implementation of
regionally agreed measures (to build a common market, in this case), a number
of technical difficulties and governance challenges are faced.
– As in a regional multi-country context one has to take institutional diversity into
account, it is not always easy to identify good region-wide data points.
– Monitoring provisions in treaty texts contribute to reaching sufficient levels of
commitment from national governments and administrations. Member state
level commitment towards the monitoring exercise is indeed an essential success
factor, independently of the many good reasons that exist to justify the moni-
toring process. National governments are not always demanding parties for
monitoring tools, especially if they expect below-average scores in cross-
country comparisons. For this reason, the private sector needs to take up a
more active role as an interested party in common market implementation, so
that they can maintain pressure on national governments to assure their enjoy-
ment of rights under the protocol in regional markets.
– Process scorecards – monitoring instruments that track commitments made by
government regulators and private sector through a process of private-public
dialogue – need to be strengthened and implemented for faster delivery of
integration commitments.
258 A. Ombudo K’Ombudo et al.
References
Governing regional integration for development: Monitoring experiences, methods and pros-
pects (pp. 1–7). London: Ashgate.
De Lombaerde, P., Estevadeordal, A. & Suominen, K. (2008b), Governing regional integration for
development: Summary and conclusions. In: P. De Lombaerde, A. Estevadeordal and
K. Suominen (eds), Governing regional integration for development: Monitoring experiences,
methods and prospects. London: Ashgate, pp. 275–283.
de Melo, J., & Tsikata, Y. M. (2014). Regional integration in Africa: Challenges and prospects,
(WIDER Working Paper, (2014/037)).
European Commission. (1997). Single market scoreboard, November (1). Brussels: European
Commission.
European Commission. (2001). European Governance: A white paper (COM(2001)428).
Brussels: European Commission.
Kondo, S. (2002). Fostering dialogue to strengthen good governance. In OECD, public sector
transparency and accountability: Making it happen (pp. 7–12). Paris: OECD.
Rillo, A.D. (2011). AEC scorecard: The road to ASEAN single market. Presentation at Network of
Reformers Meeting, Feb. 4, Mombasa.
UNECA. (2002). Annual report on integration in Africa 2002. Addis Ababa: UN Economic
Commission for Africa.
UNECA. (2004). Assessing regional integration in Africa, ECA policy research report. Addis
Ababa: UN-ECA.
World Bank/EAC Secretariat. (2014). East African common market scorecard 2014. World Bank
and EAC Secretariat: Washington – Arusha.
Chapter 12
Assessing Regional Integration at
the Country Level: A Possible Framework
as Illustrated for the COMESA Region
Rattan J. Bhatia
12.1 Introduction
The present chapter is a modified version of a report prepared for COMESA by the author to design
a performance assessment framework that can track and assess individual country’s integration
efforts periodically and provide objective basis for disbursing aid funds related to that
performance.
R.J. Bhatia (*)
African, and Policy and Development Review Departments, IMF, Washington, DC, USA
IMF Office, United Nations, New York, NY, USA
e-mail: rjbhatia@aol.com
The methodology behind the proposals here was developed on the basis of an
answer to a simple question: What it takes for a country to join a group of countries
aiming to form a mutually beneficial economic and monetary union? This issue has
been fairly extensively discussed in a AfDB/COMESA report on Multilateral Fiscal
Surveillance Framework (MFSF) for the COMESA region (Bhatia et al. 2011) and
in the integration promotion programs of the Asian Development Bank in central
and south-east Asian regions. The substance of the answer to the question empha-
sizes actions in four areas: Connectivity, Communications, Commerce (the three
Cs) and, Sustainability with its two sub-components of Competitiveness (the fourth
C) and, Productivity growth (one P). Assessment of the integration process could be
based on developing monitoring and evaluation indicators in these areas.
The general menu of assessment indicators should be comprehensive and thus
cover major actions in these areas of 4Cs and 1P that may be needed to advance
trade integration sustainably towards its immediate objective of operationalizing
customs union. Furthermore, regional integration involves legal and regulatory
processes (agreement at regional level, domestication, and implementation), policy
formulation and implementation, institutional building and strengthening, and
finally the resulting output. A comprehensive integration program will, therefore,
include a myriad of legal, institutional, and policy measures and their
corresponding indicators. Several authors have developed and proposed various
indicators of regional integration (De Lombaerde et al. 2008) that could be utilized
for the purpose of this paper. Dennis and Yusof (2003)3 have identified a
1
Even existing studies dealing only with regional trade integration deal with the end-outcome of
trade integration as noted in this quote: “Even considering RI under the restricted dimension of
trade, most. . .publications actually focus on measuring the effects/impact of trade integration,
much less . . . on the process of trade integration”.
2
Indeed, the performance matrix of COMESA MTSP is essentially on promoting intra-regional
trade facilitation and trade promotion.
3
Thus, a country could stagger the domestication of regional regulations and laws over a period of
years, and in the meantime may be reducing NTBs.
12 Assessing Regional Integration at the Country Level: A Possible Framework. . . 263
comprehensive set of 145 indicators focusing on trade in goods and in financial and
other services, investment flows and stock, infrastructure, customs, standards,
mutual recognition agreements and conformity assessments, SMEs, e-ASEAN,
and intellectual property. Salmon and Akanni-Honvo (2009) have proposed
50 ‘core’ indicators of which 9 relate to trade integration, 4 for monetary and
financial integration, and 8 for transport and communication. Similarly, the
COMESA Medium-Term Strategy Plan (MTSP) 2011–2015, had a large number
of regulatory, institutional, and policy measures that member countries, as well as
the regional actors responsible for promoting integration, needed to undertake on
the road to regional integration.
Following Dennis and Yusof, integration indicators may be categorized as
process indicators, input (implementation) indicators, output indicators, and stock
and flow indicators. Process indicators measure the progress towards the establish-
ment of an appropriate legislative, regulatory and implementation environment for
the indicator under consideration in each member country. Input indicators measure
the actual implementation of measures to realize that particular aspect of regional
economic integration. Output indicators record what the actual result is when the
new policy framework is in place and appropriate implementation measures have
been taken. Stock and flow indicators are relevant in the case of the integration of
financial services and investment regime. In the earlier stages of integration,
process indicators are likely to dominate the assessment exercise, whereas in the
later stages, once the regulatory and institutional environment is in place, imple-
mentation and outcome indicators would become predominant. However, this
sequential scenario may not be how the integration process unfolds, as countries
may be putting in place various regulatory and legal requirements gradually over a
period while simultaneously implementing other policy and institutional require-
ments to achieve some results (output) earlier on the integration roadmap (Dennis
and Yusof 2003). Thus, at any time, a country’s integration program may consist,
and generally will consist, of a combination of process, input, and output indicators
suggesting that a performance assessment framework should also provide for
simultaneous assessment of these types of indicators.
The selection of assessment indicators must satisfy certain criteria. In the first
place they, in their entirety, must ensure that their fulfillment will result in a
sustainable integrated region. Beyond that, each of the selected indicators should
be relevant to the purpose of the evaluation, be simple and clear to understand and
to monitor, should be objective to avoid value judgments, and have statistical
consistency and timely availability across member countries. However, an excep-
tion to these requirements may need to be made in respect of some indicators-
especially those denoting outcomes of policy consequences. Such indicators,
though relevant and objective, may not meet the requirement of clarity and time-
liness because the causal relationship and/or time interval between cause and effect
are indeterminate. Nevertheless, tracing the development in these indicators pro-
vides indispensable insight into the progress being made in achieving the integra-
tion objectives and an invaluable guide to assessing the efficacy of policies being
pursued. Accordingly, these indicators need to be included as ‘benchmarks’ or
264 R.J. Bhatia
4
This freedom, however, is subject to the limitation that the performance indicators are selected
from the list of indicators included in the RIIP of the country concerned and, hence, also in the
regional Matrix.
266 R.J. Bhatia
approved by the regional authorized body; if the latter deems the number of
performance indicators proposed by a member state in its RIIP as insufficient it
would not approve the program and ask the member country to revise the number
upwards. But this approach introduces unnecessary uncertainty for member states
regarding the approval of their respective RIIPs and may delay the finalization of
the program. Furthermore, in the absence of such a requirement, member states may
be tempted to identify only a very small, and easily achievable, number of perfor-
mance indicators to ensure that they could fully satisfy those criteria and draw their
entire allocated tranche. But this approach will risk that even a full satisfaction of
those fewer indicators may not ensure that integration objectives will be met if
some needed crucial indicators in the RIIP remain unfulfilled.
An efficient procedure would entail that member states are provided with some
specific guidelines to formulate their RIIPs, and which give the governments
reasonable assurance that if those guidelines are followed, their programs would
meet with the approval of the regional body. Member states must also be well
apprized of the basis for satisfactory conclusion of subsequent assessments to
unlock further tranches. These guidelines should include, inter-alia, the following:
1. Regional objectives and intervention instruments should be translated into a
regional matrix of indicators of which a few should be designated as ‘core’
indicators.
2. Country RIIPs should be consistent with the regional objectives and include
indicators5 chosen from the regional matrix, including the designated ‘core’
indicators.
3. Of the indicators included in RIIPs, countries should identify a minimum
number as performance indicators, including all the ‘core’ indicators,6 as the
basis for performance assessment and scoring.
4. In addition to the above indicators, RIIPs should include a few specified ‘bench-
marks/impact indicators’ which, although not a factor in the scoring exercise,
could assist the monitoring and evaluation process.
It is difficult to settle a priori on a specific number of performance indicators
which should be decided on a case by case basis. However, the number must be
such as to accommodate all the core indicators while still providing some flexibility
to countries in the choice of their performance indicators.
Each member country’s performance in advancing its integration within the
region in any one year will be assessed in relation to the baseline performance
indicators in its RIIP.7 Since there will be more than one such indicator, an overall
assessment of the performance would require some averaging of the performance
for each indicator. This may be done by either a simple arithmetic average where
5
Not necessarily the same for all RIIPs.
6
This requirement will also ensure a kind of a ‘level playing fields’ for scoring purposes.
7
Thus, if the base indicator relates to the decrease of 8 in non-tariff barriers (NTB) and the actual
decrease is 4, the performance ratio would be 50%.
12 Assessing Regional Integration at the Country Level: A Possible Framework. . . 267
each indicator has the same weight or by a weighted average where different
indicators have different weights according to their importance. Disbursement of
supportive aid funds, if linked to the performance assessment, could be determined
on the basis of this calculated average performance ratio. At the outset, the country
would have been ‘allocated’ a certain amount of assistance for each year should it
meet its baseline indicators for that year. The actual amount of disbursement,
however, would be a proportion of that amount equal to the average ratio of
outcomes to baseline indicators. Thus, if the initial allocation of assistance in the
first year was 20 million Euros and the average performance ratio is 0.50 actual
disbursement would be only 10 million Euros. If, however, the country
overperforms and its average performance ratio is greater than 1, disbursement
would still be limited to the initial allocation, i.e 20 million Euros even though
actual outcome is better than the baseline indicator.8
It is to be expected that in actual practice annual outcomes may differ from the
baseline indicators (targets). The question then arises whether the baseline targets
for the remainder of the program period be changed or remain fixed as originally
indicated in the implementation program. On the premise that exceptions make a
rule unfortunate, and in the interests of lending credibility to the implementation
programs, a desirable option will be not to change the baseline indicators. The
annual baseline targets should be regarded as ‘cumulative’ so that any deviations
from the baseline targets at the end of a year are automatically reflected (compen-
sated) in the baseline target for the following year. Thus, if a country has
programmed to reduce NTBs by 10 and 20 in year 1 and 2 respectively, but reduces
NTBs by only 5 in year 1, then its baseline indicator (reduction in NTBs) for the
following year will be 25 (20+5). Conversely, if the country reduced NTBs by 15 in
the first year, then its baseline indicator for the following year will be 15 (205).9
12.3.1 Background
The Common Market for Eastern and Southern Africa (COMESA) is a regional
economic grouping made up of 19 Member States with an estimated population of
400 million people. Pursuant to the 1993 Treaty COMESA was established in 1994
to succeed the Preferential Trade Area (PTA) for Eastern and Southern Africa that
had been in existence since 1981. Its current vision is the promotion of a monetary
8
This is further explained and illustrated in considering the COMESA example below.
9
Another way to indicate baseline indicators could be to express them in cumulative numbers
e.g. 10 for the first year and 30 for the second year Cumulative outcomes would then also be
expressed cumulatively for assessment purposes.
268 R.J. Bhatia
union among its member countries- a vision that builds upon its past achievements
in promoting regional integration, primarily through trade integration, but also
through promoting economic, fiscal and monetary harmonization and convergence.
In October 2000, it launched Africa’s first Free Trade Area which is operational and
now comprises 14 of its 19 member states. A few years later, in 2009, it launched
the Customs Union which was expected to be operational in 2012. The 1993 Treaty
amongst Member States of COMESA also envisages the setting up of a Monetary
Union, enjoining that member countries shall “in the field of monetary affairs and
finance, cooperate in monetary and financial matters and gradually establish con-
vertibility of their currencies and a payments union as a basis for the establishment
of monetary union.” This objective is being pursued within the framework of a
Monetary and Fiscal Harmonization Programmed that stipulates various conver-
gence criteria that member states must achieve. In that context, COMESA has
established a common Clearing House, and a regional payments union.
COMESA’s Medium-Term Strategy Plan (MTSP) for 2011–2015 summarized
this progress as follows: “Beyond achievements in trade liberalization and facilita-
tion in general, notable progress has been made in the specific areas of customs
management, transport facilitation, trade and project finance, institutional develop-
ment, technical co-operation and capacity building. Progress has also been made in
policy coordination and cooperation in the productive sectors” (COMESA
2011–2015).
Despite the progress achieved in establishing various region-wide institutions
and arrangements, however, actual progress in regional integration has been rather
limited. Two recent studies (Bhatia et al. 2010, 2011) commissioned by the African
Development Bank summarized the status of integration as follows:
1. The financial integration process still has far to go as some member states have
yet to achieve macroeconomic stability, which is an essential precondition for a
sustainable monetary union, and as the modernization and harmonization pro-
cess among member states is at different stages of implementation.
2. Stabilization and harmonization of fiscal situation and fiscal policies is even
further from being in place, with almost all member states not conforming with,
or moving towards, the intended fiscal convergence criteria.
3. Trade integration, as measured under three different concepts of integration:
intra-regional traded openness,10 Degree of regional trade integration,11 and
Relative Integration of each (member) country vis-a-vis other (member) coun-
tries,12 so far has been limited, with FTA not being ‘effective’ in terms of
promoting intra-regional trade.
10
Defined as the member country’s trade with other member countries of the region as a proportion
of its GDP.
11
Defined as the ratio of the country’s intra-regional trade to its trade with the rest of the world.
12
Defined as the ratio of the country’s trade with other member states of the region to the total
intraregional trade within that region.
12 Assessing Regional Integration at the Country Level: A Possible Framework. . . 269
The COMESA MTSP and Performance Framework clearly set the region’s prior-
ities and intervention instruments to meet its integration goals that envisage an
eventual monetary union, with time-bound intermediate goals of FTA, CU, and
Common Market. However, COMESA has also rightly adopted the ‘variable
geometry’ approach under which member countries choose their own time path
for effectively participating in the regionally agreed integration stages, and
employing intervention instruments, leading up to the monetary union. This
approach implies that member states may also choose to have a different time
frame (sequence) for implementing MTSP’s priorities and intervention instruments,
though consistent with the ultimate objective of joining the monetary union. Thus,
for the immediate purposes of RISM, country performance cannot be compared
with the time-bound objectives and intervention instruments set in the MTSP. For
example, the MTSP envisaged operationalizing Customs Union in 2012. However,
some countries that have not as yet joined the COMESA FTA may effectively aim
at joining the already operational FTA in the immediate period ahead, and follow it
with joining the CU sometime later in the medium-term period extending beyond
2012. A similar argument may be made concerning the intervention instruments.
While countries may be expected to employ those instruments eventually, they may
choose to employ different combinations of those instruments in the light of their
270 R.J. Bhatia
are more germane to the performance assessment exercise. Thus inclusion of these
two outcomes as indicators would appear to be desirable. However, the problem in
those cases is that the causal relation between outcomes and policies, as well as the
time lag between cause and effect, is not clear or determinate. For example, intra-
regional trade may respond to domestic policies with different time lags and may be
influenced by external circumstances that may negate, enhance or hide the impact
of domestic policies. In such cases it is difficult to justify including these outcomes
in the menu of indicators. It will nevertheless be useful to trace the evolution of
these outcomes as they could assist in determining the effectiveness of the policies
being followed and, in the case of RISM, be used as a guide in the formulation of
the second RISM. Table 12.2 above lists a few such indicators that, on a priori
considerations, would be the consequence of policies implemented in the earlier
years, though with indeterminate time lags. To distinguish them from performance
indicators, they are termed here as ‘benchmarks’/impact indicators, but their evo-
lution will not affect the annual PAF scoring.
The first two benchmarks would inform how far the country has succeeded in
integrating its trade with the regional economies. The comparison between bench-
marks 2 and 3 would rate the performance of the member country with that for the
region as a whole. The fourth benchmark would suggest the impact of liberalization
of financial services on intra-regional flow of financial transactions, and the fifth
benchmark would indicate improvement in the country’s business profile vis-a-vis
274 R.J. Bhatia
its regional partners and the rest of the world. Country RIIPs should include these
benchmarks as monitoring and assessment tools. Implementation of regionally
agreed regulations, procedures, and policies as reflected in RIIP should contribute
to an improvement in the above benchmarks. However, as there could be annual
fluctuations in these benchmarks/impact indicators, the assessment could be based
on some moving average formula.
The formulation of the above set of indicators and benchmark to be included in
country RIIPs is dictated by the objective that it be a comprehensive menu covering
major actions in the areas of 4 Cs and 1 P mentioned earlier. It may be noted,
however, that COMESA countries have a long history of cooperation and coordi-
nation of their policies, and of introducing region-wide regulations and institutions.
Thus, some of the indicators listed in Table 12.1 may have been partially or fully
fulfilled by some member countries. In those cases countries would be expected to
maintain those policies, regulations, and institutions but may have to review them
with a view to refocusing them on the objectives, and strengthening their imple-
mentation capacity to meet the overall integration objectives as may be identified in
their respective RIIPs. As an example, the menu of indicators in Table 12.1 includes
the establishment of a Monitoring, Evaluation, and Reporting Unit (MERU) as one
indicator. However, many countries have already established inter-ministerial
committees that monitor their implementation of regional obligations. In that
case, member countries will need to determine what the tasks of the MERU could
be in the light of their strengthened integration agenda, and determine whether
those tasks could be fulfilled by the existing arrangements and, if not, what changes
would be required in their terms of reference and in staffing and other capacities to
meet the new tasks. A choice will then need to be made whether to maintain the
inter-ministerial committee format or to set up a new MERU.
Member states’ integration status within the region and their special circum-
stances also differ greatly. For example, as of mid-2014, 13 member states have
already joined the COMESA FTA and only six member states remain to join the
Arrangement. Thus the indicator relating to reducing tariffs on intraregional trade
to zero is relevant for only those six countries. Further, as an example, in their
programs, countries are likely to adopt different time horizon for joining the
Customs Union than the 2012 target date of COMESA’s MTSP and, accordingly,
a different time sequence for the implementation of the relevant intervention
instruments (e.g. implementation of CET). In other cases, some countries may
have already fulfilled some of the indicators listed in the general menu. Thus in
their RIIPs, member states may not need, or wish, to adopt immediately the entire
menu of objectives and intervention instruments outlined in Table 12.1 above.
Their respective RIIPs in any year may, therefore, contain fewer objectives and
instruments than the full menu as may be justified by their objectives for the 3-year
RISM period, and by their specific circumstances. In this sense, the 19 indicator
framework may be regarded as an a la carte menu for member states to choose from
12 Assessing Regional Integration at the Country Level: A Possible Framework. . . 275
Table 12.3 below elucidates a ‘generic’ Matrix that reflects a country’s RIIP,14
indicating ‘objectives’ (targets) the country wishes to achieve in respect of the
19 indicators during a 3-year period (here: 2012–2014). It may be noted that in this
case not all the indicators are being addressed in each year. While some quantifiable
indicators are targeted to be met over a period longer than just one year, the
non-quantifiable indicators (e.g. implementation of CTN) by their very nature are
targeted to be met in one year. The Table also indicates hypothetical realization
(‘estimate’) of those targets. While qualitative indicators, by their nature, will be
satisfied at a point in time, the selected quantifiable indicators are targeted to be
realized gradually during the program period. These latter annual targets are
cumulative and are so shown in the RIIP Matrix of Table 12.3. This implies that
if any target is not met (or over met) in one year the shortfall (excess) is automat-
ically added to (subtracted from) the initial target for the next year. As an illustra-
tion, if the targeted reduction in the number of NTBs is 6 in each of the 3 years, the
cumulative target indicator will be 6, 12 and 18 for the 3 years respectively. If in
year 1 the reduction number is 3 (7) compared to the target of 6, the remaining
shortfall (excess) of 3 (1) is added to (subtracted from) the original annual reduction
target of 6 in year 2 and becomes 9 (3), with the cumulative target remaining as 12.
While a qualitative target cannot be cumulative in this sense, it is nevertheless
‘cumulative’ in the sense that if it is not satisfied in one year it is added as the
indicator for the following year. For example, if the initial target of domestication
of CTN is not fulfilled in the year it is targeted to be implemented, it automatically
becomes the indicator target for the following year. Conversely, if an indicator that
was targeted to be met in the second year is fulfilled in the first year it will be
credited for scoring in that year and no new target would be put in its place in the
second year.
As argued in the previous section, not all the above indicators included in the
RIIP need to be adopted as ‘performance’ indicators and their number could be
smaller, say nine. However, to ensure that the chosen performance indicators will
make a significant contribution to trade integration, and that there will be some
13
Thus, even if, for example, a country may not wish to implement CET during 2012-14, RIIP
should at least indicate when, after 2012, will the country begin to implement CET.
14
The word ‘generic’ is meant to indicate that while the Table 12.3 lists all the indicators of
Table 12.1 in practice individual country RIIPs may contain a fewer such indicators and some
other country specific indicators.
Table 12.3 COMESA illustrative regional integration implementation program (RIIP), 2012–2014
276
(1.00)
(b) Number of services liberalized (1.00) Output 10 3 1 6 3 10 7 0.33 0.50 0.70
12
13. Enforce guidelines and procedures of competition Output 2011 – – Yes No Yes No – 0.00 0.00
regulations
Adopted by member states
14. Domestication of COMESA Common Investment Area Input 2011 Yes No Yes Yes – – 0.00 1.00 –
Agreement
15. Improvement in productivity and competitivenessb
(a) WB. Sub-index Registering productivity (0.25) Output 2011 – – >0 >0 >0 >0 – 0.25 0.25
(b) WB. Sub-index Enforcing Contracts (0.25) Output 2011 – – >0 0 >0 >0 – 0.00 0.35
(c) WB. Sub-index Protecting Investors (0.25) Output 2011 – – >0 0 >0 0 – 0.00 0.00
(d) WEF Sub-index of Global Competitiveness Index Output 2011 – – >0 >0 >0 >0 – 0.25 0.25
(0.25)
Either Performance of domestic economy or Efficiency of
labour markets or Basic infrastructure
16. Implementation of COMESA Transit Transport Facili-
tation Instruments
(a) Carrier License (0.20) Input 2011 Yes Yes – – – – 0.20 – –
(b) Axle Load Limits & Overload Controls (0.20) Input 2011 – – Yes No Yes Yes – 0.00 0.20
(c) Yellow Card adopted and used where applicable (0.20) Input 2011 – – Yes Yes – – – 0.20 –
(d) Harmonized Road Transport Charges (0.20) Input 2011 – – – – Yes Yes – – 0.20
(e) Implement Phase 1 of COMESA Legal Notice # 2 for
Liberalization
(f) Air Transport within the Region (0.20) Input 2011 – – Yes Yes – – – 0.20 –
17. Reduction in cost of intra-border trade & Exchange Output 2011 – – >0 >0 >0 >0 0 1.00 1.00
Improved ranking in W Bk. Sub-index Trading Across
Borders of DBI
Assessing Regional Integration at the Country Level: A Possible Framework. . .
18. Reduction in fiscal deficit to COMESA Performance Output 2011 2 1 4 2 5 4 50 50 80
Criterion (3%) (% point cumulative reduction)
a
Especially on financial services
b
Reduction in the gap vis-a-vis COMESA average ranking (base 2011)
277
278 R.J. Bhatia
15
This set is only an illustration and actual preferred criteria could be different.
16
Some states may opt only for cabinet approval of RIIP but this runs the risk that it could easily be
reversed by a subsequent decision of the same cabinet. As a minimum, therefore, if a MS prefers
the cabinet approval option, it should be required to state that in case of the reversal of the decision,
the reversal will be available in public domain along with the explanation as to its rationale.
17
In the case of member states that have yet to join the FTA, a corresponding indicator will be the
‘Reduction in the gapvis-a-v is COMESA average ranking (base2011)’.
12 Assessing Regional Integration at the Country Level: A Possible Framework. . . 279
indicators, the intermediate indicators will not count as separate indicators but as
parts of one indicator of which they form the components. As an example, the
indicator ‘Implementation of COMESA Transit Transport Facilitation Instruments’
has five ‘intermediate’ indicators but all the five intermediate indicators together are
to be regarded as one indicator, and not five, for the purpose of meeting the
performance requirement.
The scoring system will first compare realization (‘estimate’ in Table 12.3) with the
targets and assign each such comparative outcome a numerical value. The scoring
system elucidated in Table 12.3 is a revised version of the present COMESA system
that grades every performance as ‘yes’ or ‘no’ when a target is fulfilled or not
fulfilled. ‘Yes’ is assigned a value of 1, while ’No’ is graded as 0. The overall
performance score is then the simple average of these individual scores. The system
proposed here retains the ‘Yes’ and ‘No’ classification for non-quantifiable indica-
tors, with respective scoring of 1 and 0. However, for those indicators that can be
quantified, the score is valued as a proportion of 1 corresponding to the performance
as a proportion of the target value of the performance indicator. Thus, if the
indicator target for one year is a reduction of six NTBs, but the country reduced
only three NTBs, its performance is 50% of the target and the corresponding score
assigned will be 0.50. In cases where the performance indicator has two or more
intermediate indicators, the maximum value for each intermediate indicator is
deemed as equal to the value of 1 divided by the number of intermediate indica-
tors.18 Thus, if the number of intermediate indicators is five, then each such
intermediate indicator is assigned a value of 0.20 and scoring will be the percentage
of 0.20 corresponding to the percentage of the target achieved. As in the COMESA
system, the performance scores for each relevant indicator will then be added up
and a simple average arrived at and compared with that for the average value for the
target indicators. Note that only the indicators that are identified as the performance
indicators and have value assigned to them for the year are assessed, added and
averaged for purposes of RISM scoring and fund disbursements.
The COMESA framework of performance assessment adds the individual scores
of all the indicators and a simple average is then calculated for the indicators so
graded. These averages are then grouped into three categories of performance with
corresponding three categories of percentage disbursement of budgeted allocation
(Table 12.4).
18
An alternate, perhaps simpler, system may be to value even the intermediate indicators as 1, but
this may give the impression that each intermediate indicator is a stand-alone indicator, which it is
not. To avoid this impression, intermediate indicators must be shown as sub-components of the
main indicator, whatever scoring value may be adopted.
280 R.J. Bhatia
19
Unless any requested changes in the target are approved by the COMESA FUND, as explained
earlier in the text.
12
(c) Yellow Card adopted and used where applicable – – Yes Yes – – – – 0.20 0.20 – –
(d) Harmonization of Road Transport Charges – – – – Yes Yes – – – – 0.20 0.20
(continued)
281
Table 12.5 (continued)
282
On the basis of the above calculations, and in conformity with the disbursing
framework of Table 12.4, the country’s performance will rate as category 2 perfor-
mance in the first year, and category 3 performance in the second and third years. It
will, therefore receive 50% of its allocation in the first year and 100% in the
subsequent 2 years. The question then arises as to what happens to the undisbursed
amount of the first year. Three options could be considered:
1. The undisbursed amount is not lost to the country but is added to its original
budgeted allocation for the following year and available on the basis of its
performance for that year.
2. The undisbursed amount is lost to the country, carried over to the following year
and added to the general kitty for that year for allocation among all members in
the pre-determined proportions.
3. The undisbursed amounts are reserved for distribution at the end of the RISM
period and disbursed to all the countries in the currently COMESA agreed pro-
portions, or on the basis of their average performance ratio for the RISM period.
Each option has its pros and cons. Under option 1, the country does not lose its
allocation, and the incentive for it to perform better in the following year is increased
as the amount available becomes larger. It is nevertheless penalized by losing the use
of funds for one or more years depending on its performance in the subsequent years.
On the other hand, it may also be argued that the country has less incentive to perform
in the earlier years in the certain knowledge that it will have those funds available in
the subsequent years if it performs better. Under Option 2 and 3, the country has
better incentive to perform in each year, as the amount it will receive in the following
years from the funds withheld from it for underperformance in the first year will be
only a small proportion of the amount lost in that year (assuming it performs better),
the rest being distributed to other members in the group. Option 2 has an advantage
over the third option in that it avoids accumulation of large undisbursed funds at the
end of the RISM period when their distribution may encourage political disputes.
Both options, however, imply penalizing the underperformers more severely than
under Option 1. On the other hand, option 2 still leaves the problem of undisbursed
funds at the end of the RISM period. Should they be lost for the underperforming
countries in that year, defining under performers as those below the regional average
for the full 3-year RISM period? Or should they just be redistributed to all the
members in the agreed COMESA proportions? Obviously, the first option is too
drastic, but there may be other alternatives to the second option, e.g. distribution
among all the members in the agreed COMESA proportions.
12.4 Conclusion
The main purpose of this study was to develop a system of indicators and a
Performance Assessment Framework that could be used to assess the process of
integration of a country into a regional block. To that effect it develops a
284 R.J. Bhatia
References
Bhatia, R. J., Zhang, J., & Kiptoo, C. K. (2011). Facilitating multilateral fiscal surveillance in
monetary union context with focus on COMESA region. Abidjan: COMESA, African Devel-
opment Bank.
Bhatia, R. J., Zhang, J., & Kiptoo, C. K. (2010). Financial sector integration in three regions of
Africa. African Development Bank. Tunis (Tunisia)
COMESA (2011-2015). Medium-term strategic plan: Towards an integrated and competitive
market, 2011–2015.
Dennis, D. J., & Yusof, Z. A. (2003). Developing indicators of ASEAN integration: Survey for a
preliminary survey for a roadmap, REPSF Project 02/001.
De Lombaerde, P., Pietrangeli, G., & Weeratunge, C. (2008). Systems of indicators for monitoring
regional integration processes: Where do we stand? The Integrated Assessment Journal, 8(2),
39–67.
Salmon, J. M., & Akanni-Honvo, A. (2009). ACP regional integration support- monitoring
regional integration. Brussels: The African, Caribbean and Pacific Group of States.
Part IV
Asia
Chapter 13
Monitoring the ASEAN Economic
Community
Aladdin D. Rillo
13.1 Introduction
Since the Association of Southeast Asian Nations (ASEAN)1 has envisioned building an
ASEAN Economic Community (AEC) by 2015, efforts have been strengthened to
deepen regional integration. Under the AEC, the goal is to establish a single market
and regional production base, as well as a highly competitive region with equitable
economic development, and a region that is fully integrated into the global economy.
Realization of this goal requires the elimination of barriers to free flow of goods, services,
investment, capital and skilled labor, with ultimate objective of improving the material
welfare and well-being of countries in the region. To implement the AEC, an AEC
Blueprint was also developed comprising of various policy actions, measures and
strategies, with clear timelines, to achieve the envisioned goals of establishing the AEC.2
1
ASEAN was established in 1967 by five countries (Indonesia, Malaysia, Philippines, Singapore and
Thailand; collectively known as ASEAN5) mainly for political reason. Since then the Association has
expanded to its current ten members (ASEAN5, Brunei Darussalam, Cambodia, Lao PDR, Myanmar,
and Viet Nam), and has broadened its thrust toward greater economic cooperation, beginning with the
Declaration of ASEAN Accord in 1976.
2
The decision to establish an ASEAN Economic Community (AEC) as an end-goal of regional economic
integration was made in Bali, Indonesia during the 16th ASEAN Summit in 2003. AEC is actually one
pillar comprising the ASEAN Community, which includes the other two pillars such as the ASEAN
Socio-Cultural Community (ASCC) and ASEAN Political and Security Community (APSC). In early
2007 the Leaders agreed to develop “a single and coherent blueprint” to implement the AEC, and
consequently, the AEC Blueprint was signed in November that same year. To ensure that this Blueprint is
implemented on time, the AEC Scorecard was developed in 2008 to track countries’ compliance of their
commitments to the AEC Blueprint, as well as to measure the implementation rate of the Blueprint.
A.D. Rillo (*)
Senior Economist, Capacity Building and Training, Asian Development Bank (ADB) Institute,
Tokyo, Japan
e-mail: arillo@adbi.org
However, one challenge facing ASEAN is how to monitor the progress of the
AEC. Given the diversity of the region and the capacity constraints of countries in
the implementation of various measures, building an integrated community is not
easy. It requires not only the ability to implement initiatives to support the markets,
but as well as a high level commitment by countries to ensure that policies are
supportive of integration. At the same time because regional economic integration
is a complex process, there’s a need to engage the participation of all relevant
stakeholders and players whose competing interests may sometimes hinder the
implementation of various integration initiatives and measures. Thus, it is within
this context that the monitoring of the AEC becomes crucial.
3
Monitoring is undertaken by the ASEAN Secretariat through its ASEAN Integration Monitoring
Office (AIMO) established in 2010. In developing the Scorecard, AIMO compiles all measures as
earlier identified by the various sector bodies in the AEC Scorecard Master Plan. The list is then
circulated to all working bodies for validation and approval. The approved list of measures
becomes the basis for the scorecard for that particular phase of monitoring.
4
The AEC country scorecards are also developed in a similar way as the AEC Scorecard and
consist of country-specific measures identified by the countries themselves. The country score-
cards are updated by AIMO through inputs submitted by ASEAN Member States.
5
There are two versions of the AEC Scorecard being prepared by ASEAN Secretariat. The
“official” version contains all detailed information about country commitments and specific
measures and the rate of compliance. This version is an internal document, updated twice a year
and is reported to the ASEAN Economic Ministers (AEM), ASEAN Economic Community
Council (AECC), and ASEAN Leaders’ Summit Meeting. Another version is called a “public”
version which is published by ASEC every two years for purposes of informing the general public
of the progress and achievements by AEC.
290 A.D. Rillo
Overall, ASEAN has done a good deal in advancing the AEC as shown
in Table 13.1. This translated to a total implementation rate of 82.3 percent as
of end-December 2015 (503 out of 611 ASEAN-wide measures already
implemented).
Despite this progress, monitoring the AEC through the Scorecard remains
challenging. At a practical level, compliance to AEC commitments is constrained
by implementation problems on the ground. In general, most of the bottlenecks in
implementing the AEC are due to: (1) delays in the ratification of the signed
ASEAN agreements/protocols which affect their entry into force; (2) failure to
13 Monitoring the ASEAN Economic Community 291
align regional initiatives to domestic laws and regulations; and (3) lack of political
will to implement regional and country-specific commitments. Because of these
problems, the ability of the Scorecard to keep track of the implementation is also
limited.
On a technical level, the AEC Scorecard at this juncture is only a compliance
tool to ensure ASEAN and its Member States are on track with the implementation
of their commitments. It is not a tool to evaluate the impact of these measures.
Nonetheless, the AEC Scorecard should be an evolving document that could be
improved or revised and could be used in tandem with other instruments for
evaluation purposes. Since the AEC Scorecard is a compliance tool, it is not
expected to be a rigorous monitoring tool which in a way also limits its
effectiveness.6
6
One issue about the current AEC Scorecard is the methodology being employed to measure
compliance, which is based on a simple “yes” or “no” to determine if a particular measure is fully
implemented or not. To enhance the Scorecard, two studies were undertaken by the Economic
Research for ASEAN and East Asia (ERIA) which tried to develop a rigorous methodology and a
set of indicators to assess status and progress of integration both at the national and regional levels.
The results of the studies were supposed to complement the AEC Scorecard and to assist ASEC in
improving it.
292 A.D. Rillo
7
The ACPMS Framework consists of indicators for AEC and ASCC, respectively. However, since
the purpose of this paper is about AEC monitoring, only indicators for economic integration are
reported and discussed. For complete list of ASCC indicators, see 2012 ASEAN Community
Progress Monitoring System Report.
13 Monitoring the ASEAN Economic Community 293
Pillar IV (integration into the global economy). Note that the distribution of
indicators is not even due to data availability.
In fact these perceived data gaps have prompted the revision of the ACPMS
framework in 2012, by including additional indicators that are deemed important
for monitoring the outcomes of integration. For example, in the 2007 ACPMS
framework, there were no indicators to measure the free flow of capital and
investment despite the fact that these are the core elements of AEC.
One motivation for enhancement of the framework is to provide both refine-
ments and data updates to indicators of integration in order to capture more detailed
information on the general outcomes of the AEC and thus gain more insights into
the monitoring of ASEAN economic integration. Through a series of consultations
13 Monitoring the ASEAN Economic Community 295
and reviews,8 the framework was assessed in terms of indicators included and their
overall relevance. It was recognized that as a monitoring tool for AEC, the
framework should be able to measure economic convergence outcomes particularly
in the areas of economic and financial cooperation, logistics services, and ASEAN
connectivity. At the same time indicators to better reflect the economic competi-
tiveness of the region were also sought.
8
As part of the process of enhancing the framework, extensive consultations with direct users and
producers of integration indicators were made, including desk officers from the ASEAN Secre-
tariat, officials from various government agencies in ASEAN, and other stakeholders in the region.
The original 21 ACPMS indicators were also reviewed based on their relevance to AEC monitor-
ing as well as feedback received from stakeholders who were involved in actual integration
monitoring. Finally, a number of regional integration studies and researches were also reviewed
to supplement the findings from consultations and desk review of ACPMS indicators.
296 A.D. Rillo
Thus, under the enhanced 2012 ACPMS framework (see Table 13.3), the
number of indicators was increased to 29 to address some gaps in measuring the
outcomes for price convergence, finance integration, infrastructure connectivity,
regional competitiveness and equitable economic development.9
The main output for these indicators is a statistical report called the ASEAN
Community Progress Monitoring System which is prepared by the ASEAN Secre-
tariat, and which serves as a basis for developing a monitoring system of progress
toward the ASEAN Community. The first report was published in 2008 based on the
2007 ACPMS indicators, and the second one in 2013 using the 2012 ACFMS
indicators. Both reports tried to asses, in a comprehensive manner, the various
efforts by ASEAN member countries in meeting the main goals of the ASEAN
Community, and to complement other compliance monitoring efforts and reports.10
While progress has been made, the road to the ASEAN Economic Community is
still a long one. Although various initiatives have been carried out, more efforts are
still needed to deepen the region’s economic integration. Truth is, economic
integration is a very complex agenda with many challenges, requiring greater
scope of critical actions. Still a number of measures remain pending and require
immediate action by Member States.
Despite this progress, some measures have not been fully implemented, partic-
ularly those under trade facilitation (customs modernization and standard and
conformance), services liberalization, investment, agriculture, consumer protec-
tion, and ratification of transport agreements. One reason for this shortfall is the
delay in the ratification of the signed ASEAN agreements and its protocols and
completion of countries’ specific commitments in the Blueprint. Meanwhile, the
Phnom Penh Agenda for Community Building, adopted during the 20th ASEAN
Summit in 2012, highlighted the need to double efforts to realize the AEC by 2015.
Since then efforts have been intensified to set priority activities and concrete key
measures to achieve the AEC goals. In fact under the AEC Blueprint 2025 adopted
in 2015, the immediate priority is to implement the outstanding measures in AEC
9
To construct the ACPMS indicators, data are sourced mainly from ASEAN member states
through the various national statistical offices. Data submission by NSOs is made directly to the
ASEAN stats at the ASEAN Secretariat. ASEC databases are also used such as those for finance
and surveillance, FDI, tariff, services, and tourism. Other sources of data are international
organizations such as the World Bank and the IMF, particularly for some indicators where data
are only available from these institutions.
10
One such important report is the ASEAN Baseline Report (ABR) published in 2003. The ABR
provides indicators across each of the four dimensions of ASEAN Community, namely economic,
socio-cultural, political-security, and narrowing development gap, and presents their baseline
situation as of 2003. In effect, the ABR is a predecessor of the ACPMS Reports in a sense that
both reports contain indicators of integration. The ACPMS Reports, however, re-focus the
framework of the indicators and refine/improve them.
13 Monitoring the ASEAN Economic Community 297
Blueprint 2015 as well as the new measures under AEC 2025. However, to ensure a
higher implementation rate as well as to avoid backlogs of unimplemented com-
mitments, a number of challenges remain.
First, to ensure that AEC is realized, ASEAN should strengthen the implementa-
tion of programs at the national level. ASEAN Member States have been urged to
ensure that regional commitments are transposed into national commitments through
appropriate domestic processes. At the same time, capacity building, particularly for
less developed ASEAN economies, should be given particular emphasis to enable
policy makers in those countries to follow through on their commitments.
Second, it is imperative that a stronger monitoring be put in place both at the
country and regional levels. This should be the priority. In the absence of an
effective and well-functioning mechanism to monitor the outcomes, identify issues
and address implementation gaps, the risks of the AEC falling short of achieving its
targets. Strengthening the monitoring mechanism also requires improving coordi-
nation among national agencies at a country level.
Third, given the difficult and complex process of building a single market, it is
only logical that regional institutions be developed over time to enforce rules and
monitor progress of implementation. ASEAN has taken steps to develop its insti-
tutional support to integration, like the development of enhanced dispute settlement
mechanism. But more steps are needed. One critical institutional support is the
strengthening of mechanism for private sector consultation. Formal consultations
with private sector and regional authorities may still be used, but new strategies to
involve the private sector in the integration process should be explored. Moreover,
there is a need to enhance the monitoring mechanism of the AEC. The establish-
ment of the ASEAN Integration Monitoring Office within the ASEAN Secretariat is
a step in the right direction, but this has to be complemented by well-developed
mechanisms at the country level to ensure that monitoring is effectively carried out
both at the country and regional levels.
Finally, since regional economic integration is not an end by itself, but a policy
instrument designed to achieve development goals, greater macroeconomic and policy
coordination is needed. In particular, the coordination of trade and financial policies is
crucial to ensure that both policies support each other. It goes without saying that both
financial and trade integration should go hand in hand. To facilitate trade, financial
instruments are needed to hedge the risks of trade and investment flows. In the same
manner, financial integration is needed to facilitate specialization and exploitation of
economies of scale, which are related to trade. Without significant integration of
financial systems, deeper integration of trade and investment is unlikely to happen.
References
Tristan Kohl
14.1 Introduction
T. Kohl (*)
Faculty of Economics and Business, University of Groningen, Groningen, The Netherlands
e-mail: t.kohl@rug.nl
14.2 Literature
During the past decade, numerous studies have surfaced in which scholars inves-
tigate particular features embodied in TAs. While some provide in-depth coverage
of one or two policy areas of interest, others develop measures that capture the
broader variation in a TA’s design. The former include, but are not limited to,
studies on the design of trade dispute settlement mechanisms (McCall Smith 2000),
investment provisions (Lesher and Miroudot 2006; Houde et al. 2007), trade in
services (Houde et al. 2007; Roy et al. 2007; Fink and Molinuevo 2008) or the type
of integration documented in the TA (Mansfield et al. 2008).
For example, McCall Smith (2000) looks into the governance structures of
62 regional trade pacts, in particular with respect to the design of the dispute
settlement mechanisms. For each TA, the author develops an indicator for the
level of legalism in the dispute settlement procedure. By construction, the level
of legalism is non-existent, low, medium, high or very high. Agreements with a low
level of legalism imply that participating states must settle disputes through diplo-
matic negotiations, while agreements with a high level of legalism are subject to
(binding) third-party arbitration. A number of underlying provisions are examined
to construct this measure of legalism, based on the presence of an independent
third-party review of disputes; whether third-party rulings are binding; whether
arbitrators are organized on an ad-hoc basis or as a standing tribunal; whether
individuals or bodies other than participating states have access to the dispute
mechanism; and the nature of remedies than can be imposed. Based on these
indicators, the author documents extensive variation in how states design their
dispute settlement mechanisms in TAs. Interestingly, he finds that economically
stronger states prefer lower levels of legalism and that this outcome prevails in
negotiations with partners that are more dependent on trade.
Another illustration of provision-specific studies is Houde et al. (2007), who
investigate provisions on investment and trade in services for 20 TAs. The authors’
methodology enables them to compare liberalization commitments following a
North American Free Trade Agreement (NAFTA)-inspired “negative list”
approach and those using a General Agreement on Trade in Services (GATS)-
inspired “positive list” approach. The extent to which investment provisions are
covered is captured in terms of investment liberalization, investment protection and
investment promotion, cooperation and facilitation. Provisions on trade in services
are measured based on the presence of commitments on market access, national
treatment, most favored nation treatment, temporary movement of natural persons,
recognition of qualifications and the abolition of monopoly service providers, to list
just a few. Among their findings, the authors demonstrate that the choice for either
the “positive list” or “negative list” approach is determined by factors such as the
desired degree and speed of liberalization, participants’ prior experience and
administrative capacity. However, the approach adopted by the negotiators does
not limit the extensiveness of the final degree of liberalization agreed upon in
the TA.
14 Opening the Black Box of Trade Agreements 303
Finally, the World Trade Report (WTO 2011) adopts Horn et al. (2010)‘s
methodology to identify policy areas in 96 trade agreements worldwide. The Report
shows that (legally enforceable) WTO+ provisions on tariff liberalization, intellec-
tual property rights and investment abound, while WTOX provisions on competi-
tion policy and capital mobility have also become more popular.
Our reading of the literature has so far demonstrated that a number of insightful
contributions have recently emerged in which scholars have started to tackle the
issue of trade agreement heterogeneity. Depending on the research question at
hand, (parts of) trade agreements can be successfully coded to reveal their design
and credibility. We now turn to a more thorough demonstration of how the design
of 296 TAs has recently been accounted for in a new publicly available dataset on
trade agreement heterogeneity.
14.3 Coding
1
The dataset is available at http://www.tristankohl.org
14 Opening the Black Box of Trade Agreements 305
nine indicators of the TA’s institutional quality (IQ) are also obtained. After the
policy domains that need to be coded have been identified, we can proceed by
reading the TAs in GPTAD and double-checking in the actual legal treaty.
Building on Horn et al. (2010), policy domains for which the agreement contains
a provision are coded 1 and 0 otherwise. In order for a provision to be considered
“covered” (C) and scored 1, all that is needed is for the provision to reflect
agreement by both parties to somehow cooperate with a view of trade liberalization.
The issue of legal enforceability is not relevant at this stage. So, a provision calling
for an exchange of Parties’ information on their environmental policies would score
a 1, but so would provisions that give rise to obligations to protect natural resources.
The odd provisions that only state that Parties reserve the right to protect their
natural resources are scored 0 because such measures are essentially protectionist
and do not require any form of cooperation.
A provision that is also considered to be legally enforceable scores 1 for
“enforceability” (E). The criteria build on those laid down in Horn et al. (2010).
These provisions typically use the word “shall”. For example: “Parties shall grant
service providers treatment no less favorable than that accorded to their own.”
Timing is also important. A provision calling for gradual liberalization of govern-
ment procurement policies, without indicating the date by which the liberalization
must be complete, scores 0. This is because it is unclear when the Party must be able
to meet that particular requirement. Provisions stating that Parties “shall negotiate”,
“shall consider” or “shall cooperate” are also difficult to be enforced. Negotiations
may still fail and not abolish trade barriers. It also seems very unlikely that it would
be able to prove that Parties have not given due consideration to a matter or that
they have not cooperated. Note, however, that all IQ provisions are considered fully
legally enforceable because they provide the underlying organizational mechanism
that is needed to implement the agreed upon commitments, including consultations
and dispute settlement.
For clarity and ease of replication, several excerpts from actual trade agreements
are provided below. Table 14.1 shows examples of WTO+ provisions and WTOX
provisions and classifies them as being either covered but not legally enforceable,
or both covered and legally enforceable.
Upon completion of the coding exercise, Kohl et al. (2013) construct an index of
trade agreement heterogeneity,
1 X X X
I Az ¼ WTOþ ,z
A =13 þ WTOXA, z =4 þ IQA =9 ð14:1Þ
3
Trade
Agreement
IE
IC
IPR
SPS
STE
TBT
Labor
Services
State Aid
Objectives
Investment
Definitions
Agriculture
AD & CVM
Competition
Environment
Consultations
Transparency
Capital Mobility
Plan & Schedule
Dispute Settlement
Export Restrictions
Import Restrictions
Public Procurement
Evolutionary Clause
Customs Administration
Institutional Framework
Chart 14.1 Examples of heterogeneous trade agreements (Source: Kohl et al. (2016). Notes: Undertakings not covered are white; those that are covered
(ignoring legal enforceability) are grey; provisions that are covered and legally enforceable are black)
313
314 T. Kohl
empirical results obtained in the literature thus far indicate that trade agreement
heterogeneity matters when examining, for example, their political origins and
economic outcomes. Nevertheless, coding practices are subject to a number of
challenges that call for consideration.
First, despite efforts to promote the transparency of the coding method used,
coding practices may be subject to observer bias. Having one individual code the
entire dataset is one option (Kohl et al. 2016), but methods do exist to check for
consistency among multiple scorers (Baccini et al. 2014).
Second, the extent to which a provision is credible or legally enforceable is a
matter of treaty interpretation. For example, dispute settlement rulings of the WTO
Appellate Body have shown that obligations may arise from statements using the
word “should” instead of “shall” (WTO 2011). Indeed, it is important to acknowl-
edge that there are limitations to the extent to which the legal enforceability of an
undertaking can be determined with absolute certainty. However, we have shown
that it is useful to differentiate, in one way or another, between those undertakings
that instill in the reader some sense of concrete and imminent policy liberalization
and those that merely reflect a loosely defined agreement to explore possible
avenues of future cooperation.
Third, the de jure content of TAs does not necessarily have to be the same as de
facto practices. For example, countries may not need a trade agreement with trade
partners because they liberalize trade unilaterally. Another issue is that provisions
not subject to dispute settlement may still be enforceable through political and
diplomatic channels. On the other hand, it may not at all times be possible to
enforce provisions that are subject to dispute settlement, due to political, non-legal
and/or resource considerations.
Fourth, provisions that are excluded from the agreements’ dispute settlement
system may still be subject to dispute settlement flowing from commitments that
the Parties may have elsewhere. This argument applies not only to WTO+ pro-
visions related to other commitments at the WTO, but also to WTOX commitments
arising from, for example, international treaties on labor standards and environ-
mental protection. Alternatively, the legal enforceability of a provision that allows
the use of countermeasures to enforce rights or obligations in one agreement may be
limited by commitments stemming from other agreements. For example, invest-
ment provisions in bilateral investment treaties (BITs) may overrule investment
measures in trade agreements. Future research may address this issue by further
looking into overlapping provisions across several economic cooperation agree-
ments (Estevadeordal and Suominen 2008).
Finally, there are different approaches to how indicators are weighted in the
construction of overall indices of regulation. Where Hicks and Kim (2012) use
subjective weights, Kohl et al. (2016) assume equal weights for all indicators in the
absence of a clear theoretical motivation to do otherwise. Useful robustness checks
would include experimenting with different weights and examining the relevance of
underlying (sub)components.
In conclusion, trade agreement heterogeneity matters in terms of our under-
standing of the origins of international economic integration and in terms of its
14 Opening the Black Box of Trade Agreements 315
References
Baccini, L., Dür, A., & Elsig, M. (2014). The design of international trade agreements: Introducing
a new dataset. Review of International Organizations, 9(3), 353–375.
Baier, S. L., & Bergstrand, J. H. (2007). Do free trade agreements actually increase members’
international trade? Journal of International Economics, 71, 72–95.
Estevadeordal, A., & Suominen, K. (2008). Sequencing regional trade integration and cooperation
agreements. The World Economy, 31, 112–140.
Fink, C., & Molinuevo, M. (2008). East Asian preferential trade agreements in services: Liberal-
ization content and WTO rules. World Trade Review, 7, 641–673.
Hicks, R., & Kim, S. Y. (2012). Reciprocal trade agreements in Asia: Credible commitments to
trade liberalization or paper tigers? Journal of East Asian Studies, 12, 1–29.
Horn, H., Mavroidis, P. C., & Sapir, A. (2010). Beyond the WTO? An anatomy of EU and US
preferential trade agreements. The World Economy, 33, 1565–1588.
Houde, M., Kolse-Patil, A., & Miroudot, S. (2007). The interaction between investment and
services chapters in selected regional trade agreements (OECD Trade Policy Paper 55).
Paris: OECD.
Kohl, T. (2013). I just read 296 trade agreements (UNU-CRIS Working Paper W-2013/9). Bruges:
United Nations University Institute on Comparative Regional Integration Studies.
Kohl, T., Brakman, S., & Garretsen, H. (2016). Do trade agreements stimulate international trade
differently? Evidence from 296 trade agreements. The World Economy, 39(1), 97–131.
Lesher, M., & Miroudot, S. (2006). Analysis of the economic impact of investment provisions in
regional trade agreements (OECD Trade Policy Paper 36). Paris: OECD.
Mansfield, E. D., Milner, H. V., & Pevehouse, J. C. (2008). Democracy, veto players and the depth
of regional integration. The World Economy, 31, 67–96.
McCall Smith, J. (2000). The politics of dispute settlement design: Explaining legalism in regional
trade pacts. International Organization, 54, 137–280.
Rose, A. (2004). Do we really know that the WTO increases trade? American Economic Review,
94, 98–114.
Roy, M., Marchetti, J., & Lim, H. (2007). Services liberalization in the new generation of
Preferential Trade Agreements (PTAs): How much further than the GATS? World Trade
Review, 6, 155–192.
Tinbergen, J. (1962). Shaping the world economy: Suggestions for an international economic
policy. New York: The Twentieth Century Fund.
World Bank. (2011). Global preferential trade agreements database. http://wits.worldbank.org/
gptad. Accessed 18 Dec 2011.
WTO. (2011). The WTO and preferential trade agreements: From co-existence to coherence.
Geneva: World Trade Organization.
Chapter 15
Assessing Globalization and Regionalization
Through Network Indices
15.1 Introduction
P.L. Iapadre
Dipartimento di Ingegneria Industriale e dell’Informazione e di Economia, Universita
dell’Aquila, L’Aquila, Italy
UNU-CRIS, Bruges, Belgium
e-mail: lelio.iapadre@univaq.it
L. Tajoli (*)
Dipartimento di Ingegneria Gestionale, Politecnico di Milano, Milan, Italy
e-mail: lucia.tajoli@polimi.it
network allows to emphasize the relationship between the countries in the network
and the structure, or the systemic features, of the network itself. This is exactly the
purpose of network analysis. In fact, both graph theory and NA place more
emphasis on the relationship between vertices in the graph, and on the structure
of the system itself, than on the vertices’ attributes, that are generally left in the
background. The application of NA to international trade can, therefore, nicely
complement other empirical analyses of trade, which instead put countries’ char-
acteristics or dyadic relationships at the forefront of the analysis, and that, even if
recognizing the importance of the structure of the system, leave this structure in the
background.1
In particular, NA can contribute to the analysis of changes in the level of
integration and connectivity of the world trading system as a whole: as international
trade links shift and re-arrange, this would become evident through the change of
the network structure. The extent of these changes over time is the first thing we
present, using some of the most common network indices to measure its cohesion
over time. In addition, we show how network analysis can be used to address some
issues debated in the recent trade literature, like the degree of integration within
regional blocks, and the position of single countries within the network structure.
The results obtained through this analysis provide a measure of trade integration at
the world level, showing that the world is still far from being fully connected, but
that full connection (or network completeness) is already evident in some
sub-regional components of the World Trade Network. This evidence also indicates
a strong heterogeneity in the countries’ level of connection and position. We
conclude by mentioning how these results can change when network analysis is
conducted at the product level.
In order to define the structure of the World Trade Network (WTN), we need
observations on bilateral trade flows between countries. In fact, in our network,
each country represents a node of the network, and links between nodes are given
by trade flows directed from one country to another. In the analysis of the World
Trade Network, we use aggregate bilateral imports, as reported by each country and
1
The literature analyzing world trade flows as a network is growing. Some earlier papers empha-
size the topological properties of this network, but lack to a very large extent the economic
interpretation of these findings. See for example, Bhattacharya et al. (2008), Fagiolo et al.
(2008), Garlaschelli and Loffredo (2005), Serrano and Boguna (2003). More recently, this type
of analysis was used to assess the economic implications of the properties of the international trade
network. See for example, De Benedictis and Tajoli (2010, 2011), and Chaney (2011). For a
general analysis of economic networks, see Jackson (2008).
15 Assessing Globalization and Regionalization Through Network Indices 319
measured in US dollars, drawn from the IMF Direction of Trade Statistics database,
as well as some sector-level data from the BACI-CEPII database.2
Observing the dataset, it is apparent that the bulk of the value of world trade
tends to be concentrated among a sub-group of countries and a small percentage of
the total number of flows accounts for a disproportionally large share of world
trade.3 Therefore, many of the observed patterns in trade flows are driven by a small
subsample of observations. In order to represent the entire system and evaluate all
existing trade links, which might be relevant for some small or peripheral countries
in spite of the very different orders of magnitude, we analyze the international trade
system both as a binary network, assigning a value of 1 to all existing trade flows,
and zero to all the missing bilateral links, and as a weighted network, maintaining
the dollar value of trade flows.
In Tables 15.1 and 15.2, we compare some of the trade network characteristics
over time, based on different observation samples. In Table 15.1, we included in the
network in each year only the countries for which at least one trade flow was
recorded, i.e. excluding unconnected countries, or considering only the so-called
“giant connected component”, that is the sub-network formed by the group of nodes
that are all connected to each other, directly or indirectly. While this might reduce
the overall number of errors in the observations,4 at the same time, it makes it more
difficult to compare the trade network over time because of the inherent change in
its structure given the changing number of vertices. Therefore, we computed the
network indices also for the balanced panel composed of the constant subset of
113 countries for which observations are available in the entire time span, and these
are reported in Table 15.2.
The first measure of integration or connectivity in the network is given by the
number of existing links (or bilateral trade flows) between countries. Given that
trade flows in our dataset are reported by importers, we can directly calculate the
2
The choice of the trade data is not neutral for describing the network. Even if the origin of all data
is the same database, a long time span might introduce some discontinuities. A number of
countries (especially the smallest and poorest ones) are not always reporting trade data. Additional
problems in assessing our dataset come from the fact that over time new countries were born
(e.g. the Czech Republic and Slovakia), and a few disappeared (e.g. Yugoslavia), changing the size
of the network. We use import data because they are more reliable in terms of coverage and
completeness, but import data can give rise to a network structure that is slightly different than the
one found with exports – as shown by Kali and Reyes (2007) and by De Benedictis and Tajoli
(2008) – or with average trade flows (the average of exports and imports). The same is true in a
gravity context; see Subramanian and Wei (2007).
3
In 2000, 90% of total trade value corresponded to 855 flows out of nearly 12,000 registered trade
flows, and only 82 countries out of the 157 reporting countries generate again 90% of trade.
4
Working at the aggregate level, we are aware of the fact that apparently unconnected countries
(for example Malta or United Arab Emirates, showing no link at all in some years) are in fact not
reporting data and the missing links therefore do not indicate that the country does not trade at all.
By consequence, removing vertices without any link will eliminate a few meaningful (but
unobserved) links and some meaningless zeros, but it should not introduce a systematic bias,
even if it changes the size of the network.
320 P.L. Iapadre and L. Tajoli
Table 15.1 Trade network indices over time (all reporting countries are included)
1960 1970 1980 1990 2000 2008
No. countries 113 130 143 145 157 154
No. arcs 3655 6593 8180 10289 11938 15850
Density 0.289 0.393 0.403 0.493 0.487 0.673
In-degree 0.601 0.565 0.58 0.511 0.519 0.329
centralization
Out-degree 0.546 0.51 0.438 0.469 0.484 0.329
centralization
Closeness 0.678 0.57 0.614 0.535 0.581 0.353
centralization
Betweenness 0.063 0.036 0.032 0.016 0.016 0.005
Centralization
Diameter 3 3 2 2 2 2
Aver. distance 1.657 1.567 1.52 1.486 1.497 1.327
among reachable
pairs
Source: Elaboration on IMF data
Note: Reporting countries included in the computations are the ones for which at least one trade
flow is recorded
number of incoming links for each node of our network, the so-called in-degree of
countries, but of course we can also compute the number of outgoing links, or
out-degree of each node, as we know the origin of each import flow.5 Looking at the
number of trade links among countries, measured as the number of arcs, we observe
that this has increased significantly over time.
From the number of links, an overall measure of the connectivity in the network
can be computed. This is called the density of the network, and it is defined as the
ratio between the number of existing links and the maximum number of potential
links in the given network. We observe also an increasing trend in the density of the
network in all the samples presented. Density declines slightly in 2000 compared to
ten years earlier, but this is explained by the increase in the size of the trade network
in terms of vertices.6 Density increases quite sharply again in all our samples
in 2008.
The rising trend in the network density confirms what other measures of eco-
nomic integration indicate, that linkages between countries have been increasing in
the second half of the twentieth century. Here we consider the number of linkages,
and we are not weighting for the value of trade carried by each flow, therefore this
indicator is showing something different than the standard measures that consider
openness at the individual country level. An increase in density means that on
5
We include all the definitions of the indices in the Annexes.
6
Larger networks are expected to have a lower density, because an increase in the number of
vertices requires a much more than proportional increase in the number of links to keep the density
constant.
15 Assessing Globalization and Regionalization Through Network Indices 321
average each country has a larger number of trade partners, and that the entire
system is more intensely connected. Still in 2000, though, the density index,
computed for all reporting countries, is around 0.5 and reaches 0.67 in 2008,
meaning that the network is relatively dense, but far from being complete. In
other words, this means that most countries do not trade with all other potential
partners, but that they are rather selective.
Other possible measures of integration of the system are given by the diameter of
the network and the average topological distance between nodes of the network,
also reported in the Tables. These indices measure the maximum number of steps
needed to go from one node to any other (the network diameter) and the average
number of steps between any two nodes. All the samples indicate that the network
diameter reduced from 3 to 2 in the 1970s, and that the average distance between
countries declined quite regularly over time. In other words, if we look at the world
as a network of trade linkages, the world has indeed become smaller in the past
decades.
The change in density or distance was not uniform across the network, and the
network structure changed over time also in other respects, as it can be assessed by
considering not only the number of links, but also their distribution, as well as other
network indices. Some of these indices suggest that over time the connectivity in
the network has become more evenly distributed. This assessment can be made
through centralization indices, ranging between zero and one, and measuring to
what extent the network is built around some nodes working as hubs (betweenness
centralization index), or whether there are nodes in central positions, closely
connected to a large number of other peripheral nodes (closeness centralization
index). The notion of betweenness centrality has important strategic implications
322 P.L. Iapadre and L. Tajoli
(Borgatti 2005). The central vertex could, in fact, exploit its position to its
advantage.
The decline in the betweenness centralization index, Cb, in all the tables from
1960 to 2008 implies that the increase in trade linkages has been fairly widespread,
reducing the role of hubs in the network. In line with this evidence is the trend in
degree centralization, especially evident in Table 15.2, with a fixed number of
countries. Closeness centralization, Cc, which is also influenced by the size of the
network, moves in a less regular fashion.
Do the measures discussed above indicate that integration is high or low? What we
have is a clearly increasing trend over time, but in order to assess this result, we
should know which are the predictions of international trade models in terms of the
structure of the trade network. Unfortunately, most trade models deal with the
pattern of trade of individual countries, and do not have much to say about the
structure of the whole system, and about the number of trade flows that we should
observe between countries. We know that perfect integration in theory would
require absence of trade costs and of any other barrier, even if we do not expect
to observe it in the data. As stressed by Brahmbhatt (1998), the general difficulty in
measuring economic integration using any type of outcome indicator is to provide a
standard against which actual outcomes in the real world can be judged. To do this
would require very strong assumptions, and therefore when we observe a gap
between actual results and theoretical standards, it is very difficult to tell if the
difference is due to underlying unrealistic assumptions or if it can indeed be
interpreted as the distance of the real world from full integration.
A different approach to assess the results obtained for the whole world network
is to compare them to the results obtained for different subsets of countries. In this
way, rather than setting an absolute benchmark, we can compare the average level
of world integration with the indices computed for specific groups of countries or
specific cases. In what follows, we show how the statistical tools developed in
network analysis can be applied to the study of trade regionalization. Drawing also
from the evidence emerging from other works, we compare the indices for the entire
world trade network with those obtained for groups of geographically nearby
countries, and for trade flows in specific sectors.
15 Assessing Globalization and Regionalization Through Network Indices 323
Assessing whether the degree of integration we found is high or low may be done by
checking if in the world there are regional areas at different levels of trade
integration. Here we verify if there are more trade flows between (relatively)
geographically close countries that belong to the same continent and even more
between countries that are parties to a regional trade agreement. To do so, we
present some of the characteristics of continental sub-networks of trade, reported in
Table 15.3.
If we consider density as an indicator of trade integration within each continental
sub-network, we see that both in 1980 and in 2000 the density of trade flows in each
continent – with the notable exception of Africa – is significantly higher than at the
world level, implying that among countries belonging to the same continent there
are proportionally more trade flows than with a random country elsewhere in the
world. In this respect world trade is indeed regionalized.7 However, we can also see
that over time, the density index within some continents declines, while world
density tends to increase. Europe, for instance, is close to being a complete network
in 1980, while in 2000 its density index is much lower, due to the increase in the
number and heterogeneity of trading countries after the Soviet era.
A further important push toward tighter integration for specific groups of
countries can be the existence of preferential trade agreements (PTAs), removing
barriers between partner countries signing such agreements. In the case of Europe,
the affiliation to the European Union (EU) seems indeed to play this role, as the EU
sub-continental area is a complete network with density equal to one, showing the
strength of the economic links between EU members. The fact that this indicator
can reach its maximum when computed with real data for a given group of countries
shows that there is still room for increasing the network density and the overall level
of integration at the regional and global level.
Besides density, several concepts used in network analysis can be adapted to the
study of regional trade flows, in order to better understand the topology of their
networks (Iapadre and Tironi 2009). For example, at the country level, what matters
is not only the node’s degree (the number of partners), which can be seen as a
measure of geographic diversification of regional trade, but also the average degree
of a node’s partners, which shows to what extent a country is connected to the most
important nodes of the network. Even more, for any given average partner degree, it
can be interesting to check if a country’s partners tend to be connected between
each other more than with other nodes of the network (clustering). In our context, it
tells us what the probability is that the trading partners of a country are themselves
trading partners. Table 15.4 reports some descriptive statistics of the clustering
coefficients in our sample. Generally, these clustering coefficients display an
7
This finding is in line with the evidence gathered through gravity models, showing that geo-
graphical distance is important in trade relations, as well as sharing a border and other proximity
indicators.
324
increasing trend over time. This seems to indicate that the increasing integration
that we observed in the world trading system has occurred especially through a
gradual expansion of links in the existing ‘neighborhoods’.
Conversely, the degree centrality of a node in a network is the higher, the lower
the number of connections among its partners, so that a central node plays the role
of a hub in the network. Correspondingly, the centralization of a network is often
measured as the highest centrality achieved by any of its nodes. The maximum
degree of centralization is reached in a star network, where only one country is
connected with all the others, whereas each of the others is connected only with the
hub of the network.8 Several concepts have been proposed to define centrality and
centralization in network analysis, giving rise to a wide range of different indicators
(see below).
Since regional integration areas are made of a small number of neighboring
countries, density and clustering indicators are normally very high, whereas cen-
trality measures tend to be very low. So, the interest of these simple applications of
binary network analysis (BNA) is limited to the few regions in which they give
anomalous results, such as in South Asia, or to data at product level, for which even
regional networks are not necessarily dense.
As mentioned above, BNA is based only on the number of trading partners and
neglects the intensity of their linkages. Motivated by the observation that the world
trade network is concentrated in a relatively small number of high-value bilateral
flows, weighted network analysis (WNA) of international trade represents the
intensity of linkages among the network nodes through the actual matrix of their
bilateral trade flows, expressed in absolute or relative terms.
Apart from the difference between the respective matrices, indicators used in
WNA are similar to those used in BNA. In the analysis of regional networks, the
node’s degree is replaced by the intra-regional node value, which is the value of a
country’s total trade with its region. Other WNA indicators, based on the
corresponding binary measures of average partner degree and clustering, can be
used to illustrate the topology of regional trade networks in terms of connectivity
8
A similar image is sometimes used to describe the network of preferential trade relationship
between the European Union and its partner countries, particularly in developing regions, which is
depicted as a hub-and-spoke system. The lack of preferential agreements among the spokes of the
system is sometimes considered as a factor that can inhibit their ability to reap the benefits of their
integration with the EU.
326 P.L. Iapadre and L. Tajoli
and centralization, taking into account not only direct bilateral trade flows between
a country and its regional partners, but also trade among the latter.
However, whereas in BNA the total number of possible linkages offers a natural
criterion for building normalized indicators, such as density measures, in WNA the
choice of a benchmark is more difficult, since there is no exogenously given
maximum value for trade. A widespread practice is to normalize indicators with
respect to the actual maximum value of bilateral trade in the network. However, this
choice is questionable, because it implies that the maximum levels of some
indicators could be reached only if all bilateral trade flows in the network were
equal to their maximum, which is clearly implausible given the large disparities in
country size.
Another possibility is to use revealed trade preference indicators, based on a
geographic neutrality criterion derived from the logic of trade intensity indices. A
node of a regional trade network is said to exhibit an intra-regional trade preference,
if its intra-regional trade share is higher than the region’s share of trade with the rest
of the world. The corresponding indicator at the network level may be called
regional trade introversion (Iapadre 2006).
Also for regional trade networks, it is useful to look beyond overall integration
measures to understand the potential asymmetry of the system. In a recent work,
Iapadre and Tajoli (2014) consider the centralization of some regional PTAs and the
centrality of their respective members to assess additional characteristics of the
regional process of integration, finding that the position of countries within a PTA is
often very different, some of them playing the roles of export hubs, others of
dominant suppliers of the regional system.
Table 15.5 Trade networks characteristics in the year 2000 for different sectors
In- Out-
% on tot. Aver. degree degree Betweenness
ISIC Trade
code Industry value Arcs Density Degree Central Central Central
311 Food 4.5 12710 0.259 57.25 0.485 0.644 0.049
products
313 Beverages 0.7 7078 0.144 31.88 0.432 0.705 0.076
314 Tobacco 0.3 3651 0.074 16.45 0.266 0.612 0.103
321 Textiles 3.5 12829 0.262 57.79 0.510 0.619 0.046
322 Apparel 3.9 11475 0.234 51.69 0.561 0.661 0.062
323 Leather 0.7 7842 0.160 35.32 0.453 0.644 0.061
products
324 Footwear 0.7 6623 0.135 29.83 0.437 0.633 0.104
331 Wood 1.1 8884 0.181 40.02 0.527 0.641 0.071
products
332 Furniture 1 8522 0.174 38.39 0.507 0.685 0.099
341 Paper 2.5 8984 0.183 40.47 0.407 0.666 0.078
products
342 Printing and 0.7 9585 0.195 43.18 0.472 0.695 0.064
publishing
351 Industrial 7.9 11627 0.237 52.37 0.439 0.662 0.054
chemicals
352 Other 4.1 11748 0.240 52.92 0.437 0.669 0.048
chemicals
353 Petroleum 2.8 5338 0.109 24.05 0.341 0.627 0.109
refineries
354 Petroleum 0.1 2689 0.055 12.11 0.186 0.559 0.069
and
coal prod.
355 Rubber 0.9 8951 0.182 40.32 0.385 0.676 0.078
products
356 Plastic 1.2 10286 0.210 46.33 0.399 0.662 0.060
products
361 Pottery, 0.2 6604 0.135 29.75 0.419 0.647 0.081
china,
earthenware
362 Glass 0.5 7979 0.163 35.94 0.364 0.678 0.060
products
369 Other 0.6 7846 0.160 35.34 0.367 0.689 0.066
non-metal
mineral
products
371 Iron and 2.9 8465 0.173 38.13 0.359 0.668 0.055
steel
372 Non-ferrous 2.8 7619 0.155 34.32 0.449 0.662 0.094
metals
381 Fabricated 2.9 12195 0.249 54.93 0.496 0.655 0.048
metal
products
(continued)
15 Assessing Globalization and Regionalization Through Network Indices 329
(2010) also shows that more complex goods in terms of their production processes9
are associated to more dense and articulated trade networks, suggesting the exis-
tence of an international organization of production stretching over many countries.
At the industry level, trade integration seems to be driven also by the involvement
of countries in international production networks that give rise to a number of trade
links in intermediate goods.10
15.4 Conclusion
In this chapter, we have shown how the tools of network analysis can be used to
examine the cohesion of the international trading system. Through the indices
describing the network’s properties, such as density, closeness, betweenness and
degree distribution, we see that the world trade network has indeed changed in the
past decades. In particular, the trading system has become more intensely
interconnected, while the heterogeneity among countries increased; the average
9
Goods’ complexity here is following the definition given by Nunn (2007), which refers to the
number of intermediate inputs necessary to produce a final good.
10
For further analysis of the international trade network at the sector level, see De Benedictis et al.
(2013). The paper is also providing a large set of centrality indicators for individual countries, to
understand countries’ position in the WTN and their evolution over time.
330 P.L. Iapadre and L. Tajoli
structural network distance has decreased, and the position of many countries in the
network changed.
Furthermore, network analysis can show how trade policies play a role in
shaping the trade network, at global and regional levels. An important feature of
these results is that they pertain to the trading system as a whole, which is the object
of analysis in this context, and are not due to a specific country or group of
countries. The main contribution of NA to the empirical investigation of trade
flows is probably that it offers a unified view of the system characteristics, drawing
from the underlying heterogeneity of its components and its complexity. This
approach can have relevant implications both for trade policy and for the modeling
of trade relations.
Annexes
N ¼ ðV; L; W; PÞ ð15:1Þ
Therefore, two vertices are said to be adjacent (or at just one-step distance) if
they are connected by a line. The concept of geodesic distance in networks refers to
the number of steps needed to connect two vertices Vi and Vj, named δij. The
shortest the distance between two vertices the closest is the connection
between them.
Weighted networks add to simple graph some additional information on the lines
of the graph. The additional information is contained in the line value function W,
where line values are positive weights associated to each line, usually indicating the
strength of the relation. In the ij case, wij is the link’s weight. The additional
information on the vertices is contained in the vertex value function P, assembling
different properties or characteristics of the vertices.
15 Assessing Globalization and Regionalization Through Network Indices 331
The size of a network is expressed by the number of vertices n and the number of
lines m. The set of vertices that are connected to any given Vi defines its neighbor-
hood where d 0 denotes the number of neighbors of Vi. Since, in simple directed
graphs, a vertex can be both a sender and a receiver, the indegree of a vertex is the
number of arcs it receives, and the outdegree is the number of arcs it sends. The
notion of neighborhood is associated to the one of clustering. The clustering
coefficient of a vertex Vi is the proportion of a vertex’s neighbors which are
neighbors of each other. The clustering coefficient for the network as a whole can
be derived taking a weighted or an unweighted average across vertices in the
network.
and the higher the variation in the degree of vertices, the higher the centralization of
a network. The degree centralization of any regular network is 0, while a pure star
has a degree centralization of 1.
11
For a general discussion on the concept of centrality in networks, see Bonacich (1987).
12
The standardized degree centrality of a vertex is also often used, given by the vertex degree
divided by the maximum possible degree: Cisd ¼ n1 d
.
332 P.L. Iapadre and L. Tajoli
*
At the network level, if Cci is the centrality of the vertex that attains the
maximum closeness centrality score, the closeness centralization of a network is
(Freeman 1979; Goyal 2007):
P n c
c*
i¼1 Ci Ci
Cc ¼ ð15:5Þ
ðn 1Þðn 2Þ=ð2n 3Þ
A third notion of centrality often used in the literature is based on the intuition
that a vertex Vi is central if it is essential in the indirect link between Vj and Vk. A
vertex that is located on the geodesic distance between many pairs of vertices plays
a central role in the network, because it is necessary for all periphery vertices in
order to be mutually reachable. This concept of centrality is called betweenness
centrality. The betweenness centrality of vertex Vi is the proportion of all geodesic
distances between pairs of other vertices that include this vertex (Vega-Redondo
2007):
X δjki
Cib ¼ ð15:6Þ
j6¼k
δjk
where δjk is the total number of shortest paths joining any two vertices Vj and Vk ,
and δjki is the number of those paths that connect Vj and Vk through Vi. The core of a
star network has maximum betweenness centrality, because all geodesic distances
between pairs of other vertices include the core. In contrast, all other vertices have
minimum betweenness centrality, because they are not located between other
vertices.
The betweenness centralization is the variation in the betweenness centrality of
vertices divided by the maximum variation in betweenness centrality scores possi-
ble in a network of the same size:
P n b
*
i¼1 Ci Cbi
Cb ¼ 2 ð15:7Þ
n ðn 1Þ=ð2n 1Þ
15 Assessing Globalization and Regionalization Through Network Indices 333
The simplest indicator that can be used to analyze the structure of a regional trade
network is the intra-regional node degree (INDi),14 that is the number of regional
partner countries of each country i, which can be expressed in absolute terms, or as
an intra-regional density index (IDIi), that is as a ratio of the total number of
possible regional partner countries (n 1):
The intra-regional density index can be computed also for the entire region r,
where it measures to what extent the actual number of trade linkages corresponds to
its maximum potential level:
13
The indices presented here are developed by Iapadre and Tironi (2009) to which we refer
for further details.
14
In the network analysis literature, node degree is sometimes called neighborhood degree.
334 P.L. Iapadre and L. Tajoli
where A(i) is the ith row of the adjacency matrix A describing the network and 1 is a
unitary vector. The maximum level of IAPDi is reached when all country i’s
regional partners’ IDIj are equal to one, that is when all the possible n(n 1)
trade linkages exist. This allows us to define an average intra-regional partner
density index (IPDIi) as follows:
At the regional level IAPDr and IPDIr can simply be computed as the averages of
the corresponding country indicators.
An extra-regional average partner degree (EAPDi) and an extra-regional partner
density index (EPDIi) can be defined as follows:
EAPDi ¼ EAðiÞ OA 1 =ENDi ð15:14Þ
EPDI i ¼ EAPDi =ðm 1Þ ð15:15Þ
where EA is the nm adjacency matrix of linkages between the region’s members
and the benchmark area’s countries, and OA is the m m adjacency matrix of
linkages among the benchmark area’s countries.
Finally, a relative intra-regional partner density index (RIPDIi), ranging from 1
to 1 with a neutrality threshold of zero, can be computed as:
The fact that a country has a certain average partner degree does not necessarily
imply that all its partners are connected between each other. In order to capture this
feature of the network, a third indicator has been developed, named binary cluster-
ing coefficient (BCCi), aimed at measuring to what extent a country’s partners tend
to cluster into triangles, that is to trade between each other. BCC has also been used
to detect a possible hierarchic structure of the network.
The intra-regional binary clustering coefficient (IBCCi) can be defined as:
IBCCi ¼ A3 ii =½INDi ðINDi 1Þ ð15:17Þ
where (A3)ii is the i-th entry on the main diagonal of A A A. Given INDi, IBCCi
measures the actual number of bilateral linkages between country i0 s regional
partners, relative to its potential.15
Another useful concept is the degree of centrality, which is used to assess to what
extent trade linkages tend to concentrate towards one or more hub countries. The
maximum degree of centralization is reached in a star network, where only one
15
IBCCi can be computed only if INDi >1.
15 Assessing Globalization and Regionalization Through Network Indices 335
country is connected with all the others, whereas each of the others is connected
only with the center of the network.
Several indicators have been proposed to measure the centrality of a node and
the centralization of a network. At the country level, intra-regional node centrality
(INCi) can simply be measured as:
INCi measures to what extent a country is connected to regional partners that are
not connected between each other.
At the network level, an intra-regional centralization index (ICIr) can be defined
as:
The weighted network analysis (WNA) of international trade represents the inten-
sity of linkages among the network nodes through the actual matrix of their bilateral
trade flows (W) expressed in absolute or relative terms.17
Apart from the difference between the respective matrices, indicators used in
WNA are similar to those used in BNA. In our context, node degree is replaced by
intra-regional node value (INVi), which is the value of a country’s total trade with
its region. However, since there is no given maximum value for trade, a density
index similar to that used in BNA cannot be easily defined, and there are several
options to build a normalized INVi.
If we refer to the geographic neutrality criterion mentioned in the text (Sect.
15.3.1), we can introduce intensity and revealed trade preference indices into the
context of WNA. Since INVi refers to intra-regional trade, we can define extra-
regional node value (ENVi) as the total value of country i’s trade with the bench-
mark area, and the density index of BNA can be replaced by a homogeneous
(country-size-independent) trade intensity index HIir. More precisely:
HI ir ¼ Sir =V or ð15:20Þ
16
See Kali and Reyes (2007).
17
Fagiolo, Reyes and Schiavo (2008) show why WNA is more informative than BNA in describing
the world trade network.
336 P.L. Iapadre and L. Tajoli
where:
and finally the relative intra-regional revealed trade preference index (RIRTPir) can
be computed as:
where W[1/3] is the matrix obtained by raising each element of the W matrix to 1/3
and (W[1/3])3ii is the i-th entry on the main diagonal of W[1/3] W[1/3] W[1/3]. IWCCi
measures the intensity of trade among country i0 s regional partners relative to the
total number of their potential connections. So, it is positively related to the actual
density of these connections (IBCC) and to their intensity.
For any given INDi, the maximum level of IWCCi is not scale-independent. This
problem can be solved by dividing each element in the W matrix by their maximum,
which results into an intra-regional normalized weighted clustering coefficient
(INWCCi), ranging from 0 to 1:
3
INWCCi ¼ NW½1=3 =½INDi ; ðINDi 1Þ ð15:27Þ
ii
where NW is the matrix of trade flows within region r normalized with respect to
their maximum.
where R is the matrix of intra- and inter-regional trade flows, divided by their total,
Tr is the trace operator, and kR2k is the sum of all the elements of matrix R2.
IAC is equal to zero in the case of geographic neutrality, which is when regions
trade among each other in proportion to their total trade values, and reaches a
maximum value of one in the limiting case of no inter-regional trade. On the other
hand, in the limiting case of no intra-regional trade, the minimum (negative) value
of IAC is equal to – kR2k/(1 kR2k).18
References
Bhattacharya, K., Mukherjee, G., Saramäki, J., Kaski, K., & Manna, S.S. (2008). The international
trade network: Weighted network analysis and modeling. Journal of Statistical Mechanics:
Theory and Experiment, P02002.
Bonacich, P. (1987). Power and centrality: A family of measures. The American Journal of
Sociology, 92(5), 1170–1182.
Borgatti, S. P. (2005). Centrality and network flow. Social Networks, 27(1), 55–71.
Brahmbhatt, M. (1998). Measuring Global Economic Integration: A Review of the Literature and
Recent Evidence (Working Paper). The World Bank.
Chaney, T. (2011). The Network Structure of International Trade (NBER Working Paper, 16753).
De Benedictis, L., & Tajoli, L. (2008). La rete degli scambi mondiali. L’Italia nell’economia
internazionale. ICE 2007-2008. Istituto Nazionale per il Commercio Estero, Roma.
De Benedictis, L., & Tajoli, L. (2010). Comparing sectoral international trade networks. Aussen-
Wirtschaft, 65, 167–189.
De Benedictis, L., & Tajoli, L. (2011). The world trade network. The World Economy, 34,
1417–1454.
De Benedictis L., Nenci, S., Santoni, G., Tajoli, L., Vicarelli, C. (2013). Network Analysis of
World Trade using the BACI-CEPII database (CEPII Working Paper, 2013-24).
Fagiolo, G., Reyes, J., & Schiavo, S. (2008). On the topological properties of the world trade web:
A weighted network analysis. Physica A, 387, 3868–3873.
Freeman, L. (1979). Centrality in social networks: Conceptual clarifications. Social Networks, 23
(1), 215–239.
Garlaschelli, D., & Loffredo, M. I. (2005). Structure and evolution of the world trade network.
Physica A, 35, 138–144.
Goyal, S. (2007). Connections. An introduction to the economics of networks. Princeton: Princeton
University Press.
Iapadre, L. (2006). Regional integration agreements and the geography of world trade: Statistical
indicators and empirical evidence. In P. De Lombaerde (Ed.), Assessment and measurement of
regional integration (pp. 65–85). London: Routledge.
18
The minimum IAC of 1 (perfect disassortativity) is reached when Tr(R) ¼ 0 (no intra-regional
trade) and kR2k ¼ 0.5. The latter parameter depends on the distribution of extra-regional flows and
on the number of regions. It can be shown that kR2k is equal to 0.5 only for a two-region world
with no intra-regional trade. For a symmetric matrix with a number of regions larger than 2, the
minimum IAC is higher than 1 and grows with the number of regions.
15 Assessing Globalization and Regionalization Through Network Indices 339
Iapadre, L., & Plummer, M. (2011). Statistical measures of regional trade integration. In P. De
Lombaerde, R. Flôres, L. Iapadre, & M. Schulz (Eds.), The regional integration manual
(pp. 98–123). Abingdon/Routledge: Oxon.
Iapadre, L., & Tajoli, L. (2014). Emerging countries and trade regionalization. A network analysis.
Journal of Policy Modeling., 36(S1), 89–110.
Iapadre, L., & Tironi, F. (2009). Measuring trade regionalisation: The case of Asia (UNU-CRIS
Working Paper, 2009/9).
Jackson, M. O. (2008). Social and economic networks. Princeton: Princeton University Press.
Kali, R., & Reyes, J. (2007). The architecture of globalization: A network approach to interna-
tional economic integration. Journal of International Business Studies, 38, 595–620.
Newman, M. E. J. (2003a). Mixing patterns in networks. Physical Review E, 67, 026126.
Newman, M. E. J. (2003b). The structure and function of complex networks. SIAM Review, 45,
167–256.
Nunn, N. (2007). Relationship-specificity, incomplete contracts, and the pattern of trade. Quar-
terly Journal of Economics, 122(2), 569–600.
Piccardi, C., & Tajoli, L. (2012). Existence and significance of communities in the world trade
web. Physical Review E, 85, 066119.
Serrano, M. A., & Bogu~ na, A. (2003). Topology of the world trade web. Physical Review E, 68,
015101.
Subramanian, A., & Wei, S. J. (2007). The WTO promotes trade, strongly but unevenly. Journal of
International Economics, 72, 151–175.
Vega Redondo, F. (2007). Complex social networks. New York: Cambridge University Press.
WTO. (2010). International trade statistics. Geneva: World Trade Organization.
WTO. (2011). The WTO and preferential trade agreements: From co-existence to coherence. In
World Trade Report 2011. Geneva: World Trade Organization.
Chapter 16
Measuring Actual Economic Integration: A
Bayesian State-Space Approach
16.1 Introduction
Despite the academic and policy interest and in contrast with other aspects of
institutional economics (like governance), a systematic and standard index of
regional integration is still lacking. An index that gives a quantitative measure of
the level of regional integration is deemed useful, because it would allow to
determine the trends in the world economy more precisely (e.g. the link between
globalization and regionalization), to monitor integration policy initiatives more
accurately and to assess the effectiveness of current or past policy initiatives
(e.g. aiming at indicating good practices). Yet, in their review, De Lombaerde
et al. (2008) note that “only a few attempts have been undertaken to design
composite indices of regional integration and no proposal has been systematically
and continuously used as a policy tool.”
The most plausible explanations for this dearth are data availability and meth-
odological issues. Regional integration is a complex and multidimensional process
and therefore difficult to capture by a single or a few indicators. Consequently, a
larger set of data is used, usually of very different quality in which scoring by the
analyst is not uncommon. Interpretation and analysis of the data demands a
summary indicator that integrates the information of all the available data, which
immediately brings up the problem of how to summarize (i.e. aggregate) the
individual indicators and which weighting scheme to use. For example, Feng and
Genna (2003) follow Hufbauer and Scott (1994) in their construction of Integration
Achievement Scores by taking the simple arithmetic average of the categories that
measure distinct components of (institutional) regional integration. The index of
economic integration at the bilateral country level for the member countries of the
Organisation for Economic Co-operation and Development (OECD). For these
countries a large set of indicators of good data quality are available such that due
account can be taken of the multidimensionality of regional integration. Based on
this first application, we consider in the fourth section the extensions that the
method allows. Conclusions are drawn in the last section.
16.2 Methodology
This section only aims to give a limited overview of the state-space methodology.
For more information on state-space models and how to estimate them, see Kim and
Nelson (1999) or Durbin and Koopman (2012). More detailed information on this
particular model can also be found in Standaert (2014) where it is used to combine
indicators of corruption.
The main idea in the state-space model is to estimate the unknown overall level of
regional integration RI (the state variable), using the information in the different
h i0
ð1Þ ðkÞ
indicators of regional integration, yi, t ¼ yi, t , . . . , yi, t : 1 In order to understand
how this happens, it is necessary to go back to the two equations that define its
workings: the measurement (Eq. 16.1) and state equation (Eq. 16.3).
1
For the sake of readability, the notation is sometimes simplified. y( j ) is a single indicator of
integration for all country-couples and all years. yi , t is the vector of all indicators in a given year
and for a given country-couple, while this vector for all years and all country-couples is simply y.
344 G. Rayp and S. Standaert
RIi,t. The error term Ei,t captures differences in the quality of the indicators whether
due to measurement errors or because of the influence of factors other than the level
of integration. The better an indicator y( j ) measures the level of integration, the
smaller the variance of the corresponding error term H( j, j ).
The ð1 kÞ vectors C and Z rescale the indicator variables to put them on equal
footing. The exact rescaling parameters are indicator-specific, but are kept constant
over time and country-couples. Similarly, each indicator can differ in terms of its
reliability, but the reliability of an indicator does not change over i or t if the
indicator is not missing (cf. infra).
The state equation (Eq. 16.3) allows the current level of integration to depend on
its past values. Especially in the case of natural trading partners or institutional
integration, we would expect to find a high level of persistence (with the exception
of major policy changes, institutional integration remains more or less unchanged).
As with the other parameters in the model, the level of dependence (Ti) is not
imposed ex ante but determined endogenously. It can vary for each country-couple
i, but is restricted to lie within the [1,1] interval. This rules out ever-increasing
values for the RI index and ensures that the model converges to a steady solution.
However, it does not preclude non-stationarity in the level of regional integration.
Figure 16.1 illustrates the advantage of adding the time-dependency in the state
equation. To the extent that the level of integration depends on its previous values,
both past and future information are used to predict what the level of integration is
today (step a). This prediction is governed by the state equation (Eq. 16.3). This
forecasted value is then compared to the indicators of integration today yi,t. Using
the parameter values in the measurement equation (Eq. 16.1) the estimated level of
integration is adjusted (step b). The stronger the time-dependence, the more impor-
tant step a becomes. The more reliable an indicator is, the bigger the influence of
step b is.
Because the RI i, t1 and RI i, tþ1 also depend on their past values and future values,
the entire time-series is used when estimating the current level of integration. The
advantages are manifold. First of all, it significantly increases the number of years
for which the indicator can be reliably computed. Moreover, the increase in
information helps the algorithm to better distinguish between random measurement
errors and the actual changes in the level of integration. This results in smoother
estimates made with smaller confidence bounds.
The strength of the state-space model is the ease with which it handles missing
observations. Simply put, missing observations are replaced by information which
has absolutely no value: y ¼ 0 and var(E) ¼ 1. This allows the model to run
uninterruptedly without fundamentally changing the value of missing data. More-
over, because the entire time-series is used when estimating the value of RI, it
negates the need to impute or otherwise manipulate missing data (Kim and Nelson
1999; Durbin and Koopman 2012).
An additional advantage of this model is that it encapsulates a number of other
techniques. For example, if we assume that RI does not depend on its previous
values (T ¼ 0) and all indicators have the same reliability (H( j, j ) ¼ cH), it can be
shown that this model will return a principle component analysis. If in addition it is
16 Measuring Actual Economic Integration: A Bayesian State-Space Approach 345
assumed that all indicators are scaled the same way (Zðj, 1Þ ¼ cZ and Cðj, 1Þ ¼ cC ),
then it returns a simple average.
In other words, the usefulness of the state-space approach follows directly from
the validity of the assumptions on the parameter values. If the level of integration is
not expected to depend on its previous values ( T ¼ 0 ) a principle component
analysis suffices. However, if these assumptions are incorrect, using simple tech-
niques will discard information and could lead to incorrect conclusions. It also
means that it is possible to test the validity of the state-space model ex-post using
the estimated parameter values.
In order to estimate the state-space model, it is necessary to solve for the level of RI
as well as the parameters of the state and measurement equation: H, Z, C and T. As
the number of countries and years increases, this estimation becomes more and
more cumbersome. However, using a Bayesian Gibbs sampler, it can be split up
into different sections that can be dealt with one at a time.
If the values for RI were known, the state and measurement equations would be
simple linear regressions and we could easily compute and draw from their distri-
butions. Similarly, if the parameters were known, we could draw from the distri-
bution of RI using a simulation smoother (Durbin and Koopman 2012). It can be
shown that by iteratively drawing from both conditional distributions while condi-
tioning on the last drawn value, these draws will converge to the unconditional
distribution. After discarding the first non-converged values (the burn-in), the
remaining drawn values can be used to reconstitute the original unconditional
distribution of RI as well as those of the parameters. For more information on
Bayesian econometrics and Gibbs sampling see Lancaster (2004) and Koop
et al. (2007).
Because this model is estimated in a Bayesian framework, it is necessary to be
explicit about the prior distribution of the parameters. However, seeing that there is
no ex-ante information on them we use flat priors. This means that these parameters
are not restricted in any way. The only variables that are limited are Ti, whose
values have to lie inside the [1,1] interval, and the diagonal elements of the
variance H, whose values have to be strictly positive.
346 G. Rayp and S. Standaert
pðCÞ 1 ð16:5Þ
pðZÞ 1 ð16:6Þ
p logðH ðj, jÞ Þ 1, 8j 2 f1, . . . , ng ð16:7Þ
pðT i Þ ¼ 0:5*1T2½1, 1 , 8i 2 f1, . . . , kg ð16:8Þ
This section illustrates the state-space approach by measuring the level of regional
integration between the members of the OECD. Specifically, it examines the level
of Actual Economic Integration (AEI) defined by Mongelli et al. (2005) as:
the degree of interpenetration of economic activity among two or more countries belonging
to the same geographic area as measured at a given point in time. (Mongelli et al. 2005:
p. 6)
This definition is relatively narrow and puts strict limits on the variables to be
included. It excludes institutional or cultural integration. Even within the perspec-
tive of economic integration, it focuses on actual interpenetration of activities.
Strictly speaking, this implies that co-movement of prices and GDP and other
factors from the optimal-currency-area theory should not be included. In addition,
it focuses on actual integration as opposed to measuring the potential benefits of
integration.
As a result, the AEI indicator computed here is relatively neutral. It does not rely
on any specific (economic) theory on integration, nor does it treat integration as
necessarily good or bad. It simply measures the extent to which the economic flows
of two countries are intertwined. Needless to say, this definition is one of the many
possible choices and the state-space methodology can be easily expanded to include
different aspects of regional integration.
The unit of analysis in this study is country couples, and their integration is
measured in a directional sense. In other words, the values of the index AEIB,A
express to what extent the bilateral economic flows between countries A and B are
important for country A.2 Allowing the values of AEIB,A to differ from AEIA,B
makes sense in a network where country size varies significantly. For example, that
the German-Estonian trade is important for Estonia does not necessarily imply that
the same holds for Germany.
2
In network theory the link (or edge) going from country (node) A to country B is denoted XB,A
(Newman 2010).
16 Measuring Actual Economic Integration: A Bayesian State-Space Approach 347
16.3.2 Data
3
Estonia (1991), Slovakia (1993), Slovenia (1993) and the Czech Republic (1991) are added later
to the sample.
348 G. Rayp and S. Standaert
Luxembourg Economic Union because the earliest data is often not available for
both countries separately. This gives us a total 33 times 32 or 1056 country-couples
whose level of integration is studied over a period of 41 years.
16.3.3 Results
The Gibbs sampler ran for 100,000 iterations, of which the first 50,000 were
discarded as burn-in. The remainder was used to reconstitute the distribution of
the AEI index. While the AEI index can be compared over time and over countries,
the exact values are arbitrary and can be transformed as long as the relative
differences remain unchanged. Throughout this paper, a high value of the AEI
index will mean that the level of integration is high.
Setting the variance of the annual change in integration to one (the variance of ν
in Eq. 16.3) gives us values for AEI that lie between 13 and 133. However, the
distribution is strongly skewed to the left: the mean and median are 7 and 10,
respectively. In other words, most countries have very low values of integration:
90% of country couples have negative levels of integration and less than one
percent has values greater than 35.
Figure 16.2 plots the level of integration and the 90% confidence bounds for four
country-couples. It illustrates many of the points made in the previous sections. For
example, while the bilateral flows between Mexico and the United States of
America (USA) are crucial for Mexico (they lie entirely within the top 95th
percentile of all AEI values), they are far less important for the USA. The same
holds for the flows between Austria and Germany, but to a lesser extent.
Secondly, we see that as the number of missing observations decreases, the
uncertainty bounds grow tighter. This is especially clear in the second panel. The
number of indicators for Austria-Germany increases from 4 (1970–1984) to
16 Measuring Actual Economic Integration: A Bayesian State-Space Approach 349
80 (2003–2008) falling back to 40 in 2009, and this is reflected in the size of the
confidence bounds (Fig. 16.2).
Returning to the individual indicators of integration, Fig. 16.4 graphs their
correlation with the AEI index. It shows the indicators from all four different
types of flows are highly correlated with the index, but that migration (represented
by diamonds) weighs the least. A possible explanation is the opposite influence that
different types of integration have on other flows. For example, economic migration
is expected to subside when high trade between countries creates opportunities in
the home country. On the other hand, the existence of many migrants from a
particular country increases the information on potential beneficial trade, implying
a positive correlation. To the extent that this is indeed the case, this might be
resolved by including indicators that are able to distinguish between different
motivations for migration: economic versus political migration, high-skilled versus
low-skilled, etc.
350 G. Rayp and S. Standaert
As mentioned, we can use the values for AEI to construct a weighted directed
network (Newman 2010). Using the python software package NetworkX, we can
show the shape of this network in every decade (Fig. 16.3). The values of the index
are reflected in the darkness of the arrow between the countries. The position and
color of the countries are determined by its weighted indegree: the more important a
country is for its partner countries, the darker its color and the more central its
position.4 The colors reflect only the values of the AEI index, which could lead to
the situation where a country that is strongly connected to a few small countries
takes up a central position on the network. To capture this, the position of the
country also depends on the size of the partner countries: the higher the level of
integration with countries with high GDP, the closer to the center a country is
positioned. Unsurprisingly, it shows that the most central players in the OECD are
the USA, Germany and the United Kingdom, with France, Italy and Japan as close
second.
When comparing the network’s structure over time, it becomes clear that trade
among the OECD countries is being condensed to a few central players. In 1970,
most countries had a lot of incoming and outgoing arrows and the level of AEI was
relatively similar between many country-pairs. In contrast the pattern in 2010
reveals a stark divergence between core and periphery countries. The former have
many incoming arrows representing high levels of integration, while the latter have
one predominant link to one of the core countries and fewer incoming arrows. The
standard deviation of AEI values reflects this: it rises from 7.32 in 1970 to 10.47 in
2010, falling back slightly to 9.93 in 2010. Similarly, the indegree (reflected in the
color of the countries) decreases for the periphery and increases for the core. It
should be mentioned that these changes only apply to the relative level of integra-
tion of countries. In other words, these results do not necessarily imply that trade
between peripheral countries has increased, only that the increase with core coun-
tries has been larger.
A second pattern in the evolution of the network is the rise and subsequent fall of
the USA. While it remains one of the core countries, its importance in the network
starts to decline in the year 2000. A possible explanation is the rise China and other
Asian countries as major trading partners outside the OECD, especially because this
pattern is also apparent for Japan. Conversely, during the same period Italy and
Spain take up a more central position in the network.
An often-used metric to study the overall connectivity of a network is the
network density. It is defined as the number of links (edges) between countries
divided by the total number of possible links ( nðn 1Þ with n the number of
countries). However, since AEI is a continuous variable, a threshold has to be
defined which separates the connected countries from the unconnected ones.
Figure 16.5 plots the network density with the threshold set at 10% lowest value
4
The weighted indegree of a country is the sum of the AEI index of all incoming arrows.
Fig. 16.3 Plot of the network formed by the AEI indicator (Countries are represented by their ISO
code, with the Belgium-Luxemburg Economic Union abbreviated to BLU. The color of countries
reflects the weighted indegree while level of the AEI index is reflected in the opacity of the arrow:
the darker the colors, the higher these values are. Finally, the position on the concentric circle of
the countries is determined by the importance of a country for its trading partners multiplied by the
GDP of those trading partners (weighted indegree with AEI*GDP sender as weights))
352 G. Rayp and S. Standaert
Fig. 16.5 Network density (The uninterrupted line represents the network density including 95%
confidence interval. The dotted line corrects for the inclusion of new countries by keeping the total
number of nodes constant over time)
(14.6). Because the Gibbs sampler returns the entire probability distribution of the
AEI index, it is possible to construct the 95% confidence interval for the network
density. It shows that the overall connectedness significantly decreased over time
(at 1% significance level). One possible explanation is that this result is driven by
the addition of a number of lesser-connected countries in the early nineties,
i.e. Estonia, the Czech Republic, Slovakia and Slovenia. To control for this, the
density was recomputed keeping the number of countries constant (the dotted line).
While this takes out the initial drop in the 1990s, it does not do away with the
decline in density in the 2000s. This is probably due to the concentration of trade,
which raises the AEI index for links between core and periphery, but lowers it
between peripheral countries. Seeing that there are more peripheral countries, the
overall connectivity declines.
This section briefly looks at the effect of the expansion of the European Union
(EU) and formation of the North American Free Trade Agreement (NAFTA) on the
AEI index. Do these trade agreements increase the level of actual economic
integration for the participating countries?
Table 16.2 shows the results of a difference-in-difference study of the level of
integration. The level of actual integration is regressed on two dummy variables EU
and NAFTA, which are one if both sender and target country are members of the
16 Measuring Actual Economic Integration: A Bayesian State-Space Approach 353
same agreement in a certain year. Additional controls are added to ensure that the
effect is not driven by the characteristics of the countries that joined the integration
agreements, or the time period in which they joined. As was the case with the
network density, the standard errors are adjusted for the uncertainty of the AEI
estimate.5
The first column shows that entering into the EU on average caused a 6.5-point
increase in the AEI index, while joining NAFTA causes the index to grow with
46 points. In the second column, the duration of membership is also taken into
account: e.g. EU years is the number of years both countries are members of the
EU. It reveals that the increase happens gradually in the EU, while in NAFTA the
level of integration increases quickly and then slowly degenerates of time. While
significant this effect is nevertheless small: the maximum decrease is 8. It is once
5
In Table 16.2, the following regression is estimated: AEI ¼ βX þ μ with μ ~ N(0, σ 2). This is done
by drawing
a value for β using randomly drawn values of the AEI index from the Gibbs sampler:
βðjÞ X, AEI ðjÞ N½b; σ ðX XÞ1 , with b ¼ (X X)1X AEI( j ), σ ¼ e0 e=ðn kÞ and e ¼ AEI( j ) Xb.
0 0 0
The adjusted standard deviation is then computed using the drawn values of β( j ).
354 G. Rayp and S. Standaert
again most likely driven by the emergence of Asia on the global market. Lastly, the
control variables indicate that regardless of whether they are a member themselves,
the OECD countries are significantly more likely to be highly integrated with
members of the EU and NAFTA. This confirms the central position the latter take
up in the OECD.
A similar pattern emerges when plotting the network structure of the EU
(Fig. 16.6). The edges between member countries become brighter over time.
However, at the same time, we also see the concentration of trade towards the
core countries. The combined effect of both trends causes the network density to
increase until the mid 90s, after which it decreases to its original level (Fig. 16.7).
In this section, the AEI index is compared to the two other techniques used in the
literature: the simple average and a principle component analysis (pca). Firstly,
both are applied when all data is available, but this only provides us with 68 obser-
vations. The computations are then adjusted to cope with this missing observations
problem. The mean is calculated when at least one data point is available and the
weights are adjusted every time. The weights of the pca are computed once using
the pairwise correlation matrix and the index is composed when at least one
observation is available. Keep in mind that the parameters of the state-space
model are kept constant over time themselves.
As was already mentioned, both techniques can be seen as simplified versions of
the AEI index where the values of the parameters are restricted in some way. This
means that we can test the statistical validity of the state-space approach using the
parameter values we find. Figure 16.8 plots out these values including their 95%
confidence interval. However, for clarity’s sake only the first thirty drawn values
are shown; Z, C and H actually have k ¼ 80 elements, while T has nðn 1Þ ¼ 1056
elements. From these graphs, it immediately becomes clear that the assumptions
that T ¼ 0 or that Z, C and H are constant over all indicators are invalid. T lies very
close to one for almost all parameters, and Z, C and H significantly differ even after
all indicators have been standardized to mean zero and a standard deviation of one.
However, the overall correlation (Table 16.3) indicates the mean results of the
three techniques do not differ that much from each other. The only exception is the
adjusted pca analysis, which scores relatively low. This can be explained by the fact
that the weights are kept constant for all country-couples and time-periods, ignoring
the availability of the indicators.
The last two columns of Table 16.3 decompose the overall correlation into the
correlation between the means for each country (between), and that of the
demeaned series (within). It shows that the strong overall correlation is the result
of the high correlation of the mean values. The within-variation on the other hand
differs significantly over the three methods. This implies that while the methodol-
ogy might not matter in a cross-country study, this radically changes when the level
of integration is compared over time. This could lead to substantial differences in
16 Measuring Actual Economic Integration: A Bayesian State-Space Approach 355
Fig. 16.6 Network structure EU-25 (Network structure of the EU-25 member countries, exclud-
ing Cyprus, Latvia, Lithuania and Malta. The color of the links (edges) reflects the height of the
AEI index, the position and color of the countries the weighted indegree. If both countries are a
member of the EU, the link between them changes color)
356 G. Rayp and S. Standaert
fixed effects studies, that use only the within variation, or for example in the
analysis of the effect of institutional integration on actual economic integration
(cf. Dorrucci et al. 2004).
Lastly, the adjusted mean and the state-space model differ significantly in the
confidence with which they make their predictions. Assuming normality or using
the central limit theorem, the variance of the mean is equal to σ 2 =k with σ 2 the
population variance of yi,t. This implies that as the number of available data
increases from 4 to 80, the standard deviation falls to less than one fourth of its
original level. While the AEI index’s reliability also decreases as availability
increases, the difference is far less pronounced. The reason is that the AEI index
uses all available data from the entire time-series and gives less weight to indicators
that are less reliable. This results in confidence bounds that are on average only half as
large. Resizing the AEI index and the mean to lie between zero and one, the average
standard deviation of AEI is only 0.008 versus 0.016 for the adjusted mean. More-
over, in less than 1% of the sample is the standard deviation of the mean lower.
16.4 Extensions
As was mentioned, the model estimated in this paper can be extended in multiple
ways. An obvious extension is to include a larger number of countries. As more
non-OECD countries are added, the quality and availability of data becomes
increasingly problematic. However, as was demonstrated, the model is particularly
well suited to handle these problems. The main concern would be computational
power. Expanding the current dataset with one country would add 34 additional
country-couples times 41 time periods or 1394 observations. Running the model
with 193 countries for the same time period would result in more than one and a half
million observations.
A second extension concerns the type of integration studied and the unit of
analysis. The state-space model can be used to study potential economic integra-
tion, or political integration. With respect to the latter, a powerful advantage of the
state-space model is that it can combine different types of data. For example, the
model defined in Sect. 16.2 only combines continuous variables, but through the use
of latent variables it can easily be extended to combine dichotomous information or
a combination of both:
The value of the (observed) dichotomous indicator y depends on the value of the
(unobserved) continuous latent variable y*i, t which in turn is driven by the to-be-
estimated level of regional integration RI.
358 G. Rayp and S. Standaert
The ability to combine both different types of data means that qualitative data on
integration can be added without having to impose a subjective scaling. As is the
case in Hufbauer and Scott (1994) and Feng and Genna (2003), the different aspects
of integration are viewed in a parallel rather than a sequential way. For example,
currency unions can be viewed separately from customs unions. In this way, the
index would prevent a one-track, EU-dominated view of integration.
Secondly, it also does away with the linear scaling between the different forms
of integration as the information contained in the continuous indicators provide a
natural scaling. For example, if closing a free trade agreement goes hand in hand
with a significant increase in bilateral flows, the weight of this measure will be
higher than when it leaves the bilateral flows unperturbed.
16.5 Conclusion
Acknowledgements We are indebted to Gaspare Genna, Philippe De Lombaerde and the partic-
ipants of UNU-CRIS’s Workshop on Indicator-Based Monitoring of Regional Economic Integra-
tion for their comments and suggestions. Funding provided by the Research Foundation – Flanders
and the Belgian National Bank.
References
De Lombaerde, P., Dorrucci, E., Genna, G., & Mongelli, F. P. (2008). Quantitative monitoring and
comparison of regional integration processes: Steps towards good practice. In A. K€ osler &
M. Zimmek (Eds.), Elements of regional integration (pp. 149–179). Baden-Baden: Nomos.
De Lombaerde, P., Dorrucci, E., Genna, G., & Mongelli, F. P. (2011). Composite indexes and
systems of indicators of regional integration. In L. De, R. G. Flôres, P. L. Iapadre, & M. Schulz
(Eds.), The regional integration manual. Quantitative and qualitative methods (pp. 323–346).
London: Routledge.
Dennis, D. J., and Yusof, Z. A. (2003). Developing indicators of ASEAN integration – A
preliminary survey for a roadmap. REPSF Project 02/001.
Dorrucci, E., Firpo, S., Fratzscher, M., & Mongelli, F. P. (2004). The link between institutional and
economic integration: Insights for Latin America from the European experience. Open Econ-
omies Review, 15, 239–260.
Durbin, J., & Koopman, S. (2012). Time series analysis by state space methods. Oxford: Oxford
University Press.
Feng, Y., & Genna, G. (2003). Regional integration and domestic institutional homogeneity: A
comparative analysis of regional integration in the Americas, Pacific Asia and Western Europe.
Review of International Political Economy, 10(2), 278–309.
Heston, A., Summers, R. & Aten, B. (2012). Penn World Table 7.1. University of Pennsylvania,
Center for International Comparisons of Production, Income and Prices.
Hufbauer, G. C., & Scott, J. J. (1994). Western hemisphere economic integration. Washington,
DC: Institute for International Economics.
Kim, C. J., & Nelson, C. R. (1999). State-space models with regime switching: Classical and
gibbs-sampling approaches with applications. Cambridge, MA: MIT Press.
Koop, G., Poirier, D. J., & Tobias, J. L. (2007). Bayesian econometric methods. Cambridge:
Cambridge University Press.
Lancaster, T. (2004). Introduction to modern Bayesian econometrics. Hoboken: Wiley.
Mongelli, F. P., Dorrucci, E. & Agur, I. (2005). What does European institutional integration tell
us about trade integration (European Central Bank Occasional Paper Series 40).
Newman, M. (2010). Networks: An introduction. Oxford: Oxford University Press.
360 G. Rayp and S. Standaert
Standaert, S. (2014). Divining the level of corruption: A Bayesian state-space approach. Journal of
Comparative Economics. doi:10.1016/j.jce.2014.05.007.
UNECA. (2001). Annual report on integration in Africa. Methodology for calculating indices of
economic integration effort in Africa. Addis Ababa: UN Economic Commission for Africa.
UNECA. (2002). Annual report on integration in Africa. Addis Ababa: UN Economic Commis-
sion for Africa.
UNECA. (2004). Assessing regional integration in Africa. Addis Ababa: UN Economic Commis-
sion for Africa.
UN-ESCWA. (2006). Annual review of developments in globalization and regional integration in
the Arab countries. New York: United Nations.