Professional Documents
Culture Documents
Mthuli Ncube, Issa Faye, Audrey Verdier-Chouchane (Eds.) - Regional Integration and Trade in Africa-Palgrave Macmillan UK (2015)
Mthuli Ncube, Issa Faye, Audrey Verdier-Chouchane (Eds.) - Regional Integration and Trade in Africa-Palgrave Macmillan UK (2015)
Mthuli Ncube
Senior Research Fellow, Blavatnik School of Government, University of Oxford (on leave
from the African Development Bank)
Issa Faye
Manager of the Research Division, African Development Bank
Audrey Verdier-Chouchane
Chief Research Economist, African Development Bank
© African Development Bank 2015
Individual chapters © Respective authors 2015
Foreword © Gilbert Mbesherubusa 2015
Softcover reprint of the hardcover 1st edition 2015 978-1-137-46204-1
All rights reserved. No reproduction, copy or transmission of this
publication may be made without written permission.
No portion of this publication may be reproduced, copied or transmitted
save with written permission or in accordance with the provisions of the
Copyright, Designs and Patents Act 1988, or under the terms of any licence
permitting limited copying issued by the Copyright Licensing Agency,
Saffron House, 6–10 Kirby Street, London EC1N 8TS.
Any person who does any unauthorized act in relation to this publication
may be liable to criminal prosecution and civil claims for damages.
The authors have asserted their rights to be identified as the authors of this work
in accordance with the Copyright, Designs and Patents Act 1988.
First published 2015 by
PALGRAVE MACMILLAN
Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited,
registered in England, company number 785998, of Houndmills, Basingstoke,
Hampshire RG21 6XS.
Palgrave Macmillan in the US is a division of St Martin’s Press LLC,
175 Fifth Avenue, New York, NY 10010.
Palgrave Macmillan is the global academic imprint of the above companies
and has companies and representatives throughout the world.
Palgrave® and Macmillan® are registered trademarks in the United States,
the United Kingdom, Europe and other countries.
ISBN 978-1-349-69031-2 ISBN 978-1-137-46205-3 (eBook)
DOI 10.1057/9781137462053
This book is printed on paper suitable for recycling and made from fully
managed and sustained forest sources. Logging, pulping and manufacturing
processes are expected to conform to the environmental regulations of the
country of origin.
A catalogue record for this book is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Regional integration and trade in Africa / [edited by] Mthuli Ncube, Chief Economist
and Vice President, African Development Bank Group, South Africa, Issa Faye,
Manager of the Research Division, African Development Bank, Audrey
Verdier-Chouchane, Chief Research Economist, African Development Bank.
pages cm
Summary: “Regional integration has been adopted by many African governments
as a response to difficulties facing national markets in landlocked countries.
However, despite the aspiring targets they have experienced setbacks due to poor
implementation. This book reviews the current trends, challenges and potential
benefits of regional integration and trade in Africa, in a context of rising
interest and momentum among African leaders on the topic. This study explores the
many facets of regional trade, its implementation gaps and positive effects
through empirical studies covering various African sub-regions and using different
theoretical approaches. It analyses trade performance against full potential, its
mitigated impact on the competitiveness, industrialization and economic
diversification of African countries as well as the much debated effect of monetary
unions on regional trade in the region. It provides valuable policy recommendations
aimed at stimulating the debate among the government, private sector and
development community on the ways to promote regional trade for Africa’s
economic development. Regional Integration and Trade in Africa examines various
aspects of regional integration and trade in Africa, and will be of interest to
scholars, students, researchers, policy-makers, and professionals” —
Provided by publisher.
1. Africa—Commercial policy. 2. Africa—Commerce. 3. Africa—Economic
integration. 4. Africa—Foreign economic relations. I. Ncube, Mthuli, 1963–
II. Faye, Issa, 1972– III. Verdier-Chouchane, Audrey, 1973–
HF1611.R437 2014
337.1 6—dc23 2014028329
Contents
List of Tables ix
Acknowledgments xiv
Notes on Contributors xv
v
vi Contents
Index 229
Figures
vii
Maps
viii
Tables
ix
x List of Tables
xii
Foreword xiii
integration and trade within broader policy and practical contexts, look-
ing carefully at some of the solutions available. The chapters provide
comparative discussions of the challenges and look at the many oppor-
tunities that Africa’s abundant resources and large population provide
for trade expansion and welfare enhancement. A good measure of both
theoretical and empirical tools is used in the presentations in each
chapter.
I am confident that the book will have important impacts on debates
about economic integration and trade in Africa by offering pointers to
the solutions needed to bring countries, and the continent as a whole,
up to speed in meeting the aspirations of its leaders and people. I would,
therefore, like to recommend the book to policymakers, researchers and
practitioners in economic development and other readers interested in
development issues of the day as an additional tool in their endeavors
to ensure that Africa is not left behind in this era of rapid globalization.
Gilbert Mbesherubusa
Former Vice President, Infrastructure Private Sector and Regional
Integration African Development Bank
Acknowledgments
xiv
Contributors
xv
xvi Notes on Contributors
Mthuli Ncube is on leave from the post of Chief Economist and Vice
President of the African Development Bank. He is now Senior Research
Notes on Contributors xvii
xxi
xxii List of Acronyms
Introduction
Africa has registered growth rates of more than 5% annually over the
past decade (2004–2014), with a third of African countries averaging
6.5% growth per year (African Development Bank et al., 2014). This
current growth pattern has brought Africa to a turning point, if not a tip-
ping point, in its development trajectory. A more globalized world econ-
omy has brought increasing functional and spatial specialization, as well
as growing interdependence, and vulnerability of national economies
to external shocks. Meanwhile, a process of de-industrialization in the
developed world has been paralleled by an increasing competitiveness
of developing countries in manufacturing. The changing structure of
the global economy requires Africa to integrate in order to survive.
Translating economic gains into sustainable and shared growth requires
Africa to connect its markets, deepen Regional Integration (RI) and
enhance national competitiveness (World Economic Forum et al., 2013).
Embedding RI into Africa’s development agenda, whether in the form of
preferential trading area or economic and monetary union, will move
African economies to the next stage of competitiveness and integration
in the global economy as befits the continent’s rise.
RI is a means to achieving greater global integration, especially for
countries which cannot compete on a global scale by themselves such as
small and landlocked economies (DfID, 2011). As Sub-Saharan Africa is
fragmented into 48 small states, a large number of which are landlocked
countries, RI should, in theory, be an important tool for industrializ-
ing Africa through economic diversification, export competitiveness and
facilitated access to foreign markets. According to the theory of com-
parative advantages (Ricardo, 1817), greater openness is associated on
1
2 Introduction
CEN-SAD
COMESA
UMA Seychelles
Algeria Comoros Congo, Dem. Rep*.
Egypt Ethiopia Madagascar
Malawi
IGAD Mauritius
Libya Zambia
Mauritania Djibouti Zimbabwe
Morocco Eritrea
Tunisia Uganda
South Sudan Swaziland
Sudan
Kenya Burundi* SACU
Rwanda Botswana
Lesotho
Somalia Namibia
WAEMU South Africa
EAC Tanzania
Benin
Burkina Faso
Cote d’lvoire Gambia, The
Guinea-Bissau Ghana Mozambique
Mali Guinea Sao Tome Burundi*
Niger Liberia and Principe SADC
Senegal Nigeria
Togo Sierra Leone Angola
Congo,
Central African Rep. Cameroon Dem. Rep*.
Chad Congo (Brazzaville)
Equatorial Guinea
Gabon
CEMAC
Cape verde
ECOWAS ECCAS
2.4 5.6
3.1
5.3
8.9
61.1
13.6
140
120
100
$ billions
80
60
40
20
0
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
ECCAS
AMU
IGAD
COMESA
CEN-SAD
ECOWAS
EAC
SADC
0 5 10 15 20 25
With Africa With same REC
Figure I.4 Intra-African trade by selected RECs over 2007–2011 (% total trade)
Source: Authors based on UNCTADstat database.
References
African Development Bank [AfDB] (2014), OECD Development Centre and
United Nations Development Programme, African Economic Outlook, Paris:
OECD Publishing.
Alesina, A. and Barro, R. J. (2002), ‘Currency Unions’, Quarterly Journal of
Economics, vol. 117, no. 2, pp. 409–436.
Alesina, A., Barro, R. J. and Tenreyro, S. (2002), ‘Optimum Currency Areas’,
National Bureau of Economic Research, Working Papers, no. 9072, NBER, Inc.
Barro, R. J. and Lee, J.-W. (2002), ‘IMF Programs: Who Is Chosen and What Are
the Effects?’ Working Paper,r no. 8951, National Bureau of Economic Research.
Bhagwati, J. (1988), Protectionism, Cambridge, MA: MIT Press.
Burton, H. (1998), ‘Reconsideration of Import Substitution’, Journal of Economic
Literature, vol. 36, pp. 306–936.
Department for International Development [DfID] (2011), ‘Regional Integration
and Trade in Sub-Saharan Africa’, Trade and Investment Analytical Papers, DfID:
Department for Business Innovation and Skills.
Dornbusch, R. (1992), ‘The Case for Trade Liberalization in Developing Coun-
tries’, Journal of Economic Perspectives, vol. 6, pp. 69–85.
Geda, A. (2002), ‘Debt Issues in Africa: Thinking beyond the HIPC Initiative
to Solving Structural Problems’, Working Paper Series UNU-WIDER Research
Paper, World Institute for Development Economic Research.
Glick, R. and Rose, A. K. (2002), ‘Does a Currency Union Affect Trade?’ European
Economic Review, vol. 46, pp. 1125–1151.
Kaluwa, B. M. and Reid, G. C. (1991), ‘Profitability and Price Flexibility in Manu-
facturing for a Developing Country’, The Journal of Industrial Economics, vol. 39,
no. 6, pp. 689–700.
14 Introduction
Introduction
17
18 Competitiveness and Integration through Trade in CEMAC
1 Definition of concepts
Trade integration
In economics, integration refers to the strategy of bundling activities
within a company. This enables a company to master the technical,
commercial or financial expertise to boost productivity and benefit
from synergy effects. ‘Horizontal’ and ‘vertical’ integration can also be
differentiated.
Vertical integration is when such bundling concerns the different
stages of production and distribution for the same type of goods or ser-
vices. Horizontal integration (or horizontal concentration) is when a
company develops its business at the same level of the value chain as its
products. The goal is to spread costs over a larger quantity of products,
20 Competitiveness and Integration through Trade in CEMAC
The data used are derived from the trade statistics of the Bank of Central
African States (Banque des Etats de l’Afrique Centrale – BEAC) and Cen-
tral Bank of West African States (Banque Centrale des Etats de l’Afrique
de l’Ouest – BCEAO) member countries, and downloadable from the
websites of these two institutions. The years 2002 and 2008 are cho-
sen for the harmonization of indicators and consideration of the effects
of the global economic and financial crises (2008), respectively. In the
absence of a comprehensive database on sectors, the available data show
exports and imports of goods and non-factor services and petroleum
products.
This study relies on calculation of comparative advantages in terms
of trade in CEMAC countries, which are considered as determining for
international trade. This is based on the difference between comparative
costs in a number of countries, as each country finds it advantageous to
specialize in and export goods for which it has the greatest comparative
advantage and import other goods from its partners (Lassudrie-Duchêne
and Ünal-Kezenci, 2001). However, it is difficult to measure the com-
parative advantage directly, so a large number of empirical studies
use the RCA indicator to measure the international specialization of
countries.
The RCA concept, introduced by Balassa in 1965, holds that inter-
national trade in goods reflects cost differentials between countries,
and therefore shows their comparative advantages. So observations of
business performance can enable measurement of revealed comparative
advantage. The better a country’s relative performance in trading a spe-
cific good, the greater its comparative advantage in producing that good
(Balassa, 1965).
The following formula is used to calculate RCA, which serves as a basis
for comparing the export structure of each country to that of a reference
area (CEMAC and WAEMU).
Xk(i)
Tx(i)
RCAk (i, w) = (1)
Xk(w)
Tx(w)
Joseph Parfait Owoundi 23
( − Mr)
(Xr
Zr = 100 (2)
((X + M)/2
( − M)
(X
Z = 100 (3)
((X + M)/2
4 Results
Source: Culled from global competitiveness index rankings of the 2013 World Economic
Forum report.
Competition structure
To describe the structure of competition of CEMAC countries, com-
parative advantages and contributions to trade balance are calculated.
Table 1.2 shows the RCAs of the countries concerned between 2005
and 2008.
This table shows CEMAC countries are competitive in their own mar-
ket when it comes to exports of petroleum products and non-factor
goods and services (RCA = 1). The countries are specialized in these two
sectors, except Chad, which is not specialized in the export of petroleum
products in the CEMAC zone; its comparative disadvantage ranges from
0.07 to 0.03 between 2005 and 2008.
Their competitiveness diminishes slightly from one year to another,
due to a sluggish global economic environment following the economic
and financial crises of 2008 and 2009. Cameroon and Chad lead their
26 Competitiveness and Integration through Trade in CEMAC
CEMAC Market
Equatorial 1.374 0.959 1.869 1.121 1.437 0.996 0.1439 0.088
Guinea
Cameroon 1.705 1.0925 0.901 0.722 1.475 0.952 6.193 4.372
Gabon 1.434 0.969 1.693 1.090 1.465 0.9912 0.821 0.470
Congo 1.401 0.927 1.791 1.218 1.408 0.9257 1.259 0.961
Chad 1.989 1.328 0.0713 0.0310 2.0172 1.323 1.455 1.317
WAEMU Market
Equatorial 157.84 168.85 1.01 1.094
Guinea
Cameroon 195.92 192.24 0.48 0.705
Gabon 164.75 170.65 0.91 1.064
Congo 160.92 163.15 0.967 1.189
Chad 228.58 232.75 0.038 0.030
CEMAC Goods and non-factor Petroleum products Goods exportation and Non-factor services
countries services exportation and exportation and importation exportation and
importation importation importation
CEMAC Market
Equatorial 16.04 22.02 −14.931 −8.243 −32.087 −44.059 30.97 30.27
Guinea
Cameroon 8325.62 4487.70 −1148.51 59.006 −16651.24 −8975.42 9474.13 4428.70
Gabon 36.383 24.529 −32.137 −23.270 −72.76 −49.058 68.52 47.799
Congo 56.66 141.92 −46.73 −146.59 −113.33 −283.85 103.40 288.52
Chad −22.64 −203.50 −7292.36 −3738.78 45.28 407.00 7269.72 3535.27
5 Conclusion
6 Recommendations
Notes
1. In Economics, comparative advantage is the main concept of the traditional
theory of international trade. It was broached by Robert Torrens in 1815 and
demonstrated for the first time by the British economist David Ricardo in
1817 in his On the Principles of Political Economy and Taxation. According to
Paul Samuelson, the 1970 Nobel Prize winner in Economics, it is the best
example of an undeniable but counter-intuitive economic principle.
2. Trade imbalances, whether due to economic activity (e.g. France in the years
1960–1970 recorded a trade deficit through capital goods imports following
growth) or exchange rate variations, can flaw the measurement of compara-
tive advantages. The French Centre for Prospective Studies and International
Information (CEPII) in the 1980s developed a contribution to trade balance
indicator that can correct that distortion (Cf. Gérard Lafay, 1984: Avantages
Comparatifs et Compétitivité. Commerce International, p. 52; Cf. Jean-Luc
tavernier, 1990, Echanges extérieur et avantages comparatifs: la spécialisation
de la France confrontée à celle de ses concurrents, Economie et prévision, N)
94/95.
References
Balassa, B. (1965), Trade Liberalization and Revealed Comparative Advantage, The
Manchester School of Economic and Social Studies, no. 33, May.
Desmas (2005), Analyse comparative de compétitivité: le cas de la filière
tomate dans le contexte euro-méditerranéen. Mémoire de Fin d’Etudes
Diplôme d’Agronomie Approfondie (D.A.A.); option: Politique Economique de
l’Agriculture et de l’Espace.
Djaowe, J. (2009), Investissements Directs Etrangers (IDE) et Gouvernance: les
pays de la CEMAC sont-ils attractifs? Revue africaine de l’Intégration, vol. 3, no. 1,
pp. 67–94.
Gallezot, J. (2006), les enjeux et les marges de manœuvre de la CEDEAO
face aux defis des négociations agricoles; Réseau des organisations paysannes
et de producteurs agricoles de l’Afrique de l’Ouest (ROPPA), Afrique
Nourricière.
Godonou Dossou, J. (2008), forces et enjeux de l’intégration sous régionale:
CEMAC/CEEAC; Intl. Rel., Fmr. Dean/Fac.
Heckscher, E. (1949), ‘The Effect of Foreign Trade on the Distribution of Income’
in Ellis, S., H., and L., M., Metzler (s. dir), (1919), Readings in the Theory of
International Trade, Philadelphia: Blakiston.
Lafay, G. (1984), Avantages comparatifs et Compétitivité. Commerce Interna-
tional, p. 52.
30 Competitiveness and Integration through Trade in CEMAC
Lafay G., 1990, La mesure des avantages comparatifs révélés, Economie prospective
internationale, no. 41.
Lassudrie-Duchêne, B. and Ünal-Kezenci, D. (2001), L’avantage comparatif,
notion fondamentale et controversée, in L’économie Mondiale, La Découverte,
pp. 90–104.
Mignamissi, M. D. (2008), Réformes macroéconomique et intégration par le
marché dans la CEMAC; Université Yaoundé II – DEA 2008.
Moussone, E. (2010), Insertion des pays de la zone franc africaine dans le com-
merce mondial étude d’une spécialisation appauvrissante et le problème du
financement de l’économie; Documents de travail – no. 231 novembre 2010.
Ohlin, B. (1933), International and Interregional Trade, Cambridge, MA: Harvard
University Press.
Ricardo, D. (1817), ‘On the Principles of Political Economy and Taxation’, Posner
Memorial Collection, http://posner.library.cmu.edu/Posner/books/pages.cgi?
call=330_R4.
Samuelson, P. A. (1949), ‘International Factor-Price Equalization Once Again’,
Economic Journal, vol. 59, pp. 181–197.
Smith, A. (1925[1776]), An Inquiry into the Nature and Causes of the Wealth of
Nations. London: Methuen (Cannan ed.).
Tavernier, Jean-Luc (1990), Echanges extérieurs et avantages comparatifs: la spé-
cialisation de la France confrontée à celle de ses concurrents, Economie et
prévision, no. 94/95.
UNECA (2006), Economic Report on Africa 2006. Capital Flows and Development
Financing in Africa, Addis Abeba: Economic Commission for Africa.
World Economic Forum (2013), The Global Competitiveness Report 2013–2014,
Geneva: World Economic Forum.
2
Economic Integration, Trade
Facilitation and Agricultural
Exports Performance in
ECOWAS Sub-Region
Wumi Olayiwola, Evans Osabuohien, Henry Okodua and
Oluyomi Ola-David
31
32 Economic Integration and Agriculture in ECOWAS
120000
100000
80000
60000
40000
20000
0
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Exports Imports
Figure 2.1 ECOWAS’ global exports and imports, 1999–2009 (USD millions)
Source: Computed from ECOWAS Commission, 2010b and World Bank (2010).
25
20
15
10
0
1999 2001 2003 2005 2007 2009
Intra import Intra export
Figure 2.2 ECOWAS’ intra-regional exports and imports as percent of its global
trade
Source: Computed from ECOWAS Commission, 2010b and World Bank (2010).
TF = daysi = f (Institutions
( , Infrastructure, Macro) (1)
+ π6 TF ij,t + ω (8)
Estimation techniques
Endogeneity – often a problem in dynamic panel models – can be
dealt with using instrumental variable (IV) estimation, but the instru-
ments must be genuinely exogenous yet strongly correlated with the
potentially endogenous variables. It helps to over-identify the model
to allow tests for exogeneity and excludability. Besides Two-Stage Least
Squares (TSLS), the other commonly used IV estimation method – Gen-
eralized Method of Moments (GMM) – produces identical results in
just-sufficiently identified models, but more precise results with over-
identified models. GMM also uses internal instruments, whereas with
TSLS the researcher must search for suitable external instruments.
Panel data from the 15 ECOWAS members during 2003–2008 was
therefore analyzed using GMM with the dynamic equations (4, 5 and
8). The choice of this period – besides the issue of data availability for
most members of ECOWAS – includes the need to have the number
of country dimension to be relatively larger than the time dimension.
Another reason is to focus on the period that is relatively close to the
year ECOWAP was adopted. In addition, it was from 2003 that the data
for WGI became annual. For robustness – and to account for the pres-
ence of both time-variant and – invariant explanatory factors – the same
data was also analyzed using fixed-effects estimation with the static
equations (2, 3 and 7). A preliminary analysis using Hausman (1978) test
is also carried out. Hausman test is used to differentiate between fixed
effects model and random effects model in panel data analysis. In this
case, the analysis indicated that fixed effects model gave more efficient
38 Economic Integration and Agriculture in ECOWAS
results than random effects model due to higher efficiency, while the
alternative Fixed effects (FE) is at least consistent.
An advantage of dynamic GMM estimation is that all variables not
correlated with the error term (including lagged and differenced vari-
ables) may be used as instruments (Greene, 2008). The optimal set
of internal instruments was chosen via the ‘collapse’ option in the
GMM command in STATA software. The lagged dependent variables
(xdays and mdays) appear as predetermined and endogenous variables.
Hence, endogeneity is controlled for by using internal instruments. The
dynamic GMM estimator applied in this study uses the levels equation
in each case to obtain a system of two equations: one differenced and
one in levels. The variables in the differenced equation are instrumented
with the lagged levels of the regressors in the first equation, while the
variables in levels in the second equation are instrumented with their
own first differences.
The estimated results from the fixed effect (static) models are broadly
as expected. A 1% increase in export share (as a measure of regional
integration) correlated with about 0.23% reduction in the number of
days required to process exports in the 15 ECOWAS countries (Table 2.1,
column 1). Similarly, a 1% increase in GDP per capita correlated with
about 1.05% reduction. A 1% increase in the density of Internet users
correlated with about 0.18% reduction and a 1% increase in the density
of telephone subscribers with about 0.06% reduction. All these variables
might thus be important potential channels for increasing trade facilita-
tion in ECOWAS. The constant term suggests that, in the absence of all
these variables, it would take about 98 days to process exports.
A 1% increase in GDP per capita correlated with about 1.21% reduc-
tion in the number of days required to process imports (column 2),
while a 1% increase in the density of Internet users correlated with about
0.19% reduction. The constant term suggests – again, in the absence of
all explanatory variables – that it would take about 106 days to process
imports.
Neither the rule-of-law index nor the control-of-corruption index
were statistically significant in determining either export or import days,
though both had large apparent negative effects (that is, reducing export
and import days, as desired).
As would be expected, agricultural production had a huge effect on
agricultural exports (column 3), a 1% increase in production correlated
39
∗ Measures of
trade
facilitation
xdays (lagged) 1.329∗
(6.51)
mdays (lagged) 1.233∗
(9.47)
ag – exports 0.655∗
(lagged) (3.41)
ag – product 7.220∗∗∗ 8.650∗∗∗
(1.69) (1.65)
∗ Indicatorsof
institutions
Rule of law −5.211 −8.275 −0.662 −1.274
index (−0.72) (−0.86) (−0.15) (−0.28)
Control of −3.338 −2.388 −0.447 −0.572
corruption (−1.09) (−0.56) (−0.09) (−0.11)
index
Regulatory −0.674 1.219
quality index (−0.43) (0.89)
∗ Measures of
infrastruc-
ture
Internet users −0.181∗∗ −0.193∗∗∗ −0.099∗∗ 0.094 0.052 −0.121∗∗
(−2.27) (−1.73) (−2.44) (0.89) (0.67) (−2.02)
Telephone users −0.064∗∗ −0.031 0.042∗∗ 0.049 0.075 0.046∗∗
(−2.26) (−0.79) (2.48) (0.88) (1.34) (2.30)
∗ Measureof
economic
conditions
GDP per capita −1.054∗ −1.216∗ 0.028 0.027
(−4.52) (−3.72) (0.45) (0.47)
∗ Measures of
regional
integration
Export share −0.233∗ 0.025 −0.119∗∗∗ 0.118∗∗
(−3.15) (0.79) (−1.91) (2.28)
Import share 0.031 −0.025
(0.23) (−0.24)
TF (xdays) −0.071∗ −0.086∗
(−3.02) (−2.95)
40 Economic Integration and Agriculture in ECOWAS
Notes: t-statistics are in parentheses; ∗ , ∗∗ , ∗∗∗ indicate significant at 1%, 5% and 10% levels;
all dynamic models had time dummies.
5 Policy recommendations
Annexes
Note
1. The purpose of using this measure is to take into consideration the economic
sizes of ECOWAS members.
References
AfDB (2012), African Economic Outlook 2012: Promoting Youth Employ-
ment, OECD Publishing. doi: 10.1787/aeo-2012-en (Accessed 20 November
2013).
Balchin, N. and Edwards S. (2008), ‘Trade-related Business Climate and Manufac-
turing Export Performance in Africa: A Firm Level Analysis’, Paper presented
at the Trade Costs and Business Environment: A Focus on Africa Workshop,
31 May, Entebbe, Uganda.
Central Bank of Nigeria (2006), International Economic Cooperation, http://www.
cenbank.org/monetarypolicy/ecowas.asp (Accessed 5 May 2011).
Cissokho, L., Haughton, J., Makpayo, K. and Seck, A. (2013), ‘Why Is Agricultural
Trade within ECOWAS so High?’, Journal of African Economies, vol. 22, no.1,
pp. 22–51.
Clarke, G. (2005), ‘Beyond Tariff and Quotas: Why Don’t African Manufacturing
Enterprises Export More?’ World Bank Policy Research Working Paper, r no. 3617.
Dalimov, R. T. (2009), ‘The Dynamics of Trade Creation and Trade Diversion
Effects Under International Economic Integration’, Journal of Economic Theory,
vol. 1, no. 1, pp. 1–4.
Djankov, S., Freund, C. and Pham, C. S. (2010), ‘Trading on Time’, The Review of
Economics and Statistics, vol. 92, no. 1, pp. 166–172.
ECOWAS Commission (2010a), Regional Agricultural Policy, Abuja: ECOWAS.
ECOWAS Commission (2010b), The ECOWAS External Trade Statistics, Abuja:
ECOWAS
Efobi, U. and Osabuohien, E. S. C. (2011), ‘Promotion of Non-Oil Export in
Nigeria: Empirical Assessment of Agricultural Credit Guarantee Scheme Fund’,
Journal of Economic Theory, vol. 3, no. 1, pp. 20–28.
Elbadawi, I., Mengistae, T. and Zeufack, A. (2006), ‘Market Access, Supplier Access
and Africa’s Manufactured Exports: An Analysis of the Role of Geography and
Institutions’, WB Policy Working Paper, r no. 3942.
Engman, M. (2005), ‘The Economic Impact of Trade Facilitation’, OECD Trade
Policy Working Papers, no. 21.
Essien, A. E. (2009), ‘Introducing a Common Currency: The ECOWAS Experi-
ence’, Paper Presented at the ECOWAS-GIZ Regional Integration Academy in
Comparative Regional Integration, Centre for European Integration Studies,
Bonn, Germany, 16–28 March.
Freund, C. and Bolaky, B. (2004), ‘Trade, Regulations, and Growth’, World Bank
Policy Research Working Paper,r no. 3255.
Hausman, J. A. (1978). ‘Specification Tests in Econometrics’, Econometrica, vol. 46,
no. 6, pp. 1251–1271.
International Centre for Trade and Sustainable Development [ICTSD] (2011),
‘Agricultural Export Restrictions Speak Controversy at the WTO’, Bridges Weekly
Trade News Digest,t vol. 15, no. 2, pp. 1–15.
Wumi Olayiwola et al. 45
Yang, Y. and Gupta, S. (2007), ‘Regional Trade Arrangements in Africa: Past Per-
formance and the Way Forward’, African Development Review, vol. 19, no. 3,
pp. 399–431.
Yoshino, Y. (2008), ‘Domestic Constraints, Firm Characteristics, and Geograph-
ical Diversification of Firm-level Manufacturing Exports in Africa’, WB Policy
Research Working Paper,
r no. 4575.
3
The Impact of Trade Liberalization
on Export Growth and Import
Growth in Sub-Saharan Africa
Lanre Kassim
Introduction
47
48 Trade Liberalization in Sub-Saharan Africa
Table 3.1 Average export and import growth before and after liberalization
Note: (∗ ) denotes a marginal increase or decrease while all values are the author’s calculations.
Source: WTO, 1995–2011. Policy reviews for various countries.
Model specification
A conventional export demand equation relates the level of exports to
world real income and a measure of price competitiveness. Assuming
constant price and income elasticities of demand, the export equation
can be expressed as:
δ
EPex
EXPt = A ∗
Wtγ (1)
Pim t
growth of world real income, and εit is the idiosyncratic error term. Also,
β1 and β2 denote the price and income elasticity of demand for exports
respectively. Equation (3) is augmented to include the two measures of
liberalization as follows:
Here libdum is the liberalization dummy which takes the value of 1 from
the year significant trade reforms began in an SSA country and zero
beforehand. Since trade liberalization reduces the degree of anti-export
bias, the variable libdum is expected to have a positive impact on real
export growth. epd is measured as the rate of change of export duties.
But what is the impact of trade liberalization on the price and income
elasticity of demand for exports? The implementation of trade reforms
should raise the income elasticity of demand for exports, implying
structural change in the form of resources transferred to sectors with
high income elasticities. Hence, two interaction variables are created
to examine whether trade liberalization has significantly increased or
decreased the responsiveness of export growth to world income and
relative price changes:
Random effects
Table 3.2 presents the random effects (RE) results, which the Hausman
test indicates as the appropriate estimator for Equations (4) and (5).
Liberalization significantly increases export growth by 3.32 per-
centage points, while a 10% decrease in export duties increases
export growth by 0.2%, albeit insignificant. The income elasticity of
demand for exports is 1.12 which means that a change in world
income will cause a marginally higher change in the demand for
SSA exports. In addition, the price elasticity of demand is −0. 24, imply-
ing that exports are much less responsive to changes in relative prices.
So SSA countries are still major exporters of primary commodities.
Lanre Kassim 53
RE (I) RE (II)
Note: ∗ , ∗∗ and ∗∗∗ indicate that a coefficient is statistically significant at 1%, 5% and
10% significance level, respectively. Figures in parenthesis () are absolute t/z ratios
while figures in brackets [ ] are p-values. ‘Joint Sig’ is an F-test for the joint significance
of the two slope dummies.
The second RE regression includes the two interaction terms rerlib and
wlib. The price elasticity of demand for exports drops to −0.13, and there
is also evidence that trade liberalization increased the price elasticity of
demand for exports by 0.26 percentage points. There is no significant
evidence of the impact of trade liberalization on income elasticity of
demand for exports. Testing the joint significance of the two interaction
terms and the results shows they are jointly significant.
models. Thus, equation (4) and (5) are specified in dynamic form as
follows:
Here epggit −1 is the lagged export growth variable, αi is the fixed effect,
and εit is the idiosyncratic error term. The regressors are those defined
in Section 3.1. Long-run price and income elasticities of demand for
exports can be calculated as β2 /(1 − β1 ) and β3 /(1 − β1 ), respectively;
while long-run liberalization effect can be estimated as β4 /(1 − β1 ). The
GMM results are shown in Table 3.3.
The lagged export growth variable has a low coefficient, which implies
a small difference between short-run and long-run estimates. Trade lib-
eralization raised the growth of exports by 2.67 percentage points in the
short run with this figure increasing to 3.22 percentage points in the
long run. The short-run income elasticity is 1.43 while in the long run,
the income elasticity is 1.72. The short-run price elasticity of demand is
−0. 28, while the long-run figure stands at −0. 34. This figure is above the
estimated range of long-run price elasticity of demand for SSA exports by
Olofin and Babatunde (2009). The differences in estimates can plausibly
be attributed to the different measures of real exchange rate adopted.
Again, the export duty variable remains insignificant.
The second GMM regression produces similar results to the RE regres-
sion. Liberalization increases the price elasticity of demand for exports
by 0.25 percentage points while there is no significant evidence of the
effect of trade liberalization on income elasticity of demand. The corre-
lation tests show that the error term is serially uncorrelated while the
Sargan test shows that instruments are valid.
Timing impact
The timing impact of trade reforms on export growth in SSA is investi-
gated by replacing the average post-reform dummy variable (libdum) in
the base specification with libdum2 which equals one for the year of lib-
eralization only and libdum3 which captures the year after liberalization
alone. A significant coefficient on either of these variables will imply
that the liberalization had an instant short-run effect on export growth
(see Greenaway et al. 2002). The RE and GMM estimators are applied
and results can be found in Table 3.4.
Lanre Kassim 55
Note: ∗ , ∗∗ and ∗∗∗ indicate that a coefficient is statistically significant at 1%, 5% and
10% significance level, respectively. Figures in parenthesis () are absolute t/z ratios
while figures in brackets [] are p-values.
Note: ∗ , ∗∗ and ∗∗∗ indicate that a coefficient is statistically significant at 1%, 5% and 10%
significance level, respectively. Figures in parenthesis () are absolute t/z ratios while figures
in brackets [] are p-values.
Model specification
To analyze the impact of trade liberalization on import growth, the
same countries, time period and methodology are used as in the export
growth estimations. The reason for this approach is to ensure a con-
sistent framework for comparing the impact of trade liberalization on
export growth and import growth. The static and dynamic equations to
be estimated are:
Here impggit is the real import growth; rerrit represents the rate of change
of the real exchange rate; gdpggit is the domestic real income growth;
libdumit denotes the liberalization dummy; impdit is the rate of change
of import duties; αi is the country-specific effect while εit is the idiosyn-
cratic error term. rerlib is an interaction variable between the rate of
change of the real exchange rate and liberalization dummy while ylib
is an interaction term between domestic income growth and the liberal-
ization dummy. From equation (8), the expected signs of the coefficients
are: δ1 ( + ); δ2 ( + ); δ3 ( + ); δ4 ( − ); δ5 ( + ); δ6 ( + ).
FE (I) FE (II)
Note: ∗ , ∗∗ and ∗∗∗ indicate that a coefficient is statistically significant at 1%, 5% and 10%
significance level, respectively. Figures in parenthesis ( ) are absolute t/z ratios while figures
in brackets [ ] are p-values.
insignificant. The joint significance test shows that the two interaction
variables are jointly insignificant.
The results in Table 3.6 indicate that liberalization raised import
growth by 4.33 and 4.92 percentage points in the short run and long
run respectively. This is above the long-run post-liberalization export
growth by 1.70 percentage points. A 10% decrease in import duties
increase the growth of imports by 5.6%. The short-run income elastic-
ity of demand for import is 1.05 while the long-run figure is 1.19. The
short-run price elasticity of demand is 0.30 while in the long run the
coefficient increases to 0.34. Again, we find a positive impact of trade
liberalization on price elasticity of demand while there is no significant
evidence of the effect of liberalization on income elasticity of demand
for import. Furthermore, the diagnostic tests show the error term is
serially uncorrelated and that the instruments used for the estimated
model are valid.3
Lanre Kassim 59
Note: ∗ , ∗∗ and ∗∗∗ indicate that a coefficient is statistically significant at 1%, 5% and
10% significance level, respectively. Figures in parenthesis ( ) are absolute t/z ratios
while figures in brackets [ ] are p-values.
Timing impact
Again, the timing impact of trade reforms on the growth of imports is
tested by introducing two liberalization dummies (libdum2 and libdum3)
to the base specification. Results are shown in Table 3.7.
A decrease is observed in import growth in the liberalization year,
which can be attributed to an expectation of future lower prices follow-
ing import liberalization. In the following year, however, a positive sig-
nificant import growth is witnessed. Inclusion of the average post-reform
dummy variable in second GMM result does not change the significance
of the ‘year after’ dummy variable. This implies an immediate short-run
60 Trade Liberalization in Sub-Saharan Africa
Note: ∗ , ∗∗ and ∗∗∗ indicate that a coefficient is statistically significant at 1%, 5% and 10%
significance level, respectively. Figures in parenthesis () are absolute t/z ratios while figures
in brackets [ ] are p-values.
Annexes
Real export growth epg git : Exports of goods and services; annual percentage
growth rate (constant 2000 US dollars). Source: World Development Indicators,
2011
Real import growth impg git : Imports of goods and services; annual percentage
growth rate (constant 2000 US dollars). Source: World Development Indicators,
2011
Liberalization dummy (libdumit ): This variable takes the value of one when unin-
terrupted trade reforms began in a SSA country till the end of the sample period
and zero beforehand. Source: WTO Trade Policy Reviews, 2011
Export duties epd dit : These are all levies on goods being transported out of a
country or services being delivered to non-residents by residents. Source: IMF,
2011 and World Development Bank, 2011
Import duties impd dit : These are all levies collected on goods that are enter-
ing a country or services delivered by non-residents to residents. Source: IMF
Government Financial Statistics, 2011 and World Development Indicators, 2011
Domestic income growth gdpg git : This is the annual growth rate of domestic GDP
(constant 2000 US dollars). Source: World Development Indicators, 2011
World income growth wgdpg git : This is the annual percentage growth rate of
World GDP (constant 2000 US dollars). Source: World Development Indicators,
2011
Real exchange rate (rerrit ): This is the nominal exchange rate multiplied by the ratio
of domestic export prices to foreign import prices. Source: World Development
Indicators, 2011
Aid (aid
dit ): This is defined as the annual net official development assistance and
official aid received by a SSA country (constant 2010 US dollars). Source: World
Development Bank, 2011
l.impg – 0.12
(1.81)∗∗∗
libdum 5.27 4.57
(3.50)∗ (3.35)∗
gdpg 1.17 1.01
(7.66)∗ (3.95)∗
rer 0.31 0.30
(6.00)∗ (4.33)∗
impd −0.68 −0.57
(3.03)∗ (2.12)∗∗
aid 0.63 0.78
(0.25) (0.45)
Diagnostic tests
Wald test/F-stat [p-value] [0.0000] [0.0000]
1st-Order serial correlation – 0.005
2nd-Order serial correlation – 0.217
Sargan test – 0.420
Notes
1. See Annex 1 for the list and classification of countries.
2. See Annex 2 for the definition and sources of variables used.
3. As a check for robustness, a variable is added for foreign aid in the import
growth regression. However, a statistically significant coefficient is not found.
This can be explained by the fact that donors pay directly to international
companies for imported goods rather than paying to the government of
SSA countries (see Annex 3 for results).
References
Agosin, M. R. (1991), ‘Trade Policy Reform and Economic Performance: A Review
of the Issues and Some Preliminary Evidence’, UNCTAD Discussion Papers,
no. 41.
Ahmed, N. U. (2000), ‘Export Responses to Trade Liberalisation in Bangladesh:
A Co-Integration Analysis’, Applied Economics, vol. 32, no. 8, pp. 1077–1084.
Babatunde, A. M. (2009), ‘Can Trade Liberalisation Stimulate Export Performance
in Sub-Saharan Africa?’ Journal of International and Global Studies, vol. 2, no. 1,
pp. 68–92.
Lanre Kassim 67
Introduction
71
72 Market Integration in ECCAS
1 Intra-ECCAS trade
Trade levels
Foreign trade trends in the ECCAS member countries between 1995
and 2012 suggest weak intra-community trade, compared to other RECs
worldwide. Figure 4.1 clearly shows that ECCAS foreign trade is minimal
and by far the weakest for all the sub-regions. Intra-ECCAS trade in 1995
represented 1.2% or USD 134.98 million of total trade, peaking at 1.7%
in 1998. During the period under review, intra-ECCAS trade remained
below 1% and represented 0.87% of total trade in 2012, that is about
USD 1,111.340 million in absolute value. The increase of about USD
880 million remains relatively insignificant, compared to the total vol-
ume of trade. By way of comparison, trade between ECOWAS member
countries rose during the same period from 10.34% to 17.15% of total
trade, or USD 2,294.44 million to USD 11,667.27 million – representing
Désiré Avom and Mouhamed Mbouandi Njikam 75
80
70
60
50
40
30
20
10
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
ECCAS ECOWAS SADC
UE-27 ALENA Mercosur
35000 32931.97
30000
25000
20000
15000 12988.33
10000
5523.57
5000
1638.43
515.28
0
Intra-ECCAS Intra-ECOWAS Intra-SADC
Intra-Mercosur Intra-WAEMU
Product structure
The distinctive feature of central African economies is that they rely
heavily on commodities such as agricultural produce, foodstuff and bev-
erages, oil and gas products and base metals According to Hammouda
(2005), composition by product shows ‘the continent is trapped in
the cash crop inclusion snare’, where agricultural and mining prod-
ucts account for nearly 70% of total exports (Table 4.1), although global
demand structure is developing in the opposite direction, to the detri-
ment of African exports. At the same time, imports mainly comprise
manufactured goods.
The main trading partners of the countries of the sub-region are the
EU, China and the United States. All ECCAS member countries have
abundant natural wealth. Most export oil, minerals and agricultural
products, and thus suffer deteriorating terms of trade as victims of the
‘raw material curse’ (Table 4.1).
Figure 4.3 shows averages for each country’s trade with the com-
munity between 1995 and 2010, and fairly illustrates the sub-regional
trade imbalance. Regarding imports, only Cameroon and Gabon have
acceptable levels. Gabon imports the most from its sub-regional part-
ners, followed by DRC although this level remains low for its size. The
Central African Republic, Chad and São Tomé and Príncipe import very
little. Angola’s imports account for only 20% of intra-regional exports.
Regarding exports, only Cameroon shows an upward, constant and
Désiré Avom and Mouhamed Mbouandi Njikam 77
Exports Imports
4000 3660
3500 3320
3000 2690
2460
2500
2000
1500 1310
1113
1000 920 860
670 600 490
500 320
140 0 0 120 190 6,7 34 76 86 80
0
la
go
on
AR
da
d
nd
ne
ha
ST
M
R
go
on
ab
n
ru
D
C
ui
wa
C
An
G
Bu
Exports Imports
Figure 4.3 Exports and imports within ECCAS, averages in USD thousands
Note: CAR: Central African Republic; CMR: Cameroon; DRC: Democratic Republic of the
Congo; STP: São Tomé & Prìncipe.
Source: Drawn from UNCTAD data.
sustained trend over time. The Central African Republic, Chad and São
Tomé and Príncipe contribute almost zero to sub-regional exports.
Geographical orientation
The last aspect of ECCAS’ trade is the geographical orientation of
imports and exports. Graph 4.5 shows the trend in total ECCAS trade
with some partners. The volume of foreign trade has increased steadily
since 1995, when the WTO replaced the GATT and trade liberalization
worldwide accelerated with the EU, Africa’s historic trading partner, at
the forefront. A major event since then is China’s breakthrough and
the general vibrancy of its recent trade with Africa. During the period
under review, beside the EU and China, ECCAS’ main partner in terms
of imports was the United States, which has intensified its trade with
African countries since the African Growth Opportunity Act (AGOA) in
2000. Africa is the fourth trading partner for ECCAS, ranking last in
terms of the origin of its own imports, due to a low level of diversifi-
cation and similarity in national production structures (see Figure 4.4).
78 Market Integration in ECCAS
45,000,000
40,000,000
35,000,000
30,000,000
25,000,000
20,000,000
15,000,000
10,000,000
5,000,000
0
1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
China EU27 United States Africa ECCAS
Theoretical basis
The gravity model is a generic name for the family of quantitative
models developed by Astronaut Stewart in 1940, which have been very
successful since the early 1960s (Evenett and Keller, 2002). The analyt-
ical framework assumes (i) profit maximization through competition
by monopolistic enterprises; (ii) constrained utility maximization by
consumers; and (iii) specialization in the supply of goods between coun-
tries (Anderson, 1979; Bergstrand, 1989; Anderson and Van Wincoop,
2003; Helliwell and Schembri, 2005). Since Tinbergen (1962), the gravity
Désiré Avom and Mouhamed Mbouandi Njikam 79
model has become a popular tool for empirical analysis of foreign trade.
The model, which was initially deduced to analyze bilateral trade flows
between countries, is based on the gravity principle, which holds that
the intensity of trade between two countries is proportional to their GDP
and inversely proportional to the distance separating them. The theo-
retical basis for the gravity model was developed by Anderson (1979),
Bergstrand (1985 and 1989), Deardorff (1995) and Evenett and Keller
(2002). The gravity model proposed in this study is inspired by empirical
literature, in particular by Fontagné et al. (2002).
log Xijt = α0 + 1 log Yijt + 2 log Zij + 3 Vijt + 4 Wijt + εijt (1)
j are part of the same regional integration agreement in the year t, and
the value 0 otherwise. It is also the case of the ‘MU’ variable that takes
the value 1 if countries i and j have a common currency in the year t
and the value 0 if not.
Wijt indicates the qualitative or binary variables that are specific to
partner countries i and j but are constant over time. This is the case of
the Encl variable that takes the value 0 if countries i and j both have a
sea front, the value 1 if one of the countries is landlocked and value 2
if both countries are landlocked. Two other variables are introduced to
consider the effects of geographical or linguistic closeness on bilateral
trade. The first is the frontier variable that takes the value 1 when both
countries have a common frontier and the value 0 if not. The second,
language, takes the value 1 when both countries have a common official
language and 0 if not. The parameter α0 and the model constant, 1 , 2
and 3 are vectors of coefficients associated with the different indepen-
dent variables. Lastly, εijt is the random part of the gravity model. The
log-linear form is adopted for continuous variables, to help to interpret
the coefficients directly as elasticities.
derived solely from the values of their currencies with respect to the
dollar could distort the results of the estimate (Avom and Mignamissi,
2013).
To correct these distortions, a third two-step approach borrowed from
Fontagné et al (2002) was adopted. The first step is to estimate a bilateral
trade equation using the gravity model and the second uses the equation
in the simulation for countries in the sample or for countries outside the
sample which are trading with countries in the sample. Thus, the first
step is to adjust the simulated trade flows as follows:
Xij
X −Xij
j ij
X∗ ij = X∗ i = X∗ ij (2)
−X
X j
j ij ij
−
Here X ij represents gross simulated bilateral trade flows, that is, those
obtained directly from the model estimate, Xij the level of bilateral trade
observed and X∗ ij adjusted simulated trade flows, with:
log X̂ijt = α̂0 + ˆ 1 log Yijt + ˆ 2 log Zij + ˆ 3 Vijt + ˆ 4 Wij (3)
The second step is to compute the trade potential (TP) for a given
period as the arithmetic average of gross simulated flows and adjusted
simulated flows:
1
∗
PCijt = (X ijt + X ijt ) (4)
2
Xij −Xij
ij PCijt = 1 (X
Xij
X∗ ij =
j
X∗ i = X∗ ij X
ijt + X∗ ijt ) (5)
Xij −Xij
j
2
j
Note: p < 0. 01: significant coefficients at 1%; ∗∗ p < . 0. 5: significant coefficients at 5%;
∗∗∗ p < . 0. 1: significant coefficients at 5%.
Désiré Avom and Mouhamed Mbouandi Njikam 83
Table 4.3 Average trade flows within ECCAS 1995–2010 (USD millions)
Note: This table shows exports from the first to the second country and imports from the
second to the first (Table 4.4).
Source: Calculations by the authors using estimation data.
The findings show there is strong trade potential for most countries.
Indeed, adjusted simulated exports and imports exceed observed values.
Still, trade potential is low for countries that are geographically close.
For example, during the period under review, Angola’s predicted exports
to neighboring countries such as Burundi, Congo and DRC are less than
observed exports. Conversely, for other countries such as Cameroon,
Gabon and DRC, which are less close geographically to Angola, exports
predicted by the model are much higher than exports observed. Thus,
geographical closeness is a ‘natural’ feature that has an impact on trade
between countries.
Predicted imports between Angola and Cameroon (USD 2.81 mil-
lion) are 52 times the value of imports observed (USD 0.054 million).
Predicted imports between Cameroon and DRC (USD 8.410 million)
represent more than 247 times the value of imports observed.
The main conclusion from predictions of the gravity equation is that
distance impedes trade between ECCAS member countries. The ‘frontier
effect’ plays a decisive role in trade between these countries.
Given that trade creation potential is the difference between the trade
potential and trade actually observed, for some ECCAS member coun-
tries, episodes of trade creation (positive sign) tend to dominate those
of trade disruption (negative sign). Indeed, the leading countries (high
GDP) like Cameroon and Angola, among others, have a very high trade
potential. This situation shows the existence of potential for the cre-
ation of trade in the ECCAS sub-region. Potential for the creation of
86
4 Conclusion
Annexes
Table 4.A.1 List of sample countries
Notes
1. ASEAN is composed of Brunei, Cambodia, Indonesia, Laos, Malaysia,
Myanmar, the Philippines, Singapore, Thailand and Vietnam.
2. An example is the Andean Pact or the Cartagena Agreement concluded
between Bolivia, Colombia, Ecuador, Peru and Venezuela.
3. MERCOSUR is made up of Argentina, Brazil, Paraguay and Uruguay.
4. Cameroon, Central African Republic, Chad, Congo, Equatorial Guinea and
Gabon.
Désiré Avom and Mouhamed Mbouandi Njikam 89
5. This is the trade estimate that takes into account the explanatory elements of
external trade, as well as the supply and demand structures of various coun-
tries to compute and predict the theoretical level of trade. These potentials are
usually estimated using gravity models.
6. Herfindahl-Hirschmann developed an index which has been standardized to
obtain values ranging between 0 (zero concentration) and 1 (maximum con-
centration). If the index is close to 0, exports are less concentrated, implying
that the country exports a range of products. If it is close to 1, the focus is on
a few export products.
7. Examples include the monitoring of macroeconomic indicators such as
growth and inflation (Winters, 1993), trade flows and revealed comparative
advantages (Yeats, 1997; Frankel, 1997), etc.
8. The other two forms are actual mileage and adjusted actual mileage. Actual
distance takes two forms, namely actual ground distance and actual ship-
ping distance. The first form is generally measured as the average of the road
distance and railway distance. The second is measured by the actual length
of the sea route between the two ports of both countries. For its part, the
adjusted actual mileage implies that at equal actual mileage, the quality of
infrastructure significantly influences transport cost. The better the quality of
infrastructure, the less the transport cost.
References
Agbodji Akoété, E. (2007), ‘Intégration et échanges commerciaux intra sous-
régionaux: le cas de l’UEMOA’, Revue africaine de l’intégration, vol. 1, no. 1,
pp. 161–188.
Anderson, J. E. and Van Wincoop, E. (2003), ‘Gravity with Gravitas: A Solution
to the Border Puzzle’, American Economic Review, vol. 93, pp. 170–192.
Anderson, J. E. (1979), ‘A Theoretical Foundation for the Gravity Equation’,
American Economic Review, pp. 106–116.
Avom, D. (2005), ‘Les déterminants des échanges dans la CEMAC: une évaluation
empirique’, Economie Appliquée, vol. LVIII, pp. 127–153.
Avom, D. and Mignamissi, D. (2013), ‘Évaluation et analyse du potentiel com-
mercial dans la CEMAC’, L’Actualité Économique, Revue d’Analyse Économique,
vol. 89, no. 2, pp. 115–145.
Ben Hammouda, H. (2005), Pourquoi l’Afrique s’est elle marginalisée dans le com-
merce international?, Document de travail de la CAPC no 20, Addis-Abeba:
Mai 2005.
Bergstrand, J. H. (1985), ‘The Gravity Equation in International Trade: Some
Microeconomic Foundations and Empirical Evidence’, The Review of Economic
and Statistics, pp. 474–481.
Bergstrand, J. H. (1989), ‘The Generalized Gravity Equation, Monopolistic Com-
petition, and Factor Proportions Theory in International Trade’, Review of
Economic and Statistics, vol. 71, pp. 143–153.
Deardorff, A. (1995), ‘Determinants of Bilateral Trade: Does Gravity Work in
a Neoclassical World?’ Discussion Paper, r no. 382, University of Michigan,
presented at a conference by NBER.
90 Market Integration in ECCAS
91
92 Regional Integration and Trade in Sub-Saharan Africa
45
40
35
30
25
20
15
10
5
0
00
01
02
03
04
05
06
07
08
09
10
20
20
20
20
20
20
20
20
20
20
20
Africa European Union
Developing Developing
economies: America economies: Asia
Fuels 47.4 49.6 54.2 61.1 63.2 63.0 64.9 57.6 58.6
Manufactured 24.5 23.9 21.9 17.9 16.2 16.2 15.8 18.4 17.0
goods
Food 12.2 11.4 9.4 7.8 7.1 7.1 6.6 10.2 8.8
Agricultural 3.6 3.8 3.2 2.5 2.2 2.1 1.8 2.2 2.1
materials
Others 12.3 11.3 11.3 10.7 11.3 11.6 10.9 11.6 13.5
Source: UNCTADstat,
t 2012.
ECOWAS 33.5 6.6 3.7 12.8 3.1 0.6 2.5 1.9 2.0 14.3
SADC 32.9 7.4 4.9 9.8 3.3 0.4 2.3 9.3 5.7 11.8
COMESA 25.9 5.9 2.8 8.3 4.3 1.6 1.5 12.1 5.6 11.8
IGAD 14.9 3.5 3.3 10.5 7.1 0.6 0.7 19.1 2.7 8.3
UMA 56.7 4.6 2.1 6.3 1.1 2.4 1.8 6.5 0.4 3.4
COMESA imported more from other African countries than did IGAD
and UMA.
While COMESA’s intra-regional share of its global exports is low –
especially considering its focus on integration through trade – the share
grew during 2001–2010 (as did ECOWAS’s and SADC’s), whereas IGAD’s
intra-regional share fell (Table 5.3). Fuel and food items constituted 46%
of intra-regional exports in 2010, manufactured goods 43%.
2001 2,255 27,155 8.3 3,983 44,531 8.9 1,626 27,598 5.9 828 4,635 17.9
2002 3,144 29,031 10.8 4,467 45,992 9.7 1,739 27,198 6.4 810 5,326 15.2
2003 3,298 35,928 9.2 5,663 55,649 10.2 2,004 35,254 5.7 970 6,485 15.0
2004 4,636 46,988 9.9 6,654 68,163 9.8 2,293 43,648 5.3 982 8,192 12.0
2005 5,546 58,872 9.4 7,799 83,556 9.3 2,694 58,602 4.6 1,094 10,385 10.5
2006 5,956 75,580 7.9 8,700 96,049 9.1 2,917 75,465 3.9 1,163 11,980 9.7
2007 6,806 86,504 7.9 12,051 118,670 10.2 4,021 89,557 4.5 1,319 16,391 8.0
2008 9,476 108,909 8.7 16,010 155,156 10.3 6,676 124,576 5.4 1,640 21,194 7.7
2009 7,379 73,569 10.0 12,004 106,626 11.3 6,122 85,759 7.1 1,435 15,337 9.4
2010 9,364 101,475 9.2 14,685 148,065 9.9 8,083 109,336 7.4 1,823 18,705 9.7
4 Empirical literature
The gravity model – which has had consistent empirical success analyz-
ing, for example, migration, commuting, tourism, foreign direct invest-
ment and bilateral trade flows – has nevertheless been criticized for lack
of theoretical foundation. In response, many trade economists have
formulated theoretical justifications (Anderson, 1979; Helpman and
Krugman, 1985; Bergstrand, 1985, 1989, 1990; Helpman, 1987; Deard-
orff, 1998; Eaton and Kortum, 2002; Anderson and van Wincoop, 2003).
Prior to the gravity model, the Heckscher-Ohlin (H-O) model – based
on the Ricardian theory of comparative advantage – was standard for
explaining international trade patterns. Ricardian theory explains pat-
terns as due to differences in technology, to which the Heckscher-Ohlin
model added differences in factor endowments. But it was not immedi-
ately apparent that gravity models took differences in technology and
factor endowments into account, nor, conversely, did H-O models use
country size (physical area) as a determinant (WTO, 2010).
98 Regional Integration and Trade in Sub-Saharan Africa
6 Model specification
The traditional gravity model assumes that bilateral imports and exports
can be determined by the exporting and importing countries’ GDPs
(indicative of supply and demand, increasing trade) as well as the
distance between them (a proxy for transport costs, reducing trade), or
β β β
Tij = αGDPi 1 GDP
Pj 2 Dij3 ηij (1)
β β β β β β β β β
Tij = αY
Yi 1 Yj 2 Dij3 Ni 4 Nj 5 RECij6 Hij 7 REM Mjt 9 ηij
Mit8 REM (3)
ln Tij = ln α + β1 ln Yi + β2 ln Yj + β3 ln Dij + β4 ln (N
Ni ) + β5 ln (N
Nj )
(4)
+β6 RECij + β7 Hij + β8 ln REM
Mi + β9 ln REM
Mj + ln ηij
Tij
exp ( − μij )μij
Tij ) =
Pr (T Tij = 0, 1, 2, . . . (6)
Tij !
Poisson models were originally applied to count data, but can also
be used, as here, with non-negative continuous dependent variables
(Wooldridge, 2002).
7 The data
Table 5.6 PPML estimation of Anderson–van Wincoop gravity model with time-
varying fixed effects for countries and year (N = 31,477)
of the road network is paved (UNECA, 2010), so that it costs three times
as much to ship a car overland from Addis Ababa to Abidjan as it does
to ship it by sea from Japan to Abidjan.
Again, the ratio of exporter’s to importer’s exchange rate affected
trade positively, with total trade value increasing when goods became
cheaper to import. Consistent with Linder’s (1961) hypothesis, per-
capita income differences (as proxy for differences in preferences)
affected trade negatively, indicating that similar countries traded more.
Whereas in the augmented traditional model ECOWAS had a large
positive effect on trade, here the effect is much smaller (and ‘non-
significant’). SADC’s effect is still positive (and significant), though
smaller than before. COMESA again has a negative effect (though
smaller and ‘non-significant’). The real surprise is IGAD, which is now
found to have had a quite large, positive (and ‘statistically significant’)
effect. As Anderson and Van Wincoop (2003) and Silva and Tenreyro
(2006) showed this specification handles the problem of omitted vari-
able bias, country pair fixed effects and control for multilateral trade
resistance. Hence the positive coefficient on IGAD from the Anderson
and Van Wincoop model might give sense because some country pairs
in IGAD such as Kenya and Uganda are member of a more advanced
REC, that is, East African Community (EAC), and have extensive trade
relations.
Annexes
References
Alemayehu, G. and Haile, K. (2008), ‘Regional Economic Integration in Africa:
A Review of Problems and Prospects with a Case Study of COMESA’, Journal of
African Economies, vol. 17, no. 3, pp. 357–394.
Anderson, J. E. (1979), ‘A Theoretical Foundation for the Gravity Equation’,
American Economic Review, vol. 69, no. 1, pp. 106–116.
Anderson, J. E. and van Wincoop, E. (2003), ‘Gravity with Gravitas: A Solution to
the Border Puzzle’, American Economic Review, vol. 93, no. 1, pp. 170–192.
Baldwin, R. E. and Venables, A. J. (1995), ‘Regional Economic Integration’, in
G. M. Grossman and K. Rogoff (eds.), Handbook of International Economics,
vol. 3, Amsterdam: North-Holland, pp. 1597–1644.
Balwin, R. and Taglioni, D. (2006), ‘Gravity for Dummies and Dummies for
Gravity Equation’, NBER Working Paper Series, no. 12516.
Bergstrand, J. H. (1985), ‘The Gravity Equation in International Trade: Some
Microeconomic Foundations and Empirical Evidence’, The Review of Economics
and Statistics, vol. 67, no. 3, pp. 474–481.
Bergstrand, J. H. (1989), ‘The Generalized Gravity Equation, Monopolistic Com-
petition, and the Factor-Proportions Theory in International Trade’, The Review
of Economics and Statistics, vol. 71, no. 1, pp. 143–153.
Bergstrand, J. H. (1990), ‘The Heckscher-Ohlin-Samuelson Model, the Linder
Hypothesis, and the Determinants of Bilateral Intra-Industry Trade’, Economic
Journal, vol. 100, pp. 1216–29.
Brun, J. F., Carrere, C., Guillaumont, P. and Melo, J. (2005), ‘Has Distance Died?
Evidence from a Panel Gravity Model’, World Bank Economic Review, vol. 19,
no. 1, pp. 99–120.
Cassim, R. (2001), ‘The Determinants of Intra-Regional Trade in Southern Africa
with Specific Reference to South Africa and the Rest of the Region’, Development
Policy Research Unit Working Papers, no. 01/51.
Cheng, I.-H. and Wall, H. J. (2005), ‘Controlling for Heterogeneity in Gravity
Models of Trade and Integration’, Federal Reserve Bank of St. Louis Review, vol. 87,
no. 1, pp. 49–63.
Darvas, Z. (2012), ‘Real Effective Exchange Rates for 178 Countries: A New
Database’, Bruegel Working Paper, r no. 2012/06.
Deardorff, A. (1998), ‘Determinants of Bilateral Trade: Does Gravity Work in a
Neoclassical World?’, in Jeffrey A. Frankel (ed.), The Regionalization of the World
Economy, vol. 1, Chicago: University of Chicago Press, pp. 7–32.
DeRosa, D. A. (2008), Prospects for Greater Global and Regional Integration in the
Maghreb, Washington, D.C.: Peterson Institute of International Economics.
Eaton, J. and Kortum, S. (2002), ‘Technology, Geography and Trade’,
Econometrica, vol. 70, no. 5, pp. 1741–1779.
Eichengreen, B. and D. A. Irwin (1995), ‘Trade Blocs, Currency Blocs and the
Reorientation of World Trade in the 1930s’, Journal of International Economics,
vol. 38, pp. 1–24.
Fenstra, R. (2006), ‘New Evidence on the Gains from Trade’, Review of World
Economics, vol. 142, no. 4, pp. 617–641.
Foroutan, F. and Pritchett, L. (1993), ‘Intra-Sub-Saharan African Trade: Is it too
little?’, Journal of African Economies, vol. 2, no. 1, pp. 74–105.
Edris Seid 107
Grinols, Earl L. (1981), ‘An Extension of the Kemp-Wan Theorem on the For-
mation of Customs Unions’, Journal of International Economics, vol. 11, no. 2,
pp. 259–266.
Helpman, E. and Krugman, P. R. (1985), Market Structure and Foreign Trade: Increas-
ing Returns, Imperfect Competition, and the International Economy, Cambridge:
MIT Press.
Helpman, E. (1987), ‘Imperfect Competition and International Trade: Evidence
from Fourteen Industrial Countries’, Journal of the Japanese and International
Economies, vol. 1, pp. 62–81.
IFC and World Bank (2010), Doing Business 2010: Reforming Through Difficult
Times, Washington, D.C.: IFC, Palgrave MacMillan and World Bank.
IMF (2011), Direction of Trade Statistics (DOTS) Database, CD-ROM.
Kemp, M. and Jr. HenryWan (1976), ‘An Elementary Proposition Concerning the
Formation of Customs Unions’, in Marry Kemp (eds.), Three Topics in the Theory
of International Trade: Distribution, Welfare and Uncertainty, Amsterdam: North
Holland, pp. 95–97.
Krishna, P. and Panagariya, A. (2002), ‘On Necessarily Welfare-Enhancing Free
Trade Areas’, Journal of International Economics, vol. 57, no. 2, pp. 353–367.
Linder, S. B. (1961), An Essay on Trade and Transformation, New York: Wiley
and Sons.
Linnemann, H. (1966), An Econometric Study of International Trade Flows,
Amsterdam: North-Holland.
Makochekanwa, A. (2012), ‘Impacts of Regional Trade Agreements on Trade in
Agrifood Products: Evidence from Eastern and Southern Africa’ Paper presented
at the African Economic Conference 2012, Kigali, Rwanda.
Martinez-Zarzoso, I. and Nowak-Lehmann, F. (2001), ‘Augmented Gravity Model:
An Empirical Application to Mercosur-European Union Trade Flows’, Journal of
Applied Economics, vol. 4, no. 2, pp. 291–316.
Mayer and Zignago (2011), ‘Notes on CEPII’s Distances Measures (GeoDist)’,
CEPII Working Paper, r 2011–2025, available from: http://www.cepii.fr/
anglaisgraph/bdd/distances.htm#sthash.qJGNVtLw.dpuf.
Poyhonen, P. (1963), ‘A Tentative Model for the Volume of Trade Between
Countries’, Weltwirtschaftliches Archiv, no. 90, pp. 93–100.
Silva, J. M. C. and Tenreyro, S. (2006), ‘The Log of Gravity’, The Review of
Economics and Statistics, vol. 88, no. 4, pp. 641–658.
Tinbergen, J. (1962), Shaping the World Economy, New York: The Twentieth
Century Fund.
UNCTADstat (2012), United Nations Conference on Trade and Development, t avail-
able from: http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx?sCS_
referer=&ScsChosenLang=en.
UNECA (2004), ‘Assessing Regional Integration in Africa’, ECA Policy Research
Report,
t Addis Ababa.
UNECA (2010), ‘Assessing Regional Integration in Africa IV: Enhancing Intra-
African Trade’, ECA Policy Research Report,t Addis Ababa.
Viner, J. (1950), The Customs Union Issue, New York: The Carnegie Endowment for
International Peace.
Wooldridge, J. M. (2002), Econometric Analysis of Cross Section and Panel Data,
Cambridge, MA: MIT Press.
108 Regional Integration and Trade in Sub-Saharan Africa
111
112 Malawi’s Trade Policies
Notes: Due to limitations on data availability, growth during import substitution industri-
alization was calculated over the period 1973–1981 and during trade liberalization/export
promotion over 1995–2002. Chirwa (2003) and Chirwa and Zakeyo (2003) report 6.7%
growth of manufacturing value-added during import substitution industrialization.
Source: Author’s calculations based on data from the Reserve Bank of Malawi Financial
and Economic Review (various issues).
Hopestone Kayiska Chavula 113
40 0.7
0.6
30
0.5
Price−cost margin
20
Growth rate %
0.4
10
0.3
0
0.2
–10
0.1
–20 0
1970 1974 1978 1982 1986 1990 1994 1998
Manufacturing growth rate
Price−cost margin
Figure 6.1 Price–cost margins and manufacturing output growth rate trends –
1967–2002
Source: Author’s calculations based on data from the National Statistical Office and the
Reserve Bank of Malawi Financial and Economic Review (various issues).
where the subscript i indicates the observation for the ith firm (i =
1, 2, . . . , N, where N is the number of firms in year t, 1967–2002), while
k identifies industry-level variables. DU Umt = 1 for t > Tbm and 0 otherwise,
for m = 1, 2 which is the number of structural breaks during the period.
Tb1 and Tb2 are break points to be located by grid search – expected to
coincide (with possible lag) the policy changes discussed in Section 2 –
to be located by grid-search following Baum (2001). This will let the data
reveal when policy changes actually affected manufacturing price–cost
margins.
Firm-level panel data from 141 firms in 14 four-digit ISIC industries
during 1967–2002 was obtained from responses to annual question-
naires returned to Malawi’s National Statistical Office as part of NSO’s
industrial production survey of medium and large for-profit firms. Gross
manufacturing output was deflated by the GDP deflator while capital
stock was estimated by the conventional perpetual inventory method,
taking into account the book value of all fixed assets (land and improve-
ments, buildings and other constructions, machinery, transport and
other equipment, including gross additions less depreciation). Capital
was deflated by the price of imports since most manufacturing capital is
imported.
Import and export intensities were measured at 3-digit industry level –
using data from the NSO’s (National Statistical Office, 2006) Annual
Statements of External Trade – because firm-level data was scanty. The
data was categorized and linked to the three-digit ISIC classification used
by NSO based on Malawi Foreign Trade Classification for 1984, which
Hopestone Kayiska Chavula 117
Choice of models
Model 1 (Table 6.2) is based on all 1583 observations from 1967–2002,
using unbalanced panel data, Model 2 on those from 1967–1982 (that is,
until the first structural break), Model 3 on those from 1983–1991 (that
is, between the structural breaks) and Model 4 on those from 1992–2002
(after the second break). Hausman tests were used to choose fixed-effects
(Models 1, 2 and 3) or random-effects (Model 4). The Wooldridge test
revealed first-order autocorrelation (AR1) in Models 1, 3 and 4 while
the Wald test also revealed panel heteroskedasticity in Models 1, 2 and
3. Thus Feasible Generalized Least Squares (FGLS) – which produces con-
sistent and efficient estimates in the presence of AR(1) within panels
and cross-sectional correlation and heteroskedasticity across panels if
the time period (T) is greater than the cross-sectional units (N) (Beck
and Katy, 1995) – was used for estimating Models 1 and 3. Groupwise
heteroskedasticity without first-order autocorrelation in Models 2 and 4
prompted the use of statistical software STATA’s ‘Robust’ command to
obtain robust estimates.
Table 6.2 Effects of trade policies and market structure on manufacturing price–
cost margins, 1967–2002
Notes: The values in parentheses are t-values; ∗ denotes significant at 10% level; ∗∗ 5%; and
∗∗∗ 1%.
Hopestone Kayiska Chavula 119
price–cost margins in all four models, that is, in the entire study period
as well as in each sub-period (Table 6.2). This contrasts with Kaluwa
and Reid (1991) – where the concentration variable was found to be
statistically non-significant and with the wrong sign – but their very
short study period (1969–1972) was also characterized by restrictions on
entry to manufacturing as well as by price controls on certain products –
where increases had to be approved by the Ministry of Trade and Indus-
try, which could only be justified by exogenous cost increases (Kaluwa,
1986) – hence their non-significant effect.
Consistent with Kaluwa and Reid (1991), import intensity (imp) was
found to have had negative (and highly significant) effects – indicating
that increased imports led to reduced price–cost margins – in Models 1, 2
and 4. These results are consistent with the imports-as-market-discipline
hypothesis, whereby it is expected that price–cost margins fall under
the pressure of imports (even – or especially – during import substitu-
tion industrialization), especially for firms that previously had market
power (Lopez and Lopez, 2003; Culha and Yalcin, 2005). The effect
was much less after 1982, which could imply improvement in manu-
facturing price–cost margins due to reduced costs of production after
liberalization.
Consistent with House (1973), Lopez and Lopez (2003) and Culha and
Yalcin (2005), as well as Kaluwa and Reid (1991), export intensity (exp)
was found to have had negative effects in Models 2 and 4, indicating
that – while possibly forced to reduce prices in order to export – more
export-oriented firms can also better exploit economies of scale and thus
operate on smaller price–cost margins.
Perhaps because they are mostly imported, leading to higher pro-
duction costs (Kaluwa, 1986), raw materials (rms) were found to have
had a negative effect on price–cost margins in all four models, with
the largest effect during import substitution industrialization (Model 2).
Skilled labor (skill) was also found to have had a negative effect over-
all (Model 1) and especially after liberalization (Model 4), perhaps then
reflecting improved labor mobility so that the effect of skilled wages
showed up more clearly. In contrast to Kaluwa and Reid (1991), finance
capital (fink) was also found to have had a negative effect in Mod-
els 1, 2 and (the largest effect, in Model) 3, perhaps reflecting the then
newly somewhat-liberalized financial sector increasing the availability
of finance, leading to a reduction in price–cost margins.
The capital/labor ratio (k/l) – a possible barrier to entry – was found
to have had a positive effect in all models, highest in Model 3 (dur-
ing structural adjustment programs). Perhaps liberalization, which led
120 Malawi’s Trade Policies
intensity, raw materials and skilled labor were lower in Model 1, whereas
the value for capital/labor ratio was higher. Values for export intensity
and minimum efficient scale, which were not statistically significant in
Model 1, were so in Model 4, whereas the value for finance capital,
which was statistically significant in Model 1, was not in Model 4.
Comparing variable coefficients among Models 2, 3 and 4 – covering
each of the policy periods under study – the fact that values for
market concentration, import and export intensities and raw materi-
als were highest in Model 2 indicates that those variables had most
effect on price–cost margins during import substitution industrialization
(1967–1982).
Annexes
Description Definitions
–.5
–1
lpcm
–1.5
–2
–2.5
1970 1980 1990 2000
Year
D.lpcm
1.5
1
D.lpcm
.5
0
–.5
–1
1970 1980 1990 2000
Year
References
Ahsan, A., Machinjili, C. and Mkwezalamba, M. M. (1999), The Impact of Adjust-
ment Policy Reforms on Malawi’s Manufacturing Sector, r Lilongwe: World Bank,
Malawi Field Office.
Bain, J. S. (1951), ‘Relation of Profit Rate to Industry Concentration: American
Manufacturing 1936–1940’, Quarterly Journal of Economics, vol. 65, no. 3,
pp. 293–324.
Baum Christopher, F. (2001), ‘Stata: The Language of Choice for Time Series
Analysis’, The Stata Journal, vol. 1, no. 1, pp. 1–16.
Beck, N. and Katy, J. N. (1995), ‘What To Do (and Not To Do) with Time Series
Cross-Section Data’, American Political Science Review, vol. 89, pp. 634–647.
Chirwa, E. W. (2001), ‘Privatization and Technical Efficiency: Evidence from
the Manufacturing Sector in Malawi’, Working Paper, r Economics Department,
University of Malawi.
Chirwa, E. W. (2003), ‘Trade Policy and Industrialisation in Malawi: The Need for
a Strategic Approach’, Working Paper, University of Malawi, Zomba.
Chirwa, E. W. and Mlachila, M. (2004), ‘Financial Reforms and Interest Rate
Spreads in the Commercial Banking System in Malawi’, IMF Staff Papers, vol. 51,
no. 1, pp. 96–122.
124 Malawi’s Trade Policies
Introduction
125
126 South-South and North-South Trade
1 Literature review
Other studies (Coe and Helpman, 1995; Keller, 1999) have established
a positive correlation between goods importation and technology dis-
semination. However, Higino Schneider (2005) has demonstrated that
although the importation of high-technology goods can definitely
contribute to the industrial development in both developed and devel-
oping countries through innovation, this effect is observed mostly in
developed countries.
According to Frankel and Romer (1999), international trade boosts
industrial production through several channels:
2 Scope of study
This study covers 47 African countries during the 1971–2010 period for
averages of five years. In other words, the study period is T = 40/5 = 8,
chosen because the main study variables such as trade openness and
customs duties vary little from year to year. The IMF database and
UN Comtrade base are used. Table 7.1 shows an industrialization rate
Table 7.1 Descriptive statistics
3 Methodology
The authors Hossein and Weiss (1999) studied the industrialization pro-
cess in absolute terms and in relative terms. In absolute terms, analysis
of the industrialization process is based on the value added of the
secondary sector; while in relative terms, analysis focuses on the indus-
trialization level, that is, share of the absolute value of the secondary
sector in GDP. In this study, industrialization is attributable to internal
factors such as GDP, urbanization, natural resources and external factors,
in particular trade openness. Although interesting, this approach seems
incomplete as it overlooks the impact of factors such as physical cap-
ital formation, North-South trade and especially South-South trade on
industrialization.3 The approach adopted here aims to complement that
of Rowthorn and Coutts (2004) to estimate the impact of South-South
trade and North-South trade on industrialization in Africa. To this end,
the model is depicted in equation (1):
INDi,t = θi + φXi, t + ϕY
Yi,t + i,t (7.1)
Here, IND is the industrialization rate (relation between the value added
of the secondary sector on GDP), i stands for the country, t is time, X is a
vector which specifies the internal factors and Y is a vector that specifies
external factors, ϕ and φ are parameters to be estimated and ψ an error
term. θi stands for the country fixed effect.
The internal variables used are real GDP per capita, GDP2 to
test the relationship U between the GDP and industrialization or
Henri Atangana Ondoa and Henri Ngoa Tabi 131
To this end, Blundell and Bond (1998), Arellano and Bover (1995) devel-
oped a method to use lagged but differentiated variables as instruments.
The advantage is that this estimator identifies the endogenous variables
of the model and thus the instruments, and sets limits to the lag of
the differentiated endogenous variables used as instruments. The latter
method is used in this study.
Equation (1) is estimated in first-order differences and all quantita-
tive variables in the study are in logarithmic form. This transformation
helps to reduce the magnitude of fluctuations and stabilize variances.
Variables are stationary in first-order differences, which confirms that
there is no correlation between the fixed effects and the explanatory
variables in first-order differences.5
4 Results
The estimation results obtained are shown in Tables 7.2, 7.3, 7.A.1 and
7.A.2. In Tables 7.2 and 7.3, the dependent variable is the industrial-
ization rate. In Tables 7.A.1 and 7.A.2, the dependent variable is the
growth rate of the industries’ added value. In both cases (industrial-
ization rate and value added), the results are overall significant and
pose no econometric problems. The Sargan tests performed show that
endogeneity problems have been corrected. These results are presented
in two parts. First, the impact of internal factors on industrializa-
tion in Africa are analyzed. Then, the impact of trade openness on
industrialization in Africa is studied.
Lagged endogenous 0.0372 0.0471 −0.163∗∗ −0.00116 0.0457 0.0599 −0.161∗∗ 0.0239
(0.0464) (0.0482) (0.0680) (0.0605) (0.0470) (0.0488) (0.0688) (0.0615)
Customs duties(CD) 0.256∗∗∗ 0.177∗∗∗ 0.395∗∗∗ 0.123∗ 0.253∗∗∗ 0.174∗∗∗ 0.396∗∗∗ 0.0929
(0.0582) (0.0664) (0.100) (0.0731) (0.0583) (0.0664) (0.101) (0.0730)
duration of war in 5-year −0.0168 −0.00633 −0.00577 0.0162 −0.0192 −0.00584 −0.00486 0.00555
(0.0232) (0.0234) (0.0337) (0.0295) (0.0234) (0.0235) (0.0341) (0.0295)
Urbanization rate 0.158∗∗∗ 0.155∗∗∗ 0.265∗∗∗ 0.112 0.167∗∗∗ 0.161∗∗∗ 0.269∗∗∗ 0.142∗∗
(0.0531) (0.0545) (0.0875) (0.0696) (0.0535) (0.0548) (0.0888) (0.0697)
Increase in degree of trade 0.225∗∗∗ 0.226∗∗∗ 0.182∗∗∗ 0.351∗∗∗ 0.227∗∗∗ 0.221∗∗∗ 0.189∗∗∗ 0.371∗∗∗
openness/DvgC (0.0397) (0.0396) (0.0551) (0.0694) (0.0402) (0.0402) (0.0567) (0.0695)
Increase in degree of trade −0.0428∗ −0.0346 −0.0372 −0.0516∗ −0.0428∗ −0.0413∗ −0.0290 −0.0505∗
openness/DvgC Asia (0.0228) (0.0235) (0.0334) (0.0281) (0.0237) (0.0245) (0.0364) (0.0288)
Increase in degree of trade −0.0470 −0.0200 −0.189∗∗∗ 0.0438 −0.0622∗ −0.0437 −0.176∗∗∗ 0.00614
openness/DvgC Africa (0.0315) (0.0323) (0.0557) (0.0399) (0.0343) (0.0352) (0.0636) (0.0448)
GDP growth rate (GR) 0.0435∗∗∗ 0.0592∗∗∗ 0.0654∗∗ 0.00950 0.0422∗∗ 0.0590∗∗∗ 0.0685∗∗ 0.00470
(0.0167) (0.0174) (0.0282) (0.0198) (0.0167) (0.0175) (0.0291) (0.0197)
GDP* GR growth rate −0.00251 −0.00305 0.000791 −0.00606∗ −0.00276 −0.00324 0.000661 −0.00630∗
(0.00267) (0.00288) (0.00509) (0.00363) (0.00268) (0.00290) (0.00515) (0.00359)
Investment growth rate 0.0248 0.0238 0.00238 −0.00121 0.0241 0.0238 0.00190 0.000562
(0.0161) (0.0171) (0.0239) (0.0185) (0.0161) (0.0170) (0.0242) (0.0183)
Labor force growth rate 0.0591∗∗∗ 0.0636∗∗∗ 0.0193 0.0828∗∗∗ 0.0610∗∗∗ 0.0673∗∗∗ 0.0160 0.0936∗∗∗
(0.0134) (0.0136) (0.0226) (0.0158) (0.0137) (0.0138) (0.0235) (0.0165)
133
Table 7.2 (Continued)
134
Increase in school 0.0912∗∗ 0.0799∗∗ 0.219∗∗∗ 0.0926∗ 0.0822∗∗ 0.0646∗ 0.228∗∗∗ 0.107∗∗
enrolment rate (0.0363) (0.0359) (0.0546) (0.0517) (0.0373) (0.0371) (0.0573) (0.0521)
(Increase in degree of trade −0.00548 0.000813 −0.00594 −0.0124
openness/DvgCAsia)∗ (CD) (0.00730) (0.00772) (0.0106) (0.00913)
(Increase in degree of trade 0.00956 0.0117 −0.00466 0.0224∗∗
openness/DvgCAfr)∗ (CD) (0.00701) (0.00716) (0.0125) (0.00925)
Years (1976–1980) 0.389 0.775 2.137∗∗ 0.522 0.325 0.706 2.056∗∗ 0.359
(0.475) (0.539) (0.912) (0.531) (0.479) (0.540) (0.930) (0.537)
Years (1981–1985) 1.356∗∗∗ 1.530∗∗∗ 1.907∗∗ 1.495∗∗ 1.320∗∗ 1.554∗∗∗ 1.874∗∗ 1.505∗∗
(0.516) (0.542) (0.875) (0.636) (0.518) (0.545) (0.887) (0.633)
Years (1986–1990) 1.379∗∗∗ 1.385∗∗∗ −0.0131 1.294∗∗ 1.407∗∗∗ 1.454∗∗∗ −0.138 1.417∗∗
(0.499) (0.516) (1.037) (0.613) (0.501) (0.518) (1.065) (0.606)
Years (1991–1995) 0.604 0.635 0.847 0.498 0.581 0.663∗ 0.785 0.325
(0.395) (0.396) (0.552) (0.637) (0.397) (0.399) (0.566) (0.633)
Years (1996–2000) 0.667 0.897 1.271∗ 1.140∗∗ 0.609 0.868 1.199∗ 0.973∗
(0.603) (0.761) (0.692) (0.516) (0.604) (0.760) (0.710) (0.513)
Years (2001–2005) 1.541∗∗ 1.806∗∗ 1.849∗∗∗ 1.416∗∗ 1.557∗∗ 1.828∗∗ 1.792∗∗ 1.493∗∗∗
(0.644) (0.797) (0.688) (0.584) (0.644) (0.797) (0.700) (0.577)
Years (2006–2010) 0.772 0.743 1.041∗∗ 0.559 0.733 0.750 0.976∗∗ 0.398
(0.527) (0.536) (0.467) (0.493) (0.528) (0.538) (0.480) (0.491)
Observations 318 286 152 153 318 286 152 153
Number of 05 year periods 8 8 8 8 8 8 8 8
Sargan Test 336(0.00) 316(0.00) 280(0.010) 251(0.011) 334(0.00) 314(0.00) 136(0.013) 149(0.012)
Note: Dependent variable: industrialization rate (Added value of industries on GDP); DEP dependency ratio to raw materials, SSA: Sub-Saharan Africa.
∗∗∗ p < 0. 01, ∗∗ p < 0. 05, ∗ p < 0.
Lagged endogenous −0.39∗∗∗ −0.385∗∗∗ −0.0131 −0.294∗∗ −0.407∗∗∗ −0.454∗∗∗ −0.138 −0.417∗∗
(0.049) (0.016) (1.037) (0.0613) (0.02501) (0.018) (1.065) (0.0612)
Customs duties (CD) 0.349∗∗∗ 0.324∗∗∗ 0.549∗∗∗ 0.187∗∗ 0.343∗∗∗ 0.319∗∗∗ 0.552∗∗∗ 0.176∗∗
(0.0662) (0.0707) (0.117) (0.0859) (0.0667) (0.0712) (0.119) (0.0874)
Average duration of war in 0.0189 0.0373 −0.0146 0.0615∗∗ 0.0148 0.0346 −0.0170 0.0572∗
5-year periods (0.0262) (0.0268) (0.0409) (0.0306) (0.0265) (0.0273) (0.0415) (0.0312)
Urbanization rate 0.245∗∗∗ 0.270∗∗∗ 0.520∗∗∗ 0.183∗∗ 0.253∗∗∗ 0.279∗∗∗ 0.526∗∗∗ 0.195∗∗∗
(0.0588) (0.0604) (0.104) (0.0711) (0.0595) (0.0611) (0.106) (0.0727)
Increase in degree of trade 0.180∗∗∗ 0.180∗∗∗ 0.158∗∗ 0.229∗∗∗ 0.188∗∗∗ 0.184∗∗∗ 0.192∗∗∗ 0.236∗∗∗
openness DC (0.0476) (0.0492) (0.0705) (0.0734) (0.0487) (0.0507) (0.0740) (0.0756)
Increase in degree of trade −0.000396 0.0140 −0.0380 0.0285 0.00484 0.0161 −0.0118 0.0306
openness/DvgC Asia (0.0268) (0.0278) (0.0395) (0.0308) (0.0277) (0.0290) (0.0426) (0.0317)
Increase in degree of trade −0.093∗∗∗ −0.0655∗ −0.218∗∗∗ −0.0347 −0.0963∗∗ −0.0781∗∗ −0.202∗∗∗ −0.0482
openness/DvgC Africa (0.0333) (0.0355) (0.0627) (0.0404) (0.0376) (0.0395) (0.0724) (0.0480)
GDP growth rate (GR) 0.0622∗∗∗ 0.0714∗∗∗ 0.0516 0.0160 0.0600∗∗∗ 0.0681∗∗∗ 0.0574∗ 0.0132
(0.0191) (0.0205) (0.0320) (0.0210) (0.0193) (0.0208) (0.0333) (0.0213)
GDP* GR growth rate −0.00159 −0.00304 −0.000120 −0.011∗∗∗ −0.00191 −0.00345 −0.000210 −0.011∗∗∗
(0.00312) (0.00339) (0.00620) (0.00379) (0.00315) (0.00343) (0.00632) (0.00382)
Investment growth rate 0.0219 0.0298 0.0167 0.000338 0.0207 0.0288 0.0182 0.00130
(0.0189) (0.0203) (0.0280) (0.0201) (0.0190) (0.0204) (0.0285) (0.0203)
Labor force growth rate 0.0765∗∗∗ 0.0801∗∗∗ 0.0377 0.0890∗∗∗ 0.0762∗∗∗ 0.0808∗∗∗ 0.0308 0.0921∗∗∗
(0.0150) (0.0161) (0.0274) (0.0176) (0.0153) (0.0164) (0.0285) (0.0183)
135
Table 7.3 (Continued)
136
Increase in school 0.0760∗ −0.0643 0.161∗∗ 0.175∗∗∗ 0.0746∗ −0.0559 0.176∗∗ 0.182∗∗∗
enrolment rate (0.0392) (0.0402) (0.0666) (0.0524) (0.0408) (0.0422) (0.0704) (0.0551)
(Increase in degree of trade −0.00916 −0.00679 −0.0229∗ −0.00529
openness/DvgC Asia)∗ (CD) (0.00869) (0.00951) (0.0130) (0.00996)
(Increase in degree of trade 0.00480 0.00875 −0.00645 0.00815
openness/DvgCAfr)∗ (CD) (0.00832) (0.00873) (0.0152) (0.0100)
Years (1976–1980) −0.735 −0.616 −0.0879 0.313 −0.753 −0.658 −0.111 0.322
(0.606) (0.712) (0.855) (0.921) (0.608) (0.717) (0.872) (0.927)
Years (1981–1985) 0.737 0.579 −2.649 1.114 0.710 0.658 −3.132∗ 1.118
(0.913) (1.065) (1.764) (0.924) (0.931) (1.081) (1.829) (0.965)
Years (1986–1990) 0.359 −0.0194 0.0659 1.214 0.268 −0.170 −0.131 1.129
(0.715) (0.935) (1.003) (1.256) (0.722) (0.948) (1.029) (1.270)
Years (1991–1995) 0.601 1.248 −0.782 2.105∗∗ 0.527 1.356 −1.095 2.197∗∗
(0.902) (1.334) (1.051) (0.901) (0.918) (1.361) (1.082) (0.929)
Years (1996–2000) 1.083 0.742 1.443 0.583 1.099 0.740 1.393 0.539
(1.157) (1.205) (1.282) (0.530) (1.161) (1.212) (1.310) (0.543)
Years (2001–2005) −0.595 −0.669 −0.824 0.0539 −0.620 −0.638 −1.077 0.0461
(0.777) (0.831) (0.654) (0.735) (0.783) (0.841) (0.680) (0.740)
Years (2006–2010) −0.558 −0.834 −0.610 0.390 −0.627 −0.947 −0.818 0.323
(0.629) (0.747) (0.519) (0.729) (0.633) (0.757) (0.541) (0.742)
Observations 318 286 152 153 318 286 152 153
Number of 05 year periods 8 8 8 8 8 8 8 8
Sargan Test 340(0.00) 305(0.00) 154(0.015) 178(0.012) 337(0.00) 302(0.00) 145(0.037) 176(0.022)
Note: Dependent variable: industrialization rate (added value to manufacturing industries on GDP); DEP dependency ratio to raw materials, SSA: Sub-
Saharan Africa. ∗∗∗ p < 0. 01, ∗∗ p < 0. 05, ∗ p < 0.
Source: Authors’ estimate.
Henri Atangana Ondoa and Henri Ngoa Tabi 137
Such industries sell their products easily in big cities due to concentra-
tion of demand. This result had already been established by Hausmann
and Rodrik (2006) and Brady et al. (2011), when they demonstrated
the importance of factors such as domestic demand and urbanization
in industrialization.
Customs duties and schooling also contribute to industrial develop-
ment. However, the contribution of schooling has been more clearly
demonstrated in natural resource-rich countries. These results contra-
dict some previously conducted work. According to Wood and Mayer
(1998), human resources and natural resources constitute the key deter-
mining factors of the comparative advantage of manufacturing exports.
In this regard, Africa is unable to increase the volume of its manufactur-
ing exports because it is more endowed with natural and not human
resources. Similarly, AfDB (2011) notes that competitiveness in non-
mining sub-sectors is affected by the appreciation of the real exchange
rate against that of their competitors due to raw material exports, in
accordance with the Dutch disease mechanisms.6 Yet, the contribution
of the building and public works sub-sector to GDP is particularly strong
in oil-producing countries because authorities in these countries often
use oil revenues to finance the development of public infrastructure
such as roads, schools, hospitals and government buildings.
Table 7.4 Contribution of trade with countries in transition to development of manufacturing industries in Africa
Lagged endogenous −0.295∗∗∗ −0.278∗∗∗ −0.439∗∗∗ −0.367∗∗∗ −0.289∗∗∗ −0.273∗∗∗ −0.423∗∗∗ −0.358∗∗∗
(0.0322) (0.0343) (0.0555) (0.0424) (0.0323) (0.0346) (0.0554) (0.0423)
Average duration of war – 0.0979 0.145 0.0374 0.288∗∗ 0.0938 0.138 0.0147 0.302∗∗∗
5 year periods (0.0857) (0.0921) (0.136) (0.115) (0.0854) (0.0923) (0.135) (0.114)
Urbanization rate 0.0263 0.280∗∗∗ 0.12∗∗∗ 0.0557 0.0221 0.085∗∗∗ 0.110 0.0333
(0.0519) (0.0548) (0.0111) (0.0774) (0.0517) (0.0549) (0.0906) (0.0776)
Customs duties (CD) 1.181∗∗∗ 1.108∗∗∗ 2.223∗∗∗ 0.866∗∗∗ 1.167∗∗∗ 1.100∗∗∗ 2.332∗∗∗ 0.866∗∗∗
(0.244) (0.264) (0.412) (0.306) (0.243) (0.264) (0.411) (0.303)
Increase in degree of trade 0.714∗∗∗ 0.710∗∗∗ 0.939∗∗∗ 0.741∗∗∗ 0.710∗∗∗ 0.707∗∗∗ 0.912∗∗∗ 0.734∗∗∗
openness/DvdC (0.143) (0.149) (0.221) (0.236) (0.142) (0.149) (0.220) (0.234)
Increase in degree of trade −0.734∗∗∗ −0.687∗∗∗ −1.697∗∗∗ −0.393 −0.670∗∗∗ −0.646∗∗∗ −1.539∗∗∗ −0.231
openness/trans (0.150) (0.161) (0.233) (0.239) (0.157) (0.168) (0.237) (0.254)
GDP growth rate (GR) 0.891∗∗∗ 0.903∗∗∗ 0.449∗∗∗ 1.007∗∗∗ 0.885∗∗∗ 0.900∗∗∗ 0.416∗∗∗ 0.988∗∗∗
(0.0494) (0.0530) (0.0858) (0.0607) (0.0493) (0.0530) (0.0859) (0.0609)
GDP* GR growth rate 0.0303 0.0220 −0.00962 −0.0251 0.0244 0.0179 −0.0635 −0.0270
(0.0540) (0.0590) (0.0762) (0.0717) (0.0539) (0.0591) (0.0779) (0.0709)
Investment growth rate 0.0158∗ 0.0168 −0.0104 −0.00129 0.0155∗ 0.0166 −0.0111 −0.00424
(0.009) (0.0103) (0.0153) (0.0143) (0.0093) (0.0102) (0.0152) (0.0142)
Labor force growth rate 0.301∗∗∗ 0.274∗∗∗ 0.294∗∗∗ 0.232∗∗∗ 0.300∗∗∗ 0.273∗∗∗ 0.280∗∗∗ 0.231∗∗∗
(0.0571) (0.0621) (0.0868) (0.0752) (0.0569) (0.0620) (0.0865) (0.0743)
Increase in school 0.090∗∗∗ 0.095∗∗∗ 0.146∗∗∗ 0.157∗∗∗ 0.085∗∗∗ 0.090∗∗∗ 0.113∗∗ 0.148∗∗∗
enrolment rate (0.0309) (0.0321) (0.0506) (0.0499) (0.0310) (0.0325) (0.0514) (0.0496)
(DD)∗ (Increase in degree of −0.0312 −0.0223 −0.103∗∗∗ −0.0574∗
trade openness/trans (0.0228) (0.0263) (0.0339) (0.0320)
countries)
Years (1976–1980) −5.065∗ −9.017∗∗∗ −11.14∗ −4.945 −5.402∗∗ −9.309∗∗∗ −15.60∗∗ −5.526∗
(2.709) (3.289) (6.569) (3.101) (2.708) (3.300) (6.689) (3.084)
Years (1981–1985) −6.341∗∗ −0.364 −1.569 −6.144∗ −6.043∗ −0.0826 2.965 −5.964∗
(3.200) (4.171) (8.511) (3.207) (3.194) (4.177) (8.592) (3.173)
Years (1986–1990) −1.301 −5.079 −5.430 −3.748 −1.652 −5.371 −8.054 −3.907
(4.288) (4.645) (5.050) (6.301) (4.276) (4.649) (5.091) (6.232)
Years (1991–1995) −6.980 −7.651 −7.021 −7.533 −7.016 −7.624 −7.877 −8.483
(4.806) (5.337) (5.443) (5.532) (4.785) (5.327) (5.417) (5.496)
Years (1996–2000) −7.284∗∗ −3.271 −4.651 −2.694 −6.762∗∗ −2.947 −3.541 −2.508
(3.279) (3.864) (3.753) (3.244) (3.287) (3.876) (3.748) (3.210)
Years (2001–2005) −2.135 −9.751 −3.993∗ −4.773∗ −2.947 −10.48∗ −6.050∗∗∗ −5.509∗∗
(4.674) (6.032) (2.159) (2.497) (4.690) (6.081) (2.248) (2.504)
Years (2006–2010) −0.620 5.361 −1.398 −1.760 −0.436 5.648 −1.691 −1.904
(3.022) (4.524) (1.633) (2.114) (3.012) (4.529) (1.626) (2.092)
Observations 302 271 134 155 302 271 134 155
Number of 05-year periods 8 8 8 8 8 8 8 8
Sargan Test 317(0.00) 370(0.00) 220(0.02) 245(0.01) 318(0.00) 377(0.00) 214(0.023) 254(0.011)
Note: Dependent variable: industrialization rate (added value of manufacturing industries to the GDP); DEP dependency ratio to raw materials, SSA:
Sub-Saharan Africa. ∗∗∗ p < 0. 01, ∗∗ p < 0. 05, ∗ p < 0.
Source: Authors’ estimate.
139
140
Table 7.5 Contribution of trade with countries in transition to industrial development in Africa
Lagged endogenous −0.30∗∗∗ −0.28∗∗∗ −0.447∗∗∗ −0.352∗∗∗ −0.298∗∗∗ −0.27∗∗∗ −0.431∗∗∗ −0.342∗∗∗
(0.0318) (0.0338) (0.0550) (0.0419) (0.0320) (0.0341) (0.0552) (0.0418)
Average duration war – 0.109 0.144∗ 0.0134 0.288 0.105 0.141 −0.0234 0.301∗∗∗
5year (0.0766) (0.0820) (0.117) (0.303) (0.0764) (0.0923) (0.118) (0.102)
Urbanization rate 0.0385 0.0438 0.0884 0.858∗∗∗ 0.0356 0.0427∗∗∗ 0.0850 0.0656
(0.0467) (0.0492) (0.0812) (0.0695) (0.0466) (0.0093) (0.0810) (0.0697)
Customs duties (CD) 1.179∗∗∗ 1.077∗∗∗ 2.029∗∗∗ 0.897∗∗∗ 1.170∗∗∗ 1.075∗∗∗ 2.098∗∗∗ 0.892∗∗∗
(0.216) (0.234) (0.367) (0.265) (0.215) (0.234) (0.366) (0.262)
Increase in degree of trade 0.674∗∗∗ 0.658∗∗∗ 0.794∗∗∗ 0.794∗∗∗ 0.670∗∗∗ 0.657∗∗∗ 0.765∗∗∗ 0.789∗∗∗
openness/DvdC (0.127) (0.132) (0.195) (0.211) (0.126) (0.132) (0.195) (0.209)
Increase in degree of trade −0.72∗∗∗ −0.63∗∗∗ −1.590∗∗∗ −0.356∗ −0.67∗∗∗ −0.61∗∗∗ −1.447∗∗∗ −0.220
openness/trans (0.134) (0.143) (0.210) (0.216) (0.140) (0.150) (0.215) (0.228)
GDP growth rate (GR) 0.796∗∗∗ 0.811∗∗∗ 0.412∗∗∗ 0.920∗∗∗ 0.791∗∗∗ 0.809∗∗∗ 0.382∗∗∗ 0.905∗∗∗
(0.0446) (0.0475) (0.0750) (0.0555) (0.0446) (0.0476) (0.0755) (0.0556)
GDP* GR growth rate 0.0487 0.0358 0.0199 −0.00897 0.0439 0.0338 −0.0299 −0.0108
(0.0486) (0.0528) (0.0687) (0.0645) (0.0486) (0.0530) (0.0706) (0.0639)
Investment growth rate 0.274∗∗∗ 0.240∗∗∗ 0.256∗∗∗ 0.198∗∗∗ 0.273∗∗∗ 0.239∗∗∗ 0.238∗∗∗ 0.198∗∗∗
(0.0516) (0.0559) (0.0770) (0.0677) (0.0514) (0.0560) (0.0771) (0.0670)
Labor force growth rate 0.11∗∗∗ 0.11∗∗∗ 0.138∗∗∗ 0.180∗∗∗ 0.105∗∗∗ 0.11∗∗∗ 0.106∗∗ 0.172∗∗∗
(0.0278) (0.0289) (0.0455) (0.0452) (0.0279) (0.0293) (0.0466) (0.0449)
Increase in school 0.0127 0.0133 −0.0105 −0.00233 0.0125 0.0133 −0.0102 −0.00488
enrolment rate (IDT) (0.00845) (0.00920) (0.0138) (0.0127) (0.00843) (0.00920) (0.0137) (0.0127)
(DD)∗ (IDT) −0.0250 −0.0111 −0.091∗∗∗ −0.0498∗
(0.0205) (0.0235) (0.0308) (0.0289)
Years (1976–1980) −5.550∗∗ −7.63∗∗∗ −9.198∗ −4.352∗ −5.716∗∗ −7.71∗∗∗ −12.55∗∗∗ −4.762∗
(2.229) (2.488) (4.716) (2.568) (2.226) (2.492) (4.838) (2.552)
Years (1981–1985) −5.823∗∗ −2.798 −2.651 −6.966∗∗ −5.775∗∗ −2.779 0.326 −6.912∗∗
(2.499) (2.816) (5.865) (2.796) (2.492) (2.817) (5.938) (2.766)
Years (1986–1990) −2.942 −3.851 −6.254∗ −8.711∗ −3.098 −3.928 −7.636∗∗ −8.521∗
(3.508) (3.660) (3.527) (4.786) (3.500) (3.663) (3.550) (4.737)
Years (1991–1995) −4.897 −6.986 −2.575 −3.829 −4.954 −6.983 −3.412 −5.023
(4.488) (4.951) (4.635) (4.615) (4.475) (4.951) (4.633) (4.619)
Years (1996–2000) −6.116∗∗ −1.541 −5.368 −1.721 −5.773∗ −1.416 −4.304 −1.467
(2.969) (3.393) (3.332) (2.846) (2.973) (3.403) (3.343) (2.820)
Years (2001–2005) −2.684 −8.705 −3.446∗ −5.793∗∗∗ −3.361 −9.047 −5.369∗∗∗ −6.387∗∗∗
(5.601) (6.612) (1.896) (2.221) (5.610) (6.648) (1.998) (2.225)
Years (2006–2010) −1.290 2.547 −1.093 −2.456 −1.120 2.658 −1.205 −2.524
(3.394) (4.346) (1.468) (1.966) (3.386) (4.352) (1.465) (1.946)
Observations 302 271 134 155 302 271 134 155
Number of 05-year periods 8 8 8 8 8 8 8 8
Sargan Test 350(0.00) 372(0.00) 223(0.0213) 239(0.013) 348(0.00) 387(0.00) 218(0.022) 244(0.010)
Note: Dependent variable: industrialization rate (added value of industries to the GDP); DEP dependency ratio to raw materials, SSA: Sub-Saharan Africa.
∗∗∗ p < 0. 01, ∗∗ p < 0. 05, ∗ p < 0.
5 Conclusion
Lag of endogenous variable 0.287∗∗∗ 0.334∗∗∗ 0.149∗∗ 0.343∗∗∗ 0.290∗∗∗ 0.337∗∗∗ 0.161∗∗ 0.335∗∗∗
(0.0471) (0.0503) (0.0780) (0.0616) (0.0473) (0.0505) (0.0804) (0.0614)
1.183∗∗∗ 1.078∗∗∗ 2.068∗∗ 0.676∗∗∗ 1.151∗∗∗ 1.075∗∗∗ 2.075∗∗∗ 0.628∗∗∗
(0.146) (0.161) (0.226) (0.201) (0.147) (0.161) (0.231) (0.202)
0.0435 0.0677 0.131 0.0491 0.0244 0.0535 0.133 0.0180
(0.573) (0.0594) (0.0930) (0.0748) (0.0578) (0.0601) (0.0947) (0.0758)
Costums duties (CD) −0.00306 −0.0158 0.106 −0.0465 −0.00421 −0.0159 0.0988 −0.0385
(0.0389) (0.0403) (0.0712) (0.0557) (0.0391) (0.0405) (0.0730) (0.0556)
0.240∗∗∗ 0.279∗∗∗ 0.152 0.374∗∗ 0.292∗∗∗ 0.312∗∗∗ 0.175 0.452∗∗∗
(0.0912) (0.0949) (0.137) (0.156) (0.0940) (0.0984) (0.141) (0.160)
−0.21∗∗∗ −0.178∗∗∗ −0.301∗∗∗ −0.162∗∗ −0.180∗∗∗ −0.161∗∗ −0.271∗∗∗ −0.137∗
(0.0608) (0.0639) (0.0855) (0.0814) (0.0625) (0.0659) (0.0920) (0.0823)
War duration in 05 years −0.29∗∗∗ −0.218∗∗∗ −0.885∗∗∗ −0.107 −0.299∗∗∗ −0.25∗∗∗ −0.885∗∗∗ −0.155
(0.0750) (0.0803) (0.122) (0.105) (0.0854) (0.0906) (0.142) (0.121)
0.0541 0.0391 −0.0149 −0.00790 0.0446 0.0356 −0.0172 −0.00838
(0.0415) (0.0456) (0.0575) (0.0525) (0.0418) (0.0458) (0.0587) (0.0522)
Urbanization rate 0.509∗∗∗ 0.530∗∗∗ 0.247∗∗∗ 0.606∗∗∗ 0.506∗∗∗ 0.531∗∗∗ 0.244∗∗∗ 0.613∗∗∗
(0.0372) (0.0401) (0.0559) (0.0488) (0.0378) (0.0404) (0.0584) (0.0498)
0.393∗∗∗ 0.378∗∗∗ 0.310∗∗ 0.988∗∗∗ 0.390∗∗∗ 0.388∗∗∗ 0.324∗∗∗ 0.972∗∗∗
(0.0942) (0.0970) (0.0146) (0.0151) (0.0971) (0.0999) (0.0154) (0.0155)
−0.0467∗∗ −0.01317 −0.0293 −0.0463∗
(0.0190) (0.0211) (0.0273) (0.0257)
0.0141 0.0206 0.00364 0.0376
(0.0184) (0.0196) (0.0315) (0.0259)
0.751∗∗∗ 0.293∗∗∗ 0.410∗∗ 0.97∗∗∗ 0.369∗∗∗ 0.932∗∗∗ 0.552∗∗ 0.81∗∗∗
(0.854) (0.036) (0.032) (0.0249) (0.068) (0.054) (0.0602) (0.0046)
Growth rate of trade with 4.557∗∗ 5.353∗∗∗ 5.421∗ 0.689 4.821∗∗ 5.485∗∗∗ 5.806∗ 0.792
Asian LDCs (1.867) (1.928) (3.186) (2.591) (1.876) (1.935) (3.276) (2.592)
−0.0843 −0.237 2.681 −4.716 −0.606 −0.599 1.961 −4.618
(2.250) (2.276) (2.718) (5.079) (2.267) (2.297) (2.894) (5.055)
0.995∗∗∗ 0.091 0.521∗∗ 0.20∗∗∗ 0.707∗∗∗ 0.248 0.365∗ 0.58∗∗
(0.072) (0.163) (0.059) (0.020) (0.092) (0.185) (0.0817) (0.0216)
Growth rate of trade/ 4.169∗∗ 8.093∗∗∗ 5.098∗∗ 4.714∗∗∗ 4.353∗∗ 7.945∗∗∗ 5.147∗∗ 4.606∗∗∗
African countries (2.017) (2.266) (2.289) (1.351) (2.025) (2.276) (2.357) (1.344)
4.619 3.232 6.433∗∗∗ 2.182 4.771 3.391 6.158 2.344
(3.111) (3.177) (1.366) (1.646) (3.121) (3.186) (1.433) (1.647)
4.663∗∗ 4.229∗∗ 7.151∗∗∗ 3.073∗ 4.207∗∗ 3.939∗∗ 6.839∗∗∗ 2.926∗
(1.898) (1.957) (1.211) (1.595) (1.918) (1.987) (1.301) (1.586)
Observations 316 285 150 153 316 285 150 153
8 8 8 8 8 8 8 8
396 (0.00) 341 (0.000) 174 (0.001) 175 (0.01) 389 (0.000) 337 (0.00) 167 (0.012) 173 (0.013)
Note: Dependent variable: rate of industrialization (added value of industries); DEP, dependency ratio to primary commodities, SSA: Sub-Saharan Africa.
***p < 0. 01, ∗∗ p < 0. 05, ∗ p < 0.
Source: Authors’ estimate.
145
146
Lag of endogenous variable 0.0539 0.0676 −0.0117 −0.0355 0.0583 0.0730 −0.0106 −0.0329
(0.0465) (0.0500) (0.0697) (0.0592) (0.0470) (0.0505) (0.0710) (0.0599)
Customs duties (CD) 0.414∗∗∗ 0.414∗∗∗ 0.720∗∗∗ 0.183∗∗ 0.407∗∗∗ 0.407∗∗∗ 0.728∗∗∗ 0.176∗∗
(0.0675) (0.0701) (0.121) (0.0844) (0.0683) (0.0708) (0.124) (0.0863)
War duration in 05 years 0.0156 0.0303 −0.0161 0.0491 0.0121 0.0276 −0.0178 0.0478
(0.0278) (0.0280) (0.0448) (0.0308) (0.0282) (0.0285) (0.0456) (0.0314)
Urbanization rate −0.00513 −0.0127 0.0181 0.00147 −0.00495 −0.0120 0.0125 0.00220
(0.0171) (0.0175) (0.0373) (0.0204) (0.0172) (0.0176) (0.0384) (0.0207)
Growth rate of trade with 0.301∗∗∗ 0.317∗∗∗ 0.309∗∗∗ 0.347∗∗∗ 0.310∗∗∗ 0.321∗∗∗ 0.337∗∗∗ 0.353∗∗∗
developed countries (0.0447) (0.0455) (0.0724) (0.0621) (0.0464) (0.0475) (0.0775) (0.0658)
Growth rate of trade with −0.0212 −0.00705 −0.0975∗∗ 0.0104 −0.0174 −0.00559 −0.0797∗ 0.0114
Asian LDCs (0.0283) (0.0288) (0.0418) (0.0310) (0.0291) (0.0301) (0.0449) (0.0319)
Growth rate of trade with −0.112∗∗∗ −0.104∗∗∗ −0.363∗∗∗ −0.0733∗ −0.116∗∗∗ −0.117∗∗∗ −0.36∗∗∗ −0.0814∗
African countries (0.0349) (0.0362) (0.0636) (0.0400) (0.0396) (0.0406) (0.0737) (0.0479)
Growth rate of with 0.0249 0.0336 0.00686 0.00496 0.0237 0.0325 0.00802 0.00481
physical capital (0.0199) (0.0209) (0.0306) (0.0203) (0.0199) (0.0210) (0.0312) (0.0205)
Growth rate of labor force 0.0604∗∗∗ 0.0662∗∗∗ −0.00291 0.0865∗∗∗ 0.0606∗∗∗ 0.0675∗∗∗ −0.00661 0.0882∗∗∗
(0.0153) (0.0163) (0.0284) (0.0174) (0.0157) (0.0166) (0.0302) (0.0181)
Growth rate school −0.0420 −0.0274 −0.0549 0.187∗∗∗ −0.0398 −0.0193 −0.0586 0.189∗∗∗
enrollment (gross) (0.0415) (0.0420) (0.0665) (0.0543) (0.0432) (0.0441) (0.0711) (0.0564)
(Growth rate of trade/Asian −0.00705 −0.00552 −0.0178 −0.00225
LDCs)∗ (CD) (0.00914) (0.00983) (0.0144) (0.0101)
(Growth rate of trad /Afri 0.00409 0.00779 0.000502 0.00421
countries)∗ (DD) (0.00879) (0.00904) (0.0164) (0.0102)
Years (1976–1980) −0.726 −0.639 −0.352 0.849 −0.734 −0.665 −0.386 0.867
(0.641) (0.743) (0.943) (0.924) (0.643) (0.747) (0.963) (0.931)
Years (1981–1985) 1.324 1.076 −0.139 0.990 1.352 1.207 −0.405 1.032
(0.929) (1.080) (1.790) (0.942) (0.952) (1.102) (1.876) (0.991)
Years (1986–1990) 0.824 0.960 1.096 2.159∗ 0.758 0.828 1.003 2.130∗
(0.738) (0.938) (0.964) (1.261) (0.744) (0.952) (0.988) (1.272)
Years (1991–1995) 0.183 4.75e-05 −1.018 1.367 0.115 0.0890 −1.259 1.406
(0.951) (1.372) (1.100) (0.895) (0.968) (1.399) (1.139) (0.920)
Years (1996–2000) 1.612 1.307 2.712∗∗ 0.885∗ 1.665 1.361 2.713∗ 0.886
(1.200) (1.229) (1.382) (0.534) (1.206) (1.237) (1.412) (0.547)
Years (2001–2005) 0.184 −0.0844 0.0120 0.497 0.203 −0.00697 −0.161 0.519
(0.787) (0.836) (0.667) (0.737) (0.794) (0.847) (0.700) (0.744)
Years (2006–2010) −0.595 −0.560 0.368 0.585 −0.646 −0.654 0.238 0.566
(0.666) (0.779) (0.511) (0.740) (0.670) (0.789) (0.537) (0.750)
Observations 318 286 152 153 318 286 152 153
Number of 05 years periods 8 8 8 8 8 8 8 8
Test de Sargan 328 (0.00) 309(0.000) 154(0.011) 184(0.00) 326(0.000) 306(0.00) 147(0.02) 182(0.00)
Note: Dependent variable: industrialization rate (value added of manufacturing industries); DEP, dependency ratio to primary commodities, SSA: Sub-
Saharan Africa. ∗∗∗ p < 0. 01, ∗∗ p < 0. 05, ∗ p < 0.
Source: Authors’ estimate.
147
148 South-South and North-South Trade
Notes
1. These countries are Burkina Faso, Burundi, Ghana, Senegal, Rwanda, Sierra
Leone, South Africa and Tanzania.
2. Furthermore, trade openness is sometimes marked by the dismantling of
non-tariff barriers which improves economic efficiency and corrects the nat-
ural inequalities in resource endowment of states. Indeed, the availability
of imported inputs is essential to boost industrial production in developing
economies (Lee, 1993). Thus, trade restrictions should reduce industrial pro-
duction because they prevent domestic industries from importing some inputs
at lower cost (Lee, 1995). In addition, through trade openness, countries cer-
tainly import inputs such as machinery, but they mainly increase the market
size of their domestic producers as well as returns to innovation (Harrison,
1996; Madsen, 2009).
3. However, some of these variables have been used by Rowthorn and Ken Coutts
(2004), Brady et al. (2011), Rowthorn and Ramaswamy (1999) to analyze the
industrialization process in Europe and Latin America.
4. To take into account possible cyclical effects, the practice in empirical lit-
erature is to work with data calculated as averages over a five-year period.
Thus, the study period 1971–2010 (40 years) is relatively long to work with
five-year averages. Moreover, Chang, Kaltani and Loayza (2009) have intro-
duced dummy variables for every five-year period to improve the quality of
estimates. The dummy variables for each five-year period will estimate the
effect of independent variables on the dependent variable while controlling
the impact of cyclical fluctuations.
5. In UNCTAD terminology (2008), the secondary sector consists of three cat-
egories of industries, extractive industries (mining, oil and businesses that
produce essential goods such as water and electricity), manufacturing indus-
tries and companies involved in construction (buildings and public works).
We will, on the one hand, identify the factors that explain the contribution
of the secondary sector (three types of industry) to GDP, and on the other,
identify only the factors of industrialization of manufacturing firms.
6. 50% of the sample countries derive over 12% of their GDP from exports of
raw materials. We have adopted this classification to avoid creating dummy
variables for certain raw materials. This approach enables combining all the
raw materials into a single variable and reason about various commodities
and not just one. The classification of raw materials was made by UNCTAD
(2009).
7. However, in its 2011 report, the AfDB (2011) acknowledges that the African
manufacturing sector is unable to catch up with the Asian manufacturing
industry. Indeed, in several African countries, the share of manufactur-
ing in GDP is around 10% or less. It is in South Africa, Cameroon, Côte
d’Ivoire, Egypt, Lesotho, Madagascar, Morocco, Mauritius, Namibia, Tunisia
and Zimbabwe that such industries are developed (between 15 and 20%
of GDP). According to the AfDB, the under-performance of manufacturing
industries is due to the lack of infrastructure, unreliable energy supply and
bureaucracy. It also recognizes that African manufacturing firms face fierce
competition from products imported from emerging countries such as China
(AfDB, 2011).
Henri Atangana Ondoa and Henri Ngoa Tabi 149
References
AfDB (2011), African Economic Outlook, AfDB/OECD.
Alderson, A. (1999), ‘Explaining Deindustrialization: Globalization, Failure, or
Success?’, American Sociological Review, vol. 64, no. 5, pp. 701–721.
Arellano, M. and Bond, S. (1991), ‘Some Tests of Specification for Panel Data:
Monte Carlo Evidence and an Application to Employment Equations’, Review
of Economic Studies, vol. 58, pp. 277–297.
Arellano, M. and Bover, O. (1995), ‘Another Look at the Instrumental-Variable
Estimation of Error-Components Models’, Journal of Econometrics, vol. 68,
pp. 29–52.
Barro, R. and Sala -I- Martin, X. (1995), Economic Growth, New York: McGraw-Hill.
Blundell, R. and Bond, S. (1998), ‘Initial Conditions and Moment Restric-
tions in Dynamic Panel Data Models’, Journal of Econometrics, vol. 87,
pp. 115–143.
Brady, D., Kaya, Y. and Gereffi, G. (2011), ‘Stagnating Industrial Employment in
Latin America’, Work and Occupations, vol. 38, no. 2, pp. 179–220.
Cardoso, F. and Faletto, E. (1979), Dependency and Development in Latin America,
Berkeley: University of California Press.
Cieslik, A. and Tarsalewska, M. (2011), ‘External Openness and Economic Growth
in Developing Countries’, Review of Development Economic, vol. 15, no. 4,
pp. 729–744.
Coe, D. and Helpman, E. (1995), ‘International R&D Spillovers’, European Eco-
nomic Review, vol. 39, pp. 859–887.
Clark, C. (1957), The Conditions of Economic Progress, London: Macmillan.
ECA (2013), Industrialization for Africa’s Economic Emergence, Abstract, Sixth Con-
ference of African Ministers of Finance, Planning and Economic Develop-
ment of the Economic Commission for Africa and the Conference of African
Ministers of Economy and Finance of the African Union, Abidjan.
Edwards, S. (1998), ‘Openness, Productivity and Growth: What do we Really
Know?’, The Economic Journal, vol. 108, no. 447, pp. 383–398.
Elbadawi, I. and Helleiner, G. (1998), ‘African Development in the Context of
New World Trade and Financial regimes: The Role of WTO and Its Relationship
to the World Bank and the IMF’, Presented at AERC Workshop on Africa and
the New World Trade System, Mombasa, Kenya, 27–29 April.
Frankel, J. and Rome, D. (1999), ‘Does Trade Cause Growth?’, American Economic
Review, vol. 89, no. 3, pp. 379–399.
Greenaway, D., Morgan, W. and Wright, P. (2002), ‘Trade Liberalisation and
Growth in Developing Countries’, Journal of Development Economics, vol. 67,
pp. 229–244.
Harrison, A. (1996), ‘Openness and Growth: A Time Series, Cross-Country Anal-
ysis for Developing Countries’, Journal of Development Economics, vol. 48,
pp. 419–447.
Hausmann, R. and Rodrik, D. (2006), Doomed to Choose: Industrial Policy as
Predicament,t Cambridge, MA: Harvard University, John F. Kennedy School of
Government.
Higino Schneider (2005), ‘International Trade, Economic Growth and Intellectual
Property Rights: A Panel Data Study of Developed and Developing Countries’,
Journal of Development Economics, vol. 78, pp. 529–547.
150 South-South and North-South Trade
Introduction
153
154 Impact of Monetary Unions on Trade
Until the 1980s, each member state of the Union was responsi-
ble for its macroeconomic policy and fiscal policy and all shared a
common monetary policy. This resulted in macroeconomic instabil-
ity, making the common monetary policy inefficient. Monetary policy
was also weakened by the economic and financial crisis experienced
by member states during the 1980s, after which it became necessary to
consolidate WAMU.
Considering the treaties had failed to achieve their objectives, given
the trend toward multiplication of regional blocs, and in a bid to revive
the integration process in West Africa through WAMU, the heads of state
of the Union in March 1990 instructed the then governor of the Central
Bank of West African States (BCEAO) to submit a succinct economic
integration plan. This was expected to return member states onto a path
of growth.
With support from the European Union (EU), France instigated the
establishment of WAEMU in June 1991. By 10 January 1994 in Dakar,
Senegal, following the devaluation of the CFAF, its establishment finally
became effective. The treaty to institute the Union entered into force on
1 August 1994 after ratification by the seven founding members: Benin,
Burkina-Faso, Côte d’Ivoire, Mali, Niger, Senegal and Togo. Three years
later, in May 1997, Guinea-Bissau joined the Union, becoming its eighth
member.
Among WAEMU’s objectives were to strengthen the economic and
financial competitiveness of member countries through an open and
competitive market and facilitate the free movement of people, goods,
services and capital and the establishment of a Common External Tariff
(CET) and a common trade policy through the creation of a common
market. Another was to harmonize the laws of member states, and
especially the tax system, to enable the common market to function
smoothly.
Reiterating their intention to create a common currency within
ECOWAS in Lomé in 1999, the leaders of member countries adopted
a declaration stipulating the establishment of a sub-regional monetary
Ibrahima Camara 155
1 Literature review
In recent years, there has been growing interest in monetary unions, and
especially their impact on trade. Although currency is often considered
156 Impact of Monetary Unions on Trade
Theoretical review
Several theories have been developed about monetary unions, the most
known being the Optimum Currency Area (OCA). An OCA refers to a
grouping of economic spaces that is subject to monetary sovereignty
and a single currency system, or characterized by the existence of a fixed
and irrevocable parity mechanism between their various currencies.
Exchange rates are fixed within the area, and left floating outside
the area.
The first theoretical OCA analyses were carried out by Mundell (1961)
who argues there are benefits for regions that use a common currency.
Monetary unions facilitate international trade, as a single means of
trade reduces the cost of transactions. Yet he notes that a common cur-
rency may be the source of problems in cases of asymmetric price and
salary shocks, and minimal rigidities. Mundell concludes the optimum
currency area is one within which the labor factor is mobile.
Highlighting the size of economies, McKinnon (1963) asserts that
small economies are best suited to monetary unions because they tend
to be much more open, and consequently, less subject to nominal
rigidities. When the share of international trade in economic activity
increases, the opportunity cost of exchange rate rigidity drops and its
benefit increases. He also argues that when a small country trades with
a big partner, it is unable to conduct a completely monotonic monetary
policy. So it is in the interest of small open countries to form monetary
unions which will ensure the stability of their trade without jeopar-
dizing the independence of their already limited monetary policy. For
his part, Kenen (1969) argues that broader diversification of economies
leads to less asymmetric shocks. So countries with broadly diversified
economies stand to gain nothing by conducting a national monetary
policy.
A currency facilitates trade, as a unit of account and means of trade.
Inasmuch as a monetary union may influence the reduction of exchange
rate and transaction cost uncertainties, it may reinforce trade flows
within the currency area. For many authors, the theoretical effects of
international trade are well identified and less controversial; but under-
standing the impacts of monetary unions on trade from the theoretical
viewpoint seems a complicated task. Still, this may be more feasible from
an empirical standpoint.
Ibrahima Camara 157
Empirical review
The impact of monetary unions on trade has been the topic of many
empirical studies in recent decades. It is generally admitted that a com-
mon currency strengthens trade, but the first empirical works were more
interested in the impact of exchange rate volatility on trade flows, with-
out arriving at conclusive findings (Lochard, 2005). The first and most
famous empirical study on the impact of monetary unions on trade
flows was conducted by Rose (2000).
Rose (2000) used a bilateral trade gravity model into which he intro-
duced the monetary union represented by a binary variable. The variable
was given the value 1 or 0 for two countries, depending on whether
they belonged to the same monetary union or not. By introducing
other control variables such as sharing a common border, a common
language, colonial links or belonging to the same free trade zone, he
found that two countries sharing a common currency carry out three
times more trade than similar countries with different currencies. Rose’s
(2000) paper attracted much criticism concerning the magnitude of the
impact of the common currency. Most criticism dealt with the existence
of a bias in the results, particularly a self-selection bias (Persson, 2001;
Tenreyro, 2001), an aggregation bias (Nitsch, 2004) and an omitted vari-
ables bias due to a correlation between the independent variables and
the error term.
According to Persson (2001), the results of Rose’s (2000) paper are
biased due to the existence of self-selection.3 But there could be a
correlation between the monetary union and trade determinants. For
example, small or poor countries may trade more between themselves
not because they have a common currency, but because they share
common geographical and cultural features. The existence of such a cor-
relation does not permit the comparison of countries having a common
currency with the sample countries. Therefore, it cannot be concluded
that monetary unions increase trade, everything being equal.
To remedy self-selection bias, Persson (2001) used a matching method.
First, he looked for the determinants of monetary unions and calculated
a monetary union propensity score for each pair of countries.4 Then, for
each of the pairs, he constituted a control group comprising countries
with their own currencies, but which have the features of the monetary
union pairs of countries according to the propensity score. Last, he esti-
mated the effects of the monetary union on trade using only pairs of
monetary union countries and their control groups. By so doing, he
found the impact of monetary unions on trade is low, ranging from 13%
to 65%. Tenreyro (2001) also tried to correct the self-selection bias in
158 Impact of Monetary Unions on Trade
the method used by Rose (2000). She first estimated the probability that
two countries may adopt a common currency and checked the impacts
of monetary unions on trade after explaining self-selection. She found
that monetary unions increased trade by 60%, though such impact was
not significantly different from zero.
According to Micco et al. (2002), one of the problems in the study
conducted by Rose (2000) is that most of the sample monetary unions
are made up of very small or poor countries sharing a common currency
(CFAF zone countries, for example) or of very small or poor countries
which have adopted the currency of big countries (like Réunion and
France). These findings can hardly be extended to bigger countries.
Lochard (2005) argues that introducing dummy variables which help to
isolate each type of monetary union may reduce the aggregation bias.
Using this method, she obtained a 238% trade increase. She concluded
the high impact obtained by Rose was due to the weights of unilateral
unions in her sample.
In short, most of the studies resulting from the findings of Rose (2000)
adopted other methodologies which revealed lesser impacts. Still, they
are unanimous on the fact that monetary unions increase trade.
Sub-Saharan Africa has been the subject of very many studies, though
most of them focusing on regional trade agreements rather than mone-
tary unions. Carrere (2004) assessed the impact of regional agreements
on trade in sub-Saharan Africa over the period 1962–1996 using a panel
gravity model. The findings showed the implementation of African
regional trade agreements led to a significant increase in trade between
members. She also found the monetary unions in force in the CFAF zone
strongly enhanced this positive impact.
Avom and Gbetnkom (2005) looked for intra-WAEMU trade deter-
minants using a gravity model and laying emphasis on economic
reforms undertaken from 1980–1990. Splitting their study period into
two, 1990–1994 and 1996–2000, they found regional integration sub-
stantially increases trade between the Union’s member states. Agbodji
(2007) attempted to isolate the impacts of monetary and economic
unions on intra-WAEMU trade. Using an augmented dynamic gravity
model, he showed that WAEMU membership and the implementation
of economic reforms produced significant effects in terms of diversion
of imports and exports; although economic distortions, considered to
encourage illegal trade, significantly reduced bilateral trade within the
Union.
By controlling the impact of geographical and structural factors, Diop
(2007) showed how the existence of a monetary union affects the level
of intra-community trade in West Africa. Using a gravity model with
Ibrahima Camara 159
data on bilateral trade between ECOWAS member states over the period
1997–2004, he found that geographical and structural factors as well
as WAEMU membership determine the intensity of bilateral trade flows
in West Africa. Further, use of a common currency leads to considerable
trade development. He concluded that, all things being equal, extending
WAEMU to the other ECOWAS countries should considerably increase
the volume of intra-regional trade. However, the study does not take
into account the period preceding the establishment of WAEMU to allow
for comparison.
15,000,000
10,000,000
5,000,000
–5,000,000
–10,000,000
–15,000,000
–20,000,000
in
re
er
al
al
ga
s
sa
t
en
oi
ig
M
Fa
To
To
ne
is
Iv
N
B
B
na
Se
d'
a-
ki
ne
ôt
ur
ui
B
Figure 8.1 Trade balance 2000–2009 for WAEMU’s countries (CFAF millions)
Source: Central Bank of West African States (CBWAS), author’s calculations.
Gravity model
During the last decades, many authors have used the gravity model to
assess the impact of customs or monetary unions on bilateral trade.
Tinbergen (1962) used the model to explain the intensity of migration
flows based on the size of regions and distance between them. Since
Rose (2000), the gravity model has been increasingly used to assess the
impact of monetary unions on trade.
Theoretical model
Empirical studies on international trade increasingly rely on gravity
models. Inspired by the Newtonian theory, this type of model expresses
trade flows between two countries as proportional to their economic
weight and inversely proportional to the geographical distance between
them. In its simplest form, the model is presented as follows:
β β
Mij = A × (Y
COM Yi Yj )j 1 × Dij2 (1)
Here COM Mij , represents the value of trade flows between country i and
country j, Y, their national income, Dij the distance between these
countries and A, a proportionality coefficient. In addition to these two
traditional variables (GDP and distance), the model has been augmented
by other variables to take into account some specificities of bilateral
Ibrahima Camara 161
relations. They include sharing a common land border, the fact that the
country is an island (landlocked) or not, having the same former colo-
nial rulers and GDP per capita. These variables are often called control
variables.
To isolate the effects of monetary unions on trade since Rose (2000),
some authors introduce a dummy variable with the value 1 for two
member countries of a monetary union and the value 0 if not. In addi-
tion to the dummy variable, control variables are represented by a vector
Xij . So, the augmented model becomes:
β β β
Mij = A × (Y
COM Yi Yj )j 1 × Dij2 × Xij2 (2)
Mij ) = β0 + β1 × ln(Y
ln(COM Yi Yj ) + β2 × ln(D
( ij ) + β3 × ln(X
( ij ) (3)
Empirical specification
Many empirical gravity model equations were derived from the theo-
retical model equation (3). Most of the empirical studies were inspired
by Rose (2000), who used the augmented gravity model to estimate the
effects of monetary unions and exchange rate volatility on trade. He
added many variables in this augmented model which take into account
monetary aspects, among others. The model used is presented as follows:
( ijt ) = β0 + β1 ln(Y
ln(X Yi Yj )t + β2 ln(yi yj )t + β3 ln(D
( ij ) + β4 Conttij + β5 Langgij
+ β6 FTAijt + β7 ComNattij + β8 ComColij + β9 Colonyij + γ CU
Uijt
+ δV(eij )t + εijt (4)
Here i and j represent the two countries, t expresses time and other
variables are defined as:
( ij ) = β0 + β1 × ln(Y
ln(X Yi Yj ) + β2 × ln(yi yj ) + β3 × ln(Supi Sup
pj ) + β4+k × Dk
k=1
(5)
Here X is bilateral trade, Y real GDP, y GDP per capita, Sup surface
area and D the vector of dummy variables. For the purposes of proper
presentation, the time index t was omitted.
While most of the empirical studies that used the gravity model
focus on the overall impact of monetary or customs unions for the
entire study period, Micco et al. (2002) assessed year-on-year impact of
the European Monetary Union on trade. To that end, they conducted
a cross-analysis by estimating the year-on-year gravity equation using
ordinary least squares. The model estimated year-on-year is as follows:
Ibrahima Camara 163
( ij ) = β0 + β1 ln(Y
ln(X Yi Yj ) + β2 ln(Y
Yi Yj ) + β3 ln(D
( ij ) + β4 Conttij + β5 Langgij
+ β6 FTAijt + β7 LandLockij + β8 Islandij + β9 Surffij
+ β10 EMU
Uij + εij (6)
Here the variables, save for FEMU and EU, are described as above. FEMU Uijt
is a dummy variable with value 1 where countries i and j formally belong
to the European Monetary Union during year t and EU Uijt is a dummy
variable equal to 1 where the pair of countries belongs to the EU during
year t. In this equation, however, the other trade determinants – like
the colonization, border or common language variables – are not direct
interest values and do not vary, or vary much, over time. In fact, most of
these determinants have already been taken into account in the bilateral
specific impact (βij ) and might lead to a bias if they are not well specified
(Pakko and Wall, 2001).5 Lochard (2005) also used a fixed-impact panel
model which did not include these dummy variables.
To check the robustness of the findings of the estimation of equa-
tion (7), they estimated another panel model which, this time, included
control variables.
( ijt ) = β0 + β1 ln(Y
ln(X Yit Yjt ) + β2 ln(yit yjt ) + β3 ln(D
( ij ) + β4 Conttij + β5 Langgij
+ β6 FTAijt + β7 LandLockij + β8 Islandij + β9 Surffij + β10 FEMU
Uijt
+ β10 EU
Uijt + εijt (8)
bilateral trade over the years. Before joining a monetary union, coun-
tries first meet a number of criteria some of which help to strengthen
trade between them. Besides, countries do not all join a monetary union
at the same time.
For the purposes of this study, a year-on-year cross-analysis is con-
ducted to better capture and monitor WAEMU’s impact on trade flows
over time. To that end, the equation to be estimated will be similar
to equation (6). The choice of this model seems appropriate given the
events within the WAEMU area, which experienced several changes dur-
ing the last decades. The early 1990s was marked by the implementation
of Structural Adjustment Programmes (SAPs) by countries of the sub-
region, which probably modified the structure of their trade. Then there
was the devaluation of the CFAF in 1994, Guinea-Bissau’s late join-
ing of the Union in 1994 and the coup d’état and rebel uprising in
Côte d’Ivoire6 in 1999 and 2002, respectively. All these events probably
modified the intra-WAEMU trade structure. The model to be estimated
therefore is:
( ij ) = β0 + β1 ln(Y
ln(X Yi Yj ) + β2 ln(Dist
( tij ) + β3 Borderrij + β4 Langgij + β5 Colonyij
+ β6 LandLockij + β7 ln(Supi Sup
pj ) + β8 WAEMU
Uij
+ β9 WAEMU1ij + εij (9)
Here the variables are defined as follows: Xij , is the export flows of coun-
try i to country j; Supi and Sup pj are respectively the surface areas of
countries i and j; Colonyij , is a dummy variable equal to 1 where the two
countries had the same colonial ruler and 0 if not; Border, is equal to 1
where the two countries share a common border and 0 if not, WAEMU,
is a dummy variable equal to 1 where the two countries belong to the
WAEMU zone and 0 if not7 ; WAEMU1, is a dummy variable with value
1 where the partner country is a WAEMU member state and 0 if not.
The coefficient of the WAEMU1 variable helps to capture effects on the
diversion of trade flows while the WAEMU variable helps to identify the
trade surplus created by the establishment of the Union. The expected
signs are therefore positive for the WAEMU variable and negative for the
WAEMU1 variable.
Owing to the existence of zero export values for pairs of countries
at a given date, there is a modeling problem as exports are recorded
in logarithm. To solve this problem, we proceed like Bangake and
Eggoh (2008) did. Inspired by Rose (2000), they added the value 100
to each bilateral trade value. In this study, therefore, the export value
Ibrahima Camara 165
This section presents data and empirical findings as well as related com-
mentaries. The study sample comprises all ECOWAS member countries
and covers the period 1990–2005. Data on exports (in 1987 constant
USD million) is taken from DOTS. GDP is provided by the World Bank’s
‘World Development Indicators & Global Development Finance, 2012’.
It is evaluated in 1987 constant USD million. Data on distance is culled
from the Centre d’Etudes Prospectives et d’Informations Internationales
(CEPII).
Empirical findings
Ordinary least squares were used for the year-on-year estimation by
equation (9). The results of the estimation are presented in Figure 8.2.
Only the coefficients of the WAEMU and WAEMU1 variables for each of
the cross-cutting regressions have been entered, and the overall findings
are presented in Table 8.1.
The GDPs of exporting and importing countries, their surface areas,
the distance between them and the number of landlocked countries in
the pairs of countries are significant and show expected signs for almost
all the years. As such, large countries with high GDPs tend to carry out
much more trade while the more they are landlocked and distant from
35.0%
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
each other the less trade they tend to conduct. The WAEMU1 variable,
for its part, is not significant for each of the study years, but has the
expected negative sign (with the exception of 1990). It is therefore diffi-
cult to conclude that there is an effect of diversion in trade flows to the
Union each year.
The WAEMU variable coefficient is significant and positive for each
of the years, meaning WAEMU has a significant and positive impact on
export flows. However, there seems to be various phases in the Union’s
marginal effects trend on exports (Figure 8.2): the first phase from 1990
to 1996 was characterized by a reduction in coefficients with a peak in
1995, the second phase covering 1997 and 1998 was characterized by
growth and the third fluctuation phase extended from 1999 to 2005.
On the whole, these effects recorded a decline.
In 1990, intra-WAEMU trade was 28.3% (exponential [0.2494]−1)
higher than trade between countries with different currencies.8 This
marginal effect stood at 23% in 1994 and 22% in 1996. The decline
in this effect can be explained, among other things, by the lack of trade
and production policy harmonization and coordination. It can also be
explained by poor infrastructure and similarity in production structures,
most of which are based on agro-industry. These findings show that the
Ibrahima Camara 167
40.0%
30.0%
20.0%
10.0%
0.0%
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Toward Africa WAEMU
4 Conclusion
Coefficient LPIB LSUP Ldistance Landlock WAEMU WAEMU1 Border Comland Comcol Constance
T-Stat
1990 0.177 −0.018 −0.153 −0.091 0.249 0.009 0.123 0.269 −0.179 −3.997
(5.82)∗∗∗ (−0.6) (−2.95)∗∗∗ (−1.27) (2.46)∗∗ (0.11) (1.36) (1.94)∗ (−1.15) (−5.2)∗∗∗
0.110 0.037 −0.078 −0.210 0.243 −0.013 0.236 0.074 0.050 −2.958
1991 (3.37)∗∗∗ (1.31) (−1.48) (−2.81)∗∗∗ (2.28)∗∗ (−0.17) (2.46)∗∗ (0.47) (0.3) (−3.4)∗∗∗
1992 0.134 0.032 −0.095 −0.156 0.234 −0.036 0.167 0.202 −0.115 −3.630
(5.28)∗∗∗ (1.28) (−1.72)∗ (−2.08)∗∗ (2.57)∗∗ (−0.52) (1.85)∗ (1.26) (−0.66) (−4.77)∗∗∗
1993 0.105 0.063 −0.075 −0.193 0.221 −0.018 0.148 0.219 −0.079 −3.293
(4.33)∗∗∗ (2.76)∗∗∗ (−1.36) (−2.62)∗∗∗ (2.44)∗∗ (−0.26) (1.63) (1.36) (−0.45) (−4.22)∗∗∗
1994 0.112 0.060 −0.115 −0.175 0.207 −0.007 0.086 0.187 −0.090 −3.113
(4.59)∗∗∗ (2.63)∗∗∗ (−2.12)∗∗ (−2.4)∗∗ (2.29)∗∗ (−0.1) (0.97) (1.17) (−0.52) (−4.04)∗∗∗
1995 0.126 0.046 −0.135 −0.126 0.214 −0.034 0.062 0.229 −0.151 −3.278
(5.55)∗∗∗ (2.1)∗∗ (−2.54)∗∗ (−1.79)∗ (2.44)∗∗ (−0.51) (0.71) (1.47) (−0.89) (−4.5)∗∗∗
1996 0.126 0.050 −0.160 −0.104 0.199 −0.067 0.084 0.243 −0.119 −3.240
(5.77)∗∗∗ (2.34)∗∗ (−3.13)∗∗∗ (−1.51) (2.36)∗∗ (−1.03) (1) (1.63) (−0.73) (−4.66)∗∗∗
1997 0.143 0.049 −0.133 −0.106 0.236 −0.075 0.051 0.179 −0.085 −4.073
(6.07)∗∗∗ (2.22)∗∗ (−2.67)∗∗∗ (−1.54) (2.87)∗∗∗ (−1.19) (0.62) (1.25) (−0.55) (−5.73)∗∗∗
1998 0.153 0.042 −0.133 −0.097 0.247 −0.103 0.061 0.151 −0.099 −4.290
(6.79)∗∗∗ (1.98)∗∗ (−2.7)∗∗∗ (−1.46) (3.05)∗∗∗ (−1.65) (0.75) (1.08) (−0.65) (−6.25)∗∗∗
1999 0.155 0.058 −0.194 −0.152 0.197 −0.035 −0.025 0.181 −0.090 −4.345
(6.89)∗∗∗ (2.72)∗∗∗ (−3.97)∗∗∗ (−2.32)∗∗ (2.46)∗∗ (−0.57) (−0.31) (1.31) (−0.61) (−6.36)∗∗∗
2000 0.127 0.069 −0.183 −0.129 0.166 −0.073 −0.026 0.074 0.045 −3.427
(5.2)∗∗∗ (2.97)∗∗∗ (−3.53)∗∗∗ (−1.84)∗ (1.95)∗ (−1.12) (−0.31) (0.51) (0.29) (−4.78)∗∗∗
2001 0.122 0.046 −0.186 −0.121 0.262 −0.077 0.099 −0.025 0.114 −2.739
(4.32)∗∗∗ (1.82)∗ (−3.64)∗∗∗ (−1.69)∗ (2.72)∗∗∗ (−1.05) (1.19) (−0.14) (0.59) (−3.36)∗∗∗
169
170
Coefficient LPIB LSUP Ldistance Landlock WAEMU WAEMU1 Border Comland Comcol Constance
T-Stat
2002 0.136 0.040 −0.195 −0.147 0.222 −0.037 0.084 −0.011 0.111 −3.153
(5.54)∗∗∗ (1.65)∗ (−3.6)∗∗∗ (−2.05)∗∗ (2.53)∗∗ (−0.55) (0.96) (−0.07) (0.69) (−4.43)∗∗∗
2003 0.107 0.062 −0.153 −0.156 0.181 −0.055 0.081 0.016 0.138 −2.739
(4.37)∗∗∗ (2.56)∗∗ (−2.82)∗∗∗ (−2.18)∗∗ (2.07)∗∗ (−0.81) (0.93) (0.11) (0.84) (−3.81)∗∗∗
2004 0.098 0.079 −0.155 −0.179 0.245 −0.107 0.088 −0.031 0.133 −2.722
(4.09)∗∗∗ (3.27)∗∗∗ (−2.84)∗∗∗ (−2.48)∗∗ (2.77)∗∗∗ (−1.56) (1) (−0.2) (0.81) (−3.82)∗∗∗
2005 0.104 0.075 −0.161 −0.174 0.161 −0.065 0.098 −0.008 0.167 −2.846
(4.46)∗∗∗ (3.09)∗∗∗ (−2.94)∗∗∗ (−2.41)∗∗ (1.82)∗ (−0.94) (1.11) (−0.05) (1.02) (−4.07) ∗∗∗
Notes: This table shows the results of the estimations based on a sample of ECOWAS countries; 1990–2005 Period. Dependent Variable: Logarithm of
Exports (LExport).
T-stat are presented in brackets. ∗∗∗ 1% significant; ∗∗ 5% significant; ∗ 10% significant.
Source: Author’s estimates.
Ibrahima Camara 171
Notes
1. These regional economic communities are: the Arab Monetary Union (UMA),
the Economic Community of West African States (ECOWAS), the Common
Market for Eastern and Southern Africa (COMESA), the Economic Commu-
nity of Central African States (ECCAS) and the Southern African Development
Community (SADC).
2. This date was set in 1987 during the establishment of the Monetary Coopera-
tion Programme (MCP).
3. The probability of two countries adopting the same currency is non-random
and could depend on some independent variables.
4. This score evaluates the probability of belonging to monetary unions accord-
ing to other trade determinants.
5. Quoted by Lochard (2005).
6. The specific case of Côte d’Ivoire was considered owing to its economic and
trade impact within the Union.
7. The WAEMU variable had the value 1 for two Union member countries even
before the creation of the Union, the goal of the model being to monitor the
coefficient of this variable over time.
8. WAEMU’s effects on trade were calculated taking into account the exponential
of coefficients as expressed in logarithm in the model.
References
Agbodji, A. E. (2007), ‘Intégration et échanges commerciaux intra sous-régionaux:
Le Cas de l’UEMOA’, Revue Africaine de l’Intégration, vol. 1, no. 1, pp. 161–188.
Avom, D. and Gbetnkom, D. (2005), ‘Intégration par le marché: le cas de
l’UEMOA’, Région et Développement no. 22–2005.
Bangake, C. and Eggoh, J. (2008), ‘The Impact of Currency Unions on Trade:
Lessons from CFA Franc Zone and Implications for Proposed African Mone-
tary Unions’, Document de Recherche no. 2008–29, Laboratoire d’Economie
d’Orléans.
Carrere, C. (2004), ‘African Regional Agreements: Impact on Trade with or with-
out Currency Unions’, Journal of African Economies, vol. 13, no. 2, pp. 199–239.
De Monchy, G., Nancy, G., Aloy, M. and De Boisdeffre, L. (1995), ‘Effets et
perspectives de la dévaluation au Sénégal’, DIAL 1995-01/T.
Diop, C. A. (2007), ‘L’UEMOA et la perspective d’une zone monétaire unique de
la CEDEAO: les enseignements d’un modèle de gravité’, Document d’Etude et
de Recherche no DER/07/01 – BCEAO.
Kenen, P. (1969), ‘The Theory of Optimum Currency Areas: An Eclectic View’,
in Mundell and Swoboda (eds.), Monetary Problems of the International Economy,
Chicago: University Press, pp. 41–60.
Lochard, J. (2005), ‘Mesurer l’influence des unions monétaires sur le commerce’,
Economie internationale, vol. 2005/3, no. 103, pp. 5–24.
Masson, P. and Patillo, C. (2004), ‘A Single Currency for Africa?’, Finance and
Development,t vol. 41, no. 4, pp. 9–15.
Masson, P. and Patillo, C. (2005), The Monetary Geography of Africa, Washington,
DC: Brookings Institution Press.
172 Impact of Monetary Unions on Trade
Introduction
Most WTO members are also party to one or more regional trade
agreements. Since the early 1990s, the sharp increase in the num-
ber of Regional Trade Agreements has continued. Some 250 Regional
Trade Agreements (RTAs) have been notified to the GATT/WTO up to
December 2002, of which 130 were notified after January 1995. Over
170 RTAs are currently in force; an additional 70 are estimated to be
operational although not yet notified.
(WTO, 2003)
173
174 Trade Agreements and Flows in ECOWAS
However, some studies show that RTAs in Africa have not led to
any increase in trade (Elbadawi, 1997; Yeats, 1999; Diouf, 2002). Diouf
(2002) attributes the lack of success of African Regional Agreements
to the adoption of an integration process that is market-based to the
detriment of transport and communication channels. Sylla (2003) and
others suggest the failure of integration in Africa is due to governments
that jealously protect national sovereignty even when states are barely
viable in the context of globalization. Similarly, the works of Gunning
(2001), ECA (2004), Yang and Gupta (2005) and Chacha (2008) sup-
port the assertion that RTAs in Sub-Saharan Africa have not improved
trade between member countries, because products are not comple-
mentary, insufficient infrastructure, minimal product differentiation,
high costs discouraging imports, narrow markets and a lack of political
commitment.
In contrast, Deme (1995), Cernat (2003), Carrère (2004), Coulibaly
(2007) and Afersorgbor and Bergeijk (2011) find that Sub-Saharan
African RTAs have increased trade flows between member countries.
In a recent study by Afersorgbor and Bergeijk (2011), a gravity model
was applied to 35 countries to analyze the impact of ECOWAS and The
Southern Africa Development Community (SADC)1 trade agreements
on bilateral trade. The impact is considered relatively more significant
than the impact from European Union-African Caribbean and Pacific
States agreements (EU-ACP agreements). Definitively, the contribution
of RTAs to regional and sub-regional integration in Africa remains con-
troversial. Similarly, different arguments are provided to explain these
divergent results. Thus, many ECOWAS actors believe that monetary
integration might contribute to the effectiveness of RTAs. Recently,
Interbank Electronic Banking Group of the West African Economic and
Monetary Union (GIM-WAEMU)2 type interbank payment project was
initiated within ECOWAS by the World Bank and GIM-WAEMU, who
advised that use of the CFA franc, Cedi or Naira should no longer be
an obstacle to trade development in the ECOWAS zone. Removing the
currency constraint in cross-border payments within ECOWAS would
be a key driver of integration. So do trade agreements associated with
the sharing of a single currency have a significant impact on bilat-
eral trade within ECOWAS? In the case of CEMAC,3 Avom (2005)
observes that economic and monetary union had a relatively no sig-
nificant impact on the development of internal trade. Whereas Debrun
et al. (2002), Masson and Patillo (2005) and Bénassy-Quéré and Coupet
(2005) point to the sub-optimal nature of the ECOWAS monetary union
project.
Benjamin Ndong and Sokhana Diarra Mboup 175
1,600
1,400
1,200
1,000
800
600
400
200
0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
World North America Central South America
Europe CEI Africa
Middle East Asia
2,500
2,000
1,500
1,000
500
0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
World North America Central South America
Europe CEI Africa
Middle East Asia
AFRICA
AGRIC. PROD 59.49 3.5 2.04 24.82 1.19 12.02 4.81 10.55
F. EXTRAC I. P 382.21 86.92 14.65 127.34 0.37 26.84 3.48 115.24
MANUF, P. 110.3 10.6 2.68 48.29 0.25 28.18 5.86 13.68
TOT. EXPORT. 594.24 101.64 19.45 205.21 1.85 77.03 21.34 145.84
INTER. TRADE 1% 17% 3% 35% 0% 13% 4% 25%
AFRICA
TOT. IMPORT. 538.08 37.47 21.35 199.39 12.49 77.03 37.87 152.48
INTER. TRADE 1% 7% 4% 37% 2% 14% 7% 28%
Export effort
Like most African countries, ECOWAS countries specialize in a limited
number of commodities. Nigeria’s exports mainly comprise oil products
(94%). Many countries export cotton, but most other products are spe-
cific to one or two countries (see Table 9.A.4 in Annex and Table 9.4).
The average export effort for the least developed countries (LDCS) in
ECOWAS was 25.72% in 2010, with countries such as Togo (37%) and
Guinea (32%) showing exports below the global level. The export efforts
of LDCs in ECOWAS are significant. The weakness of intra-community
trade, Africa’s low share in general, and ECOWAS’s in particular, is
undoubtedly due to specialization in a limited number of commodities
and a failure to diversify economies. Table 9.A.4 (in Annex) shows how
LDCs in ECOWAS specialize in commodities (oil, extractive and agricul-
tural products) and how their exports fluctuated on average by 19.2%
during 2005–2011, far above the global level of 12%.
Dependence on primary products with particularly unstable prices
leads to structural instability and high terms of trade volatility.
Table 9.2 Intra-African trade by economic zone in 2011, percentages of total exports
AMU CEMAC COMESA ECCAS ECOWAS FRANC ZONE SADC WAEMU AFRICA WORLD
−−−−−−−−−−−−→ Exports to
AMU 3.6 0.2 1.7 0.3 0.8 0.7 0.1 0.5 5.6 100.0
CEMAC 0.2 1.0 0.2 1.3 0.9 1.4 0.4 0.3 2.7 100.0
COMESA 2.8 0.2 8.9 2.1 0.5 0.4 7.6 0.2 16.1 100.0
ECCAS 0.1 0.4 1.8 0.6 0.4 0.6 3.0 0.1 4.1 100.0
ECOWAS 0.3 1.3 0.1 1.4 6.3 4.7 3.0 3.4 10.9 100.0
FRANC ZONE 0.4 1.3 0.3 1.6 7.5 5.1 1.6 3.8 10.9 100.0
SADC 0.3 0.1 6.5 1.7 1.1 0.4 9.9 0.3 12.5 100.0
WAEMU 0.9 2.0 0.4 2.5 23.2 14.1 4.3 12.1 30.5 100.0
AFRICA 1.4 0.5 3.5 1.3 2.3 1.6 4.6 1.1 10.1 100.0
WORLD 0.7 0.1 0.8 0.2 0.7 0.3 0.9 0.2 2.9 100.0
Note: AMU: Arab Maghreb Union/CEMAC: Central African Economic and Monetary Union/COMESA: Common Market of Eastern and Southern
Africa/ECCAS: Economic Community of Central African States/SADC: Southern Africa Development Community.
Source: African Statistical Yearbook, 2013.
179
180 Trade Agreements and Flows in ECOWAS
Table 9.4 ECOWAS’ Least Developed Countries (LDC), export effort and ratio of
exports to GDP
Benin 6558 18 25 13 20 4 5
Burkina Faso 8825 10 21 9 18 1 3
The Gambia 1050 29 24 16 16 13 8
Guinea 4736 30 32 29 31 1 1
Guinea Bissau 835 16 20 16 15 1 5
Liberia 988 39 28 24 24 15 4
Mali 9422 26 26 21 22 5 4
Niger 5411 17 24 14 21 2 2
Senegal 12855 26 24 18 17 8 7
Sierra Leone 1910 21 22 15 19 6 3
Togo 3176 37 40 30 31 7 9
World 63134700 28 30 23 24 5 6
Specifications
The gravity model has many applications in analysis of international
trade. Since Tinbergen’s (1962) pioneering work, Newton’s gravitational
equation has been adapted to international trade flows. Newton’s theory
of gravity holds that the gravitational force, or attraction, between two
objects (planets for example) is proportional to their respective masses
and inversely proportional to the distance between them.
M i × Mj
Aij = (1)
Dij
Yi × Yj )
(Y
Xij = α0 (2)
Dij
182 Trade Agreements and Flows in ECOWAS
Here Xij represents the value of bilateral trade between country i and
country j; Y is GDP; and Dij measures the distance between country i
and country j.
Expressed in logarithmic form, the equation (3) becomes:
ln Xij = α0 + α1 ln ((Dij ) + α2 ln (Y
Yi × Yj ) (4)
ln Xij = α0 + α1 ln ((Dij ) + α2 ln (Y
Yi × Yj ) + α3 Cij (5)
Here α3 is the vector of the coefficients associated with the vector of the
control variables Cij . The control variables are added because bilateral
trade is not only explained by the economic size of partners and the
distance between them, but also by other variables, such as historical ties
(language, former colonizer) and geographical characteristics (common
border, landlockedness).
In this analysis, a dummy variable is introduced and expressed as
follows: trade agreements plus monetary union (RTA&MUij ) takes the
value 1 where two countries share both a trade bloc and monetary
union, otherwise it is 0. This variable allows us to take into account our
problem and provide an innovating factor in comparison with the pre-
vious studies where the two effects were separated in the same equation.
Initially, a cross-sectional analysis is selected where the gravity equa-
tion is estimated using the Ordinary Least Squares (OLS) method. The
OLS method is considered appropriate, when most of the variables are
time-bound (Micco and Ordônez, 2002; Diop, 2007).
ln ((Xij ) = α0 + α1 ln (Y
Yi × Yj ) + α2 ln (yi × yj ) + α3 ln Dij + α4 Comlanggij +
α5 FTRij + α6 LANDLOCKij + α7 ISLANDij + α8 Comcolij
+ α9 RTA&MU
Uij + εij (6)
Benjamin Ndong and Sokhana Diarra Mboup 183
ln ((Xij )t = α0 + α1 ln (Y
Yi ∗ Yj )t + α2 ln (yi ∗ yj )t + α3 RTA&MU
Uij + αij + (εij )t (7)
Here αij is individual effect, and εij is residual effect. However, in a second
specification, the dummy variables are introduced to test the robustness
of the previous model (equation 7).
ln ((Xij )t = α0 + α1 ln (Y
Yi × Yj )t + α2 (yi × yj )t + α3 ln Dij + α4 Comlanggij
+ α5 FTRij + α6 LANDLOCKij + α7 ISLANDij + α8 Comcolij
+ α9 RTA&MU
Uij + αij + (εij )t (8)
It should be noted that the variables of the different models are classified
into two categories. The first concerns continuous variables, expressed
in logarithms, which is the case of the product of GDPs, the prod-
uct of GDPs per capita and the relative distance. This form allows
184 Trade Agreements and Flows in ECOWAS
• Product of GDPs and GDPs per capita are indicators of the poten-
tial size of the market and of purchasing power, respectively. The
higher these products, the greater the opportunities for trade for both
partners. The sign of the associated coefficients should be positive.
• Relative distance allows transaction costs to be calculated approx-
imately, especially, the cost of transport for trade. Distance is a
negative factor on trade, as it generates transport costs that are
particularly high when distance between the two partners is greater.
• Landlockedness variable makes it possible to estimate what effect the
lack of a seaboard has on trade for the countries concerned. Transport
statistics show that 80% of trade is carried out by ship. A country
which is an island is thus considered to be handicapped in terms of
trade.
• Studies have shown that a common colonial past shared between
partners influences bilateral trade. Thus, common former colonial
ruler and common language variables should impact bilateral trade
positively.
• The RTA&MU variable makes it possible to calculate the assumed pos-
itive effect of a monetary union coupled with a trade agreement.
In theory, trade agreements and shared currencies foster bilateral
trade between member countries (Table 9.5).
Note: (∗ ) significant at 1%, (∗∗ ) significant at 5%, (∗∗∗ ) significant at 10%. Student’s t statistics
are reported in brackets.
and present the expected signs except for the common colonial ruler
variable. On average, the value of bilateral trade rises in line with the
size of the economies and falls with distance. The existence of a com-
mon border fosters trade; whereas the landlocked or insular position of
a partner country or two partner countries does not foster trade. Overall,
geographical and structural factors have a determining impact on intra-
regional trade. The GDP and GDP per capita variables have a positive
effect on bilateral trade, so their coefficients are positive. Geographical
characteristics such as relative distance, landlockedness and islandness
impede bilateral trade, although a common border, impacts positively
on bilateral trade.
For the cross-sectional and panel data, the coefficient associated with
the colonial ruler variable is negative, even though it is not statistically
significant. The cross-sectional data results are confirmed by the panel
data with the exception of the sign of the coefficient of the product
of GDP per capita variable which is not statistically significant in the
Benjamin Ndong and Sokhana Diarra Mboup 187
Note: (∗ ) significant at 1%, (∗∗ ) significant at 5%, (∗∗∗ ) significant at 10%. Student t statistics
are reported in brackets.
equation (8), which was used to control the results of equation (7).
Stability is noted among the different relationships described in their
dual time and spatial or static and dynamic dimensions.
The coefficients associated with the product of GDPs, product of GDP
per capita and relative distance variables are elasticities. When asso-
ciated with the product of GDPs variable, for example, these indicate
that, all other things being equal, a 1% rise in the product of GDPs is
reflected in a 0.54% increase in trade and a rise in the product of GDPs
per capita results in a 0.86% increase. Other things being equal, a 1%
increase in distance results in a 1.15% drop in trade over the 2003–2012
period.
Estimation of the model also showed that partners sharing a com-
mon border traded, on average, over the 2003–2012 period 2.27 times
188 Trade Agreements and Flows in ECOWAS
(e0.82 ) more between themselves than with others. Also on average for
2003–2012, landlockedness and insularity respectively resulted in a drop
of 1.79 (e0−58 ) and 9.68 (e2−27 ) times more.
Concerning the research subject, the results showed that sharing trade
agreement and single currency has a real impact on trade between mem-
ber countries. This result is confirmed both by the cross-sectional and
panel data estimations. During 2003–2012, countries sharing a trade
agreement and a single currency traded with each other, on average,
8.41 (e2.15 ) times more than countries outside the same monetary and
free trade zone. In the ECOWAS area, countries sharing a trade agree-
ment and single currency belong to WAEMU and the result is consistent
with those of Carrère (2004) and Masson and Patillo (2005) that showed
WAEMU countries trade more with each other than with the other
ECOWAS countries.
In the case of ECOWAS, the border effect is thought to be more signifi-
cant than the effect related to the use of the same language. For example,
trade of Togo and Benin (French-speaking countries) with, respectively,
Ghana and Nigeria (English-speaking countries) is buoyant. This sit-
uation is undoubtedly due to the statistical non-significance of the
coefficient associated with the common official language variable, while
the negative sign of the coefficient of the common colonial ruler vari-
able may suggest excessive polarization of trade from former colonies
to the former colonizing country rather than to each other. This was
noted in the case of WAEMU countries and their relations with France:
the nature of their monetary system (pegged to the Euro, a guarantee of
convertibility) should favor greater trade openness (Guillaumont, 2002)
and may also explain this result.
The result of the panel data estimation relating to GDP per capita also
appears interesting. It would indicate that, in a dynamic perspective,
an increase in GDP per capita, that is, purchasing power, widens the
national market. The need to satisfy this national market above all is
likely to reduce foreign trade. Overall, it was noted that trade potential
is significant and community development avenues multifarious. The
potential knock-on effects of intense economic cooperation between
ECOWAS member countries are enormous due to the number and
diversity of the populations making up the community.
Annexes
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2008–09 2009–10 2001–11
WORLD 729.1 693.1 724.8 814.9 901.1 1026 1148.2 1287.4 1056 1278.3 1480.4 −18% 21% 103%
NORTH.AM 265 258.7 267.7 301 332.6 364.9 385.4 413.2 334.3 413.2 479.8 −19% 24% 81%
CEN.SOUTH AM 57,6 50.9 51.5 60.7 71.5 87.2 105.4 135 108 137.1 166.4 −20% 27% 189%
EUROPE 177.7 160.1 168.9 187 204 235.8 272.5 311.1 249.9 276.3 315.7 −20% 11% 78%
CIS 3.8 3.8 3.7 4.8 5.9 7.1 10.5 13,8 8.1 9.2 12.9 −41% 14% 239%
AFRICA 12.3 10.7 10.7 13.4 15.5 18.8 24.1 28.8 24.7 28.7 33.2 −14% 16% 170%
M. EAST 19.2 18.9 19.4 23.5 31.5 37.2 45.1 55 44.8 48.7 58.4 −19% 9% 204%
ASIA 193.1 189.7 202.7 224.1 239.5 274.2 304.4 329.4 285.4 364 412.1 −13% 28% 113%
Africa’s share of global exports
AFRICA/WORLD 0.017 0.015 0.015 0.016 0.017 0.018 0.021 0.022 0.023 0.022 0.022
Note: NORTH. AM = North America/CEN.SOUTH.AM = Central and South America/CSI = Commonwealth of Independent States/M. EAST = Middle
East/Africa/World = ratio of African imports/global imports.
Source: International Trade Statistics, WTO, 2012 and authors’ calculations.
Table 9.A.2 Trend in imports of goods by region
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2008–09 2009–10 2001–11
WORLD 1179.2 1200.2 1303.1 1732.7 1910.1 1910.1 2020.4 2169.5 1605.3 1969.2 2265.9 −26% 23% 92%
NORTH.AM 352.6 349.2 366.5 417.7 464.6 507.9 531.6 559 407.1 512.2 585.7 −27% 26% 66%
CEN.SOUTH AM 71.7 73.7 83.8 105.2 129.9 140.4 141.8 167.4 113.5 136.4 178.8 −32% 20% 149%
EUROPE 252.8 259.7 282 317.5 346.8 369.9 393.5 409.6 314.5 359.2 416.5 −23% 14% 65%
CIS 8.1 8.4 10.4 14.9 19.8 25.3 26.5 38.5 24.1 32.2 42 −37% 34% 419%
AFRICA 26.8 23.3 33.9 48,3 67.9 83.8 95.2 117.3 64.6 87.5 94.9 −45% 35% 254%
M. EAST 38.7 35.8 43.9 54.2 65.7 74.7 80.3 115.3 61 76.9 106.3 −47% 26% 175%
ASIA 428.6 450.2 482.6 567.7 638 716 751.4 762.4 620.4 764.7 841.7 −19% 23% 96%
Africa’s share of global imports
AFRICA/WORLD 0.023 0.019 0.026 0.028 0.036 0.044 0.047 0.054 0.040 0.044 0.042
Note: NORTH. AM = North America/CEN.SOUTH.AM = Central and South America/CSI = Commonwealth of Independent States/M. EAST = Middle
East/Africa/World = ratio of African imports/global imports.
Source: International Trade Statistics, WTO, 2012 and authors’ calculations.
191
192
Table 9.A.3 Intra-African trade by economic zone in 2011, percentages of total imports
AMU CEMAC COMESA ECCAS ECOWAS FRANC ZONE SADC WAEMU AFRICA WORLD
Imports to −−−−−−−−−−−−→
AMU 4.0 0.1 2.2 0.1 0.4 0.3 0.5 0.2 6.4 100.0
CEMAC 1.8 2.7 1.1 2.8 6.5 4.8 1.7 2.1 13.5 100.0
COMESA 1.7 0.1 6.1 1.3 0.1 0.1 9.4 0.1 14.8 100.0
ECCAS 0.9 1.4 4.9 1.6 3.2 2.5 8.2 1.1 16.6 100.0
ECOWAS 0.8 0.2 0.3 0.3 6.8 3.4 1.6 3.2 9.7 100.0
FRANC ZONE 1.8 1.2 0.7 1.3 11.9 6.2 1.6 5.0 16.9 100.0
SADC 0.1 0.1 3.4 2.0 2.2 0.2 11.4 0.1 14.7 100.0
WAEMU 1.8 0.4 0.4 0.5 14.8 7.0 1.4 6.5 18.7 100.0
AFRICA 1.4 0.2 2.5 0.8 2.6 1.1 4.4 0.9 10.1 100.0
WORLD 0.7 0.2 0.6 0.6 0.8 0.3 1.2 0.1 3.3 100.0
Exports Imports
2011 2005–11 2009 2010 2011 2011 2005–11 2009 2010 2011
Notes
1. The Southern Africa Development Community (SADC).
2. The Interbank Electronic Banking Group (GIM) of the West African Economic
and Monetary Union (WAEMU) is an interbank payment system that provides
access to all automatic teller machines (ATMs) for the withdrawal of money in
the WAEMU zone. Each withdrawal is charged CFA F 500 (A transaction costs
CFAF 500).
3. The Central African Economic and Monetary Union (CEMAC).
4. A large body of empirical research finds that a pair if regions within a coun-
try tends to trade 10 to 20 times much than s an otherwise identical pair
of regions across countries. This result has been called the ‘border effect’.
McCallum (1995) was the first to find this effect.
5. This type of model has been used in three previous studies to underscore the
achievements of monetary union countries with regard to trade integration
(see Guillaumat, 2002), but this study addresses a more specific question using
more recent data.
6. The same behavior is noted for imports, see Figure 9.2 and Tables 9.A.1 and
9.A.2 in Annex).
7. WAEMU Trade Integration: Article 4 of the WAEMU Treaty (1994) stipulates
the creation of a common market with the free movement of persons, goods,
services and capital. The Common External Tariff (CET) has been in force since
January 2000.
8. ECOWAS comprises two separate sub-groups: First are the WAEMU countries
(Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal, Togo)
which form an economic and customs union with the CFAF as a common
currency. The second group is known as WAMZ, a project for a second West
African Monetary Zone with five countries: Gambia, Ghana, Guinea, Nigeria
and Sierra Leone). This group represents 75% of regional GDP and 70% of the
population. Cape Verde and Liberia are in neither group.
9. The gravity model is the name used for the family of quantitative models
developed by the astronomer J. Q. Stewart in 1940.
References
Afersorgbor, S. K. and van Bergeijk, P. A. G. (2011), ‘Multi-Membership and the
Effectiveness of Regional Trade Agreements in Western and Southern Africa. A
Comparative Study of ECOWAS and SADC’, ISS Working Paper Series, vol. 520.
Agbodji, A. E. (2007), ‘Integration and Intra-Sub-Regional Trade: Case of
WAEMU’, African Integration Review, vol. 1, no. 1, January 2007.
Avom, D. (2005), ‘The Determining Factors of Trade in CEMAC: An Empirical
Assessment’, Economie Appliquée, vol. LVIII, no. 2, pp.127–153.
Bénassy-Quéré, A. and Coupet, M. (2005), ‘On the Adequacy of Exchange
Rate Arrangements in Sub-Saharan Africa’, World Economy, vol. 28, no. 3,
pp. 349–373.
Carrère, C. (2004), ‘African Regional Agreements: Impact on Trade with or with-
out Currency Unions’, Journal of African Economies, vol. 13, no. 2, pp. 199–239.
Benjamin Ndong and Sokhana Diarra Mboup 195
Since its creation in 1963, the African Union has sought to foster
regional economic integration through monetary unions. But of six
major attempts at currency union, only the CFA Zone – using both
West African and Central African CFA francs – has been successful
(Carrère, 2004; Masson and Pattillo, 2004; Tsangarides et al., 2006,
2008; Tapsoba, 2009a, 2009b). Many have argued that African insti-
tutions are not sufficiently developed to deal with the heterogeneous
shocks to which African monetary unions have often been subject
(Eichengreen, 1992; Krugman, 1993; Fielding and Shields, 2001, 2005a,
2005b; Benigno, 2004; Debrun et al., 2003, 2005; Tsangarides and
Qureshi, 2006; Kabundi and Loots, 2007; Karras, 2007; Corsetti et al.,
2008; Masson, 2008).
But monetary unions create structural changes – in both trade and
risk-sharing – that mitigate the effects of heterogeneous shocks (Frankel
and Rose, 1997, 1998; Baxter and Kouparitsas, 2005; Caldéron et al.,
2007; Houssa, 2008; Inklaar et al., 2008). Thus the success of the
CFA Zone has renewed the interest of African policymakers in mon-
etary unions as a means of regional integration (Tsangarides et al.,
2005; Guillaumont-Jeanneney, 2006; De Grauwe, 2009; Tapsoba, 2009b;
Sénégas, 2010).
197
198 A DSGE Model of Trade and Risk-Sharing Effects
Trade and risk-sharing between the two CFA regions – the Economic and
Monetary Community of Central Africa (CEMAC) and the West African
Economic and Monetary Union (WAEMU) – were analyzed in a ‘two-
country’ model based on Badarau-Semenescu and Levieuge (2011) but
extended with key features of African economies. The model includes –
separately for each CFA region – the behavior of households in labor
supply, consumption and investment; of firms in both wholesale and
capital goods production; of retailers in price-setting; of commercial
banks in financing; and of the central bank and governments in pol-
icymaking. (Despite the existence of two central banks – BEAC and
BCEAO, one for each CFA region – they will be treated as one because
they maintain a common currency and thus a common monetary
policy.)
As is usual in less developed countries, households own the retail-
ers which – following Calvo (1983) – are assumed to exhibit nominal
price-stickiness. Producers are net-worth constrained, needing to bor-
row to finance production. Commercial banks use central bank credit
(refinancing) as well as deposits from households to finance production.
Risk-sharing is modeled via credit markets for bank loans because finan-
cial markets are not well developed in the CFA Zone. Heterogeneous
Thierry Kame Babilla 199
Householders’ program
We assume a continuum of infinitely-lived households represented by
the unit interval [0, 1], each maximizing the one-period utility function
σC −1 σh −1
∞
σC Ct C
σ
σh Ht h
σ
max Et βk
− (1)
k=0
σC − 1 σh − 1
Cγ1 C1−γ
CEMAC, C= 2
(2)
γ γ (1 − γ )1−γ
∗ 1−γ ∗ γ
C1 C2
WAEMU, C∗ = (2∗ )
γ γ (1 − γ )1−γ
1 − σ1C
λt = C (5a)
Pt t
Pt+1
0 = λt − βRDt+1 Et [λt+1 ] Et (5b)
Pt
0 = λt − βRAt+1 Et [λt+1 ] (5c)
Ht = (λt Pt Wt )σh (5d)
∗
D Pt+1 Pt+1
Rt+1 Et = RD∗
t+1 E t (6)
Pt Pt∗
λt+1 λ∗t+1
= ∗ (7)
λt λt
Pt∗
t = (9)
Pt
Producers’ program
Producers use the Cobb–Douglas constant-returns-to-scale technology
where (Y) is total output, (K) is capital, (L) is labor, (at ) is total factor
productivity and (α) is the output elasticity of capital. An autoregressive
productivity (technological) shock (ρa ) is included when
1−
Lt = Ht HtF (12)
Kt+1 = (1 − δ) Kt + It (13)
2
φ It
(IIt , Kt ) = −δ Kt (14)
2 Kt
(16)
and symmetrically for the other CFA zone. The first two conditions
define labor demands, while the third gives Tobin’s Q and the last rep-
resents expected gross return on capital. At optimum, firms’ demand for
capital ensures equality between the expected marginal cost of external
financing and the expected marginal return on capital.
Thierry Kame Babilla 203
Retailers’ program
Retailers purchase goods from wholesale producers, repackaging and
slightly differentiating them – introducing nominal price-stickiness –
and retailing them in a monopolistically competitive market at price P1,t
which is marked up at rate μt over their wholesale purchase price μ1 P1,t
t
so that their real marginal cost is ρt = μ1 .
t
The total supply of retail goods in the economy – which includes the
value of capital goods used by producers – is the total product of the
economy.
The composite index CES production function and price index are
then
1 ε−1
ε
ε−1
Yt = Yt j ε dj (18)
0
1−ε
1
P1,t = (19)
()
1,t j 1−ε dj
ε
Following Gali and Monacelli (2005), firms apply mark-up ε−1
to real
marginal costs (ρt+k ), yielding first-order condition
∞
Et (βς)k lt,t+k Yt +k j μ 1 P1,t+k
= ε t+k
P1,t
k=0
∞ (22)
ε−1
Et (βς)k lt,t+k Yt +k j
k=0
In each period, wholesale producers use these funds – plus their net
worth (NFt ), which is assumed insufficient to cover the entire cost – to
acquire Kt+1 of physical capital (at price Qt ) to be used during the next
period. Thus
Bt = Qt Kt+1 − NFt (26)
Optimally, producers’ expected return on capital RKt equals their
expected marginal financing cost (the real interest rate plus a credit
premium), or
Pt
Et RKt+1 = t+1 RBt+1 (Bt ) ηt+1 (28)
Pt+1
−ω
NF
where = q Kt+1 describes how the premium depends on the pro-
t t+1
ducers’ leverage ratio (Ben Bernanke et al., 1999), with (ω) being the
elasticity of the premium with respect to leverage and qt the real
price of capital qt = QPt , while (t+1 ) < 0 and (1) = 1. Thus, the credit
premium is an equilibrium inverse function of the aggregate financial
position in the economy, expressed by the leverage ratio.
Equation (28) provides the basis for the financial accelerator. If whole-
sale producers’ net worth goes up, the credit premium and the overall
cost of borrowing fall, yielding greater profits. Demand for capital equi-
librates its marginal return to its marginal productivity mpct – which
is the rental rate of capital – or
mpct + (1 − δ) qt
RKt = (29)
qt−1
where (δ) is the capital depreciation rate and (1 − δ) qt is the value of one
unit of capital. Wholesale producers’ accumulated net worth depends
on previous profits plus inheritance (t ) from exiting producers, which
evolves as
−ω
Pt NFt
NFt+1 = R K
q Kt − R (Bt ) ηt+1
t+1 −1
B
t qt−1 Kt − NFt
Pt+1 qt−1 Kt
+ (1 − ) t (30)
1
π̂tCU = π̂t + π̂t∗ (32)
2
1
ŷt =
CU
ŷt + ŷt∗ (33)
2
where π̂tCU and ŷtCU designates, respectively, the log deviations of infla-
tion and output of the CFA zone to the steady-state values. π̂t and π̂t∗ are
the log deviations of inflation respectively of CEMAC and WAEMU to
the steady-state values. ŷt and ŷt∗ represents the log deviations of output
of each of the two economies of the CFA zone to the steady-state values.
where, as before, asterisk designates the other CFA region.
Governments’ program
The two ‘national’ governments – for CEMAC and WAEMU – maintain
independent exogenous autoregressive fiscal policies with public spend-
ing ((ggt ) equal to lump-sum taxes (T
Tt )
CEMAC, Tt = Pt gt (34)
WAEMU, Tt∗ = Pt∗ gt (35)
where ρg between zero and one and εg,t → i. i. d. (0 σεg,t ) are random
public-spending shocks.
The labor market clears when first-order conditions for labor supply (5d)
and labor demand (15a) are equal. Substituting X into Y yields
σh +1
− σ1
Ht )
(H σh
= (Ct ) h ρt (t )1−γ (1 − α) Yt (38)
3 Dynamic Models
REAL INTEREST RATE FIRM EXTERNAL FINANCE PREMIUM BANK EXTRNAL FINANCE PREMIUM
×10–3
0.05 0.025 4
0 0.02 3
–0.05
0.015 2
–0.1
–0.15 0.01 1
–0.2 0.005 0
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
Figure 10.1 Impulse response functions to a negative productivity shock in CEMAC (dashed) and WAEMU (solid), years
211
OUTPUT CONSUMPTION INVESTMENT
0
212
0 0
–2
–10 –10
–4 WAEMU
–6 CEMAC –20 –20
–8
–30 –30
–10
–12 –40 –40
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
Figure 10.2 Impulse response functions to a negative monetary shock in CEMAC (dashed) and WAEMU (solid), years
OUTPUT CONSUMPTION INVESTMENT
0.2 –0.6 0.3
WAEMU –0.8
0.15 –1 0.2
CEMAC
–1.2
0.1 0.1
–1.4
0.05 –1.6 0
–1.8
0 –2 –0.1
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
TERM OF TRADE INFLATION NOMINAL INTEREST RATE
1.4 0.3 0.018
1.2 0.017
1 0.2 0.016
0.8 0.015
0.1
0.6 0.014
0.4 0 0.013
0.2 0.012
0 –0.1 0.011
2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
REAL INTEREST RATE FIRM EXTERNAL FINANCE PREMIUM BANK EXTERNAL FINANCE PREMIUM
×10–3 ×10–4
0.01 0 –2
0 –0.5
–4
–1
–0.01 –1.5 –6
–0.02 –2 –8
–2.5
–0.03 –10
–3
–0.04 –3.5 –12
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
Figure 10.3 Impulse response functions to negative public-spending shock in CEMAC (dashed) and WAEMU (solid), years
213
214 A DSGE Model of Trade and Risk-Sharing Effects
References
Adebiyi, M. A. and Mordi, C. N. (2010), ‘Estimating Small Scale
Macroeconometric Model (SSMM) for Nigeria: A Dynamic Stochastic General
Equilibrium (DSGE) Approach’, Central Bank of Nigeria.
Alesina, A. and Barro, R. J. (2002), ‘Currency Unions’, The Quarterly Journal of
Economics, vol. 117, no. 2, pp. 409–436.
Araujo, J., Bin Li, Poplawski-Ribeiro, M. and Zanna, L.-F. (2013), ‘Current Account
Norms in Natural Resource Rich and Capital Scarce Economies’. IMF Working
Paper,
r no. 13/80, Washington, DC: International Monetary Fund.
Thierry Kame Babilla 215
Avouyi-Dovi, S., Matheron, J. and Fève, P. (2007), ‘DSGE models and their
importance to Central Banks’, Banque de France Bulletin, vol. 161, no. 41.
Badarau-Semenescu, C. and Levieuge, G. (2011), ‘Assessing the Effects of Finan-
cial Heterogeneity in a Monetary Union: A DSGE Approach’, Economic Modeling, g
vol. 28, no. 6, pp. 2451–2461.
Barro, R. and Gordon, D. (1983), ‘A Positive Theory of Monetary Policy in a
Natural Rate Model’, Journal of Political Economy, vol. 91, pp. 589–610.
Baxter, M. and Kouparitsas, M. (2005), ‘What Determines Bilateral Trade Flows?’,
Working Paper Series, WP-05-11, Federal Reserve Bank of Chicago.
Bean, C. (2005), ‘Monetary Policy in an Uncertain World’, World Economics,
Economic & Financial Publishing,g vol. 6, no. 1, pp. 31–53.
Bean, C. (2006), ‘Impact of Globalization on Monetary Policy: Commentary’,
Proceedings Economic Policy Symposium, Jackson Hole, Federal Reserve Bank of
Kansas City, pp. 307–313.
Ben Bernanke, Gertler, S. M. and Gilchrist, S. (1999), ‘The Financial Accelerator in
a Quantitative Business Cycle Framework’, Handbook of Macroeconomics, vol. 1,
pp. 1341–1393.
Benigno, P. (2004), ‘Optimal Monetary Policy in a Currency Area’, Journal of
International Economics, vol. 63, pp. 293–320.
Berg, A., Portillo, R., Yang, S. and Zanna, L.-F. (2012), ‘Public Investment
in Resource Abundant Low-Income Countries’, IMF Working Paper/12/274,
Washington, D: International Monetary Fund.
Buffie, E. F., Berg. A., Patillo, C., Portillo, R. and L.-F. Zanna (2012), ‘Public
Investment, Growth, and Debt Sustainability: Putting Together the Pieces’, IMF
Working Paper,r no. 12/144, Washington, DC: International Monetary Fund.
Caldéron, C., Chong, A. and Stein, E. (2007), ‘Trade Intensity and Business
Cycle Synchronization: Are Developing Countries any Different?’, Journal of
International Economics, vol. 71, no. 1, pp. 2–21.
Calvo, G. A. (1983), ‘Staggered Prices in a Utility-Maximizing Framework’, Journal
of Monetary Economics, vol. 12, no. 2, pp. 383–398.
Carrère, C. (2004), ‘African Regional Agreements: Impact on Trade with or with-
out Currency Unions’, Journal of African Economies, vol. 13, no. 2, pp. 199–239.
Corsetti, G., Dedola, L. and Leduc, S. (2008), ‘International Risk-Sharing and the
Transmission of Productivity Shocks’, The Review of Economic Studies, Oxford
University Press.
Debrun, X., Masson, P. R. and Pattillo, C. A. (2003), ‘West African Currency
Unions: Rationale and Sustainability’, CESifo Economic Studies, vol. 49, no. 3,
pp. 381–413.
Debrun, X., Masson, P. R. and Pattillo, C. A. (2005), ‘Monetary Union in West
Africa: Who Might Gain, Who Might Lose, and Why?’, Canadian Journal of
Economics, vol. 38, no. 2, pp. 454–481.
De Grauwe, P. (2009), ‘The Fragility of the Eurozone’s Institutions’, Open
Economies Review, vol. 21, no. 1, pp. 167–174.
Del Negro, M. and Schorfheide, F. (2007), ‘Monetary Policy Analysis with
Potentially Misspecified Models’, Working Paper no. 13099, Cambridge, MA:
National Bureau of Economic Research.
Diop, M. B. (2011), ‘Un modèle de prévision de l’inflation à moyen terme dans
l’UEMOA: une approche DSGE’, Direction de la Recherche et de la Statistique de la
BCEAO.
216 A DSGE Model of Trade and Risk-Sharing Effects
Masson, P. R. and Patillo, C. A. (2004), ‘A Single Currency for Africa’, Finance and
Development, t December, Washington, DC: International Monetary Fund.
McKinnon, R. I. (1963), ‘Optimum Currency Areas’, American Economic Review,
vol. 53, no. 4, pp. 717–725.
Mundell, R. (1961), ‘A Theory of Optimum Currency Areas’, American Economic
Review, vol. 51, no. 4, pp. 657–665.
Peiris, S. and Saxegaard, M. (2007), ‘An Estimated DSGE Model for Monetary
Policy Analysis in Low-Income Countries’, IMF Working Paper, r no. 07/282,
Washington, DC: International Monetary Fund.
Rogoff, K. (1985), ‘The Optimal Degree of Commitment to an Intermediate
Monetary Target’, Quarterly Journal of Economics, vol. 100, pp. 1169–1190.
Rose, A. (2000), ‘One Money, One Market: Estimating the Effect of Common
Currencies on Trade’, Economic Policy, vol. 30, pp. 9–45.
Rose, A. K. and Stanley, T. D. (2005), ‘A Meta-Analysis of the Effect of Common
Currencies on International Trade’, Journal of Economic Surveys, vol. 19, no. 3,
pp. 347–367.
Samaké, I. (2010), ‘A Macro Model of the Credit Channel in a Currency Union
Member: The Case of Benin’, IMF Working Paper, r no. 10/191, Washington, DC:
International Monetary Fund.
Sénégas, M.-A. (2010), ‘Bilateral Trade, Colonial Heritage and Common
Currency Arrangement: An Oceanian Perspective’, Economie Internationale,
CEPII Research Center, vol. 129, pp. 63–98.
Tapsoba, J.-A. Sampawende (2009a), ‘Union Monétaire en Afrique de l‘Ouest:
Quelles Réponses à l’Hétérogénéité des Chocs?’, Etudes et Document, t CERDI.
Tapsoba, J.-A. Sampawende (2009b), ‘Trade Intensity and Business Cycle
Synchronicity in Africa’, Journal of African Economies, vol. 18, pp. 287–318.
Tavlas, G. (1993), ‘The “New” Theory of Optimum Currency Areas’, World
Economy, vol. 16, no. 6, pp. 663–685.
Taylor, J. B. (1980), ‘Output and Price Stability: An International Comparison’,
Journal of Economic Dynamics and Control, vol. 2, pp. 109–132.
Tsangarides, G., Charalambos, G. and Van den Boogaerde, P. (2005), ‘Ten Years
after the CFA Franc Devaluation: Progress Toward Regional Integration in
the WAEMU’, IMF Working Paper, r no. 05/145, Washington, DC: International
Monetary Fund.
Tsangarides, G., Charalambos, G. and Qureshi, M. (2008), ‘Monetary Union
Membership in West Africa: A Cluster Analysis’, World Development, t vol. 36,
no. 7, pp. 1261–1279.
Tsangarides, G. and Qureshi, M. S. (2006), ‘What Is Fuzzy about Clustering in
West Africa?’, IMF Working Papers, no. 06/90, Washington, DC: International
Monetary Fund.
Conclusion: Enhancing
Intra-African Trade through
Regional Integration
Mthuli Ncube, Issa Faye and Audrey Verdier-Chouchane
Introduction
218
Mthuli Ncube et al. 219
growing supply potential into the regional and global markets. This is
why trade facilitation and other measures to reduce transport costs and
increase supply capacity are so vital.
African RECs have set ambitious targets toward achieving full integra-
tion. But in practice, RECs suffer from delays in the ratification of
regional treaties and protocols by member states, gaps in the imple-
mentation of regional trade agenda and non-compliance with regional
arrangements (Erasmus, 2011).
The role of institutions is crucial in both implementing regional trade
agreements and in improving competitiveness and the cost of doing
business in Africa. For Joseph Parfait Owoundi (Chapter 1), the coun-
tries that form the Economic and Monetary Community of Central
Africa (CEMAC) would enhance their competitiveness and register better
macroeconomic performances by improving their business environ-
ment, institutional integration and regulatory frameworks. This requires
goodwill from political leaders, commitment from the nation and will-
ingness to surrender some national sovereignty to the regional level.
The absence of sanctions or an incentive mechanism is one of the most
important reasons for the lack of commitment from African govern-
ments for the regional trade integration agenda. Most countries fail to
clearly see what is in it for them to promote and implement a regional
integration agenda.
According to Wumi Olayiwola, Evans Osabuohien, Henry Okodua
and Oluyomi Ola-David (Chapter 2), further incentives are needed to
enhance the implementation of the Economic Community of Western
African States’ (ECOWAS) Agricultural Policy (ECOWAP) and Trade Lib-
eralization Scheme (ETLS), in order to enhance the agricultural exports
within the region. ECOWAS members should be committed to the eco-
nomic integration agenda and especially to reducing bureaucratic delays
associated with preparing export documents. The authors also advo-
cate strengthening the institutional framework to curb the threat of
corruption and ensure adherence to the rule of law.
Also looking at the ECOWAS zone, Benjamin Ndong and Sokhana
Diarra Mboup (Chapter 9) argue that Regional Trade Agreements (RETs)
within the frame of a monetary union significantly trigger bilateral
trade. Therefore, they recommend that ECOWAS evolves toward a wider
monetary union in order to increase regional trade. However, the cur-
rency union should be complemented simultaneously by more credible
commitments at the institutional level and structural reforms aimed at
promoting complementarity between the ECOWAS economies. Bad gov-
ernance and informal trade would also need to be tackled to increase
regional trade. For Thierry Kame Babilla (Chapter 10), currency unions
Mthuli Ncube et al. 221
failed to sustain RI in the CFA Franc zone because of the lack of insti-
tutions and mechanisms for sharing risk within the zone. Financial
asymmetries have led to the widening of national differences among
members countries.
Edris Seid (Chapter 5) recommends that countries adopt and imple-
ment coherent and coordinated trade policies to promote intra-regional
trade. In particular, simplifying custom procedures makes the transit
system more efficient, faster and less costly for both importers and
exporters. This argument is corroborated by the World Bank (2011) for
which a well-functioning legal system is also a key requirement for con-
tract enforcement, enabling firms to expand their markets and networks.
Appropriate business environments facilitate risk and cost management.
The private sector, whether foreign or domestic, is more likely to commit
to new projects or the expansion of existing ones in a business-friendly
environment.
Institutions
7
Innovation Infrastructure
6
5
Business Macroeconomic
sophistication 4 environment
3
2
Health and
Market size 1 primary
education
5 Conclusion
Africa. Yet the continent lacks the adequate institutions and regula-
tions to improve the overall business environment and facilitate trade.
Moreover, hard infrastructure, in particular transport, connectivity and
energy constitute a pervasive long-term challenge. Sub-Saharan Africa’s
deficit in infrastructure largely explains its low levels of intra-African
trade.
Furthermore, there are a myriad of challenges at the firm level which
limit the development of the private sector and the emergence of
high-quality exports. Referring to the industrialization policy, African
countries should move from an inward to an outward focus so as to
encourage competition and generate regional industrial capacity. Value
chain-related activities constitute an important driver of RI. African
countries need to develop regional markets and trade and to achieve
a relevant structural transformation of their economies through the
production of higher-value commodities. Improved access to afford-
able finance and skills development and training would greatly support
innovation and industrial technology. Mobilizing domestic resources
would be a first step for African countries to exploit their comparative
advantages and enter regional and then global value chains.
Finally, for RI to be more successful in Africa, political leadership is
needed to champion the process at the national level. African countries
should also reduce the disparity between actors with regards to regional
policy goals, the implementation process and the benefits from regional
trade. For instance, in the ECOWAS region, Nigeria, Côte d’Ivoire and
Senegal account for almost 90% of intra-regional exports and 50% of
intra-regional imports (UNCTAD, 2011).
Note
1. www.enterprisesurveys.org, accessed 21 April 2014.
References
African Development Bank [AfDB] (2014a), Regional Integration Strategy 2014–
2023, Tunis: AfDB.
African Development Bank [AfDB] (2014b), African Development Report 2013/14:
Leveraging Regional Integration for Inclusive Growth, Tunis: AfDB.
African Union (2012), Action Plan for Boosting Intra-Regional Trade. Available at
http://www.au.int/ [Accessed 23 May 2014].
Ancharaz, V., Mbekeani, K. and Brixiova, Z. (2011), ‘Impediments to Regional
Trade Integration in Africa’, Africa Economic Brief,
f vol. 2, no. 11, pp. 1–8, African
Development Bank.
228 Conclusion
Arvis, J. -F., Duval, Y., Shepherd, B. and Utoktham, C. (2010), ‘Trade Costs in
the Developing World 1995–2010’, Policy Research Working Paper, r no. 6309, The
World Bank.
Department for International Development [DfID] (2011), ‘Regional Integration
and Trade in Sub-Saharan Africa’, Trade and Investment Analytical Papers, DfID:
Department for business Innovation and Skills.
Erasmus, G. (2011), ‘Deeper Regional Integration in SADC. Will EPAs Undermine
the Process?’ Working Paper,r Trade Law Centre for Southern Africa.
Kumar, C. S. (2006), ‘Human Capital and Growth Empirics’, Journal of Developing
Areas, vol. 40, no. 1, pp. 153–179.
McCarthy, C. (2007), ‘Is African Economic Integration in Need of a Paradigm
Change? Thinking Out of the Box on African Integration’, in A. Bösl et al.
(eds.), Monitoring Regional Integration in Southern Africa, Yearbook Vol. 6–2007.
Stellenbosch: Trade Law Centre for Southern Africa.
OECD/WTO, Aid for Trade at a Glance 2013: Connecting to Value Chains.
WTO/OECD Publishing, 2013.
Programme for Infrastructure Development in Africa (PIDA) (2014), website,
section ‘PIDA sectors’. Available from http://www.au-pida.org/ (Accessed
23 May 2014).
Sackey, H. A. (2011), Human Capital Development in Africa: Agents, Drivers and
Implications for Growth and Structural Transformation, Addis-Ababa: UNECA.
United Nations Conference on Trade and Development [UNCTAD] (2011),
Strengthening Regional Economic Integration for Africa’s Development,t UNCTAD,
Division for Africa, Least Developed Countries and Special Programmes,
Geneva.
United Nations Economic Commission for Africa [UNECA] (2013), Economic
Development in Africa Report – Intra-African Trade: Unlocking Private Sector
Dynamism, Addis-Ababa: UNECA.
United Nations Economic Commission for Africa [UNECA] (2011), ‘Industrial
Policies for the Structural Transformation of African Economies: Options
and Best Practices’, Policy Research Paper, r no. 2, United Nations Economic
Commission for Africa, Addis Ababa, Ethiopia.
World Bank (2008), World Development Report 2008. Washington, DC:
World Bank.
World Bank (2010), World Development Indicators, Washington, DC: World Bank.
World Bank (2011), Doing Business 2011: Making a Difference for Entrepreneurs,
Washington, DC: World Bank Group.
World Bank (2013), World Development Indicators, Washington, DC: World Bank.
World Economic Forum, World Bank and AfDB (2013), The Africa Competitiveness
Report,
t Geneva: World Economic Forum.
Index
Note: The letter ‘n’ ‘t’ ‘f’ following locators refers to notes, tables and figures
respectively.
229
230 Index