The objective of the study is to determine the effect of the dynamics of regional integration on economic growth in the West African Economic and Monetary Union (WAEMU). The autoregressive vector analysis using the Cholesky decomposition used data from the Central Bank of West African States, United Nations Conference on Trade and Development statistics, and world development indicators. The results show that the overall contribution of a WAEMU country to integration accounts for 0.5% of the variation in growth in the community area over the long term. The effects of the free trade area and the customs union are respectively higher on economic growth than economic union and economic and monetary union. The establishment of an effective free trade area would lead to better economic growth. In addition, taking the common market into account would more than capture the contribution of the integration process to economic growth and would be a prospect for future research.
Economic integration aims at the adoption of a common currency and the opening of the economic markets of the member countries of the group and proceeds through the creation of a free-trade area, a customs union, a common market, and an economic and monetary union. Increasing the size of an economic market through economic integration would have positive effects on economic growth. The process of regional integration becomes an instrument of economic policy and a result to improve the living conditions of the inhabitants [
In Africa, recent or emerging supranational organisations are, in this global context, reflecting the development model “a country is stronger when it is alone”. Economic groupings, such as the West African Economic and Monetary Union (WAEMU), raise questions about the outcome of the phases of the process of building integration. Indeed, the success of WAEMU requires positive effects in terms of economic growth [
The main objective of this article is to empirically determine the effect of the integration process on economic growth. Specifically, it is a question of measuring the interdependence between WAEMU’s level of economic integration and its economic growth. The assumption is that the stages of the process of economic integration have a differentiated positive influence on economic growth. An autoregressive vector model was used on data from the Central Bank of West African States, World Development Indicators, and UNCTAD statistics to measure the contribution of the integration process phase to growth. The results show erratic positive effects of the integration phases.
After the introduction, the empirical effects of economic integration are presented first. Second, the methodology and data sources of the study are described. The results and the debates of the results are discussed before the general summary as a conclusion.
The elimination of any discrimination or barrier between the economies of the member countries and the establishment of a supranational structure is the aim of economic and monetary integration. A government with a tax and economic policy specific to the integration of member countries would then be established. Several intermediate phases exist between the beginnings of the process until the advent of economic integration. Thus, one successively has the zone of free trade, the customs union, the common market, the economic and monetary union, and finally, the political union [
Economic integration processes include positive integration through the harmonisation and coordination of existing economic instruments and negative integration with the removal of all barriers to the movement of goods, services, and factors of production between countries [
Economic growth differs according to the zones of integration. Indeed, the European Union contributes to the growth of 0.6 to 0.8 percentage points. In South America, economic growth went hand in hand with the increase in intra-Mercosur trade, from 10% to almost 20% between 1991 and 1994 [
The fall in tariffs on regional imports has favoured the exports of regional partner countries in the WAEMU zone [
The effects of regional integration on growth are apprehended by an index of economic integration. The model regressed on the panel data allows the effects of regional integration on the growth in the regions to be analysed separately and jointly [
An increased endogenous growth model enhances the phases of the integration process [
Y i t = β i t + Γ i t ( L ) Y i t + ε i t , (1)
where Y i t is the dependent variable, ε i t is the error term, Γ i t ( L ) is a polynomial matrix defined on the delay operator, and L and β i t are the period and individual coefficient matrices. The functional form of Γ i t ( L ) is as follows:
Γ i t ( L ) = β i t + Γ i t 1 L 1 + Γ i t 2 L 2 + ⋯ + Γ i t p L p + μ i , (2)
with Γ i t k ; k = 1 to p.
The heterogeneity of individuals is considered by the unobservable individual effects μ i that affect the growth dynamics of the countries. The temporal effects are introduced by temporary dummies variables, d t , intended to measure shocks that will uniformly affect all WAEMU countries. A composite error specification is made after the Honda test [
Y i t = β 0 + Γ ( L ) Y i t + d t + μ i + ε i t . (3)
The use of least ordinary squares and least generalised squares provides non-convergent intra-individual and inter-individual estimators due to endogeneity [
Y i t ˜ = Γ ( L ) Y ˜ i t + ε i t ˜ , (4)
where Y i t ˜ = Y i t 1 ˜ , Y i t 2 ˜ , ⋯ , Y i t M ˜ and ε i t ˜ = ε i t 1 ˜ , ε i t 2 ˜ , ⋯ , ε i t M ˜ .
Growth is the completion of a production process influenced by trade and investment. Empirically, two models were used to estimate the effects of integration steps on economic growth. The first model seeks to capture the effect of economic integration in general on growth and is written as follows:
Y i t = ( I C C E i t , V C I i t , T X _ I N V i t , T P I B T i t ) , (5)
With TPIBT the growth rate of GDP per capita, TX_INV is the investment rate, VCI is the variation of intra-community trade, and ICCIE is the composite index of contribution to economic integration. The second model individually integrates the different phases of the integration process and is written as follows:
Y i t = ( I C L E i t , I C U D i t , I C U E i t , I C U E T i t , V C I i t , T X _ I N V i t , T P I B T i t ) , (6)
where ICLE, ICUD, ICUE, and ICUET are the indices of contribution to free trade, to the customs union, to the economic union, and to the total economic union, respectively. Integration always begins with the free-trade area [
Moreover, the stationary with (Im, Pesaran, and Shin tests) analysis shows that all the variables are stationary at level with the exception of stationary CLE and ICUET in the first difference at the level 1 (
The information criteria of Akaike (AIC), Schwarz (SC), and Hannan and Quinn (HQ) show an optimal delay of 1 for both models (
Economic phenomena are sometimes explained by indices, and the contribution of each phase of the integration process to growth is analysed by indices [
and is a tool for analysing specialisation models [
The variables used for constructing the indices are grouped into five broad categories. Each of them corresponds to one of the integration phases [
Estimated data from WAEMU countries cover the period 2000-2013 with 112 observations. They come from the Central bank of west Africa State for sectorial added values (gross fixed capital formation, changes in intra-regional trade, public expenditures, and interest rates) the UNCTAD statistics for the statistics on customs duties, the World Bank for data on the WAEMU Common External Tariff, and GDP per capita growth rates.
This section presents the analysis of the impulse response functions and the decomposition of the variance.
With respect to the common external tariff, the signing of free-trade agreements
Variables | Statistics | Stationarity | |
---|---|---|---|
Level | First difference | ||
TPIBT | −6.534* | - | I (0) |
TX_INV | −6.046* | - | I (0) |
VCI | −6.643* | - | I (0) |
ICUET | - | −9.603* | I (1) |
ICUE | −6.367* | - | I (0) |
ICLE | - | −10.092* | I (1) |
ICUD | −6.790* | - | I (0) |
ICCIE | −2.763* | - | I (0) |
*stationary; Authors’ own calculation.
Delay | Model 1 | Model 2 | ||||
---|---|---|---|---|---|---|
AIC | SC | HQ | AIC | SC | HQ | |
1 | 10.679* | 11.274* | 10.918* | −9.317* | −7.650* | −8.649* |
2 | 10.891 | 11.963 | 11.321 | −9.008 | −5.881 | −7.754 |
3 | 11.146 | 12.695 | 11.767 | −9.085 | −4.500 | −7.247 |
4 | 11.294 | 13.319 | 12.106 | −8.673 | −2.629 | −6.250 |
(*) indicates lag order selected by the criterion; Authors’ own calculation.
Indices | Signification |
---|---|
I C L E c , t = 1 − T D M c , t | TDMc,t is the average customs tariff that the country applies to the rest of the region. ICLEc,t measures the country’s openness to intra-regional trade and is the arithmetic average of the intra-zone tariffs of the member countries to capture the liberalisation of trade within the WAEMU. |
I C U D c , t = min ( T M c , t ; T M R t ) max ( T M c , t ; T M R t ) | TMc,t is the country’s average fare c at date t, and TMRt is the regional average fare for all countries in the zone. ICUD is the simple arithmetic average of tariffs applied to member countries of the World Trade Organization. |
I C U E c , t = min ( P S c , t P S c , t − 1 ; V M P i , t ) max ( P S c , t P S c , t − 1 ; V M P i , t ) | PSc,t and PSc,t−1 are the sectorial productions of country c, respectively, for periods t and t − 1. VMPi,t is the average variation of production in the sector considered for region i at date t. |
I C P T I c , t = min ( T I M R i , t ; T I M c , t ) max ( T I M R i , t ; T I M c , t ) I C H B c , t = min ( P D P c , t ; V M P D P i , t ) max ( P D P c , t ; V M P D P i , t ) I C P I c , t = min ( V I P C M i , t ; I P C c , t I P C c , t − 1 ) max ( V I P C M i , t ; I P C c , t I P C c , t − 1 ) | The ICUET is an aggregation of three indices and is built in three stages: 1) TIMRi,t is the regional average interest rate at date t, TIMc,t t is the average interest rate of the country c at date t, and ICPTI is the Index of Contribution at the proximity of the interest rate; 2) ICHB is the Budget Harmonisation Contribution Index, PDPc,t is the share of public expenditure in the country’s GDP c at date t, and VMPDPi,t is the average value of the share of public expenditure in the GDP of region i at date t; and 3) ICPI is the Inflation Proximity Index, where IPCc,t and IPCc,t−1 are the country’s Consumer Price Index c at dates t and t − 1, respectively, and VIPCMi,t is the average change in the Consumer Price Index in region I on date t. The last step in the construction of the ICUET is to aggregate the three previously constructed indicators through a simple arithmetic average. |
Authors’ own synthesis
between the WAEMU member countries is a kind of shock to raise the level of indices for each step of integration. The reaction of the growth rate of GDP per capita is observed.
The extreme curves represent confidence intervals at the 5% threshold, and the middle curve represents the evolution of GDP per capita. The shock of a variable justifies 95% of the growth behaviour. The shock of economic integration is positive on the GDP per capita growth rate following a shock of the ICCIE variable. This shock has a persistent effect on growth (
A shock on free trade positively affects much of the economic growth and can be explained by the nature of the products traded in WAEMU (see
The ICUD shock shows a negative dynamic on GDP growth up to the seventh year (see
The reaction of the GDP per capita following a shock of the economic union is negative the first three years and positive for the rest of the period (see
The Gross domestic product reacts positively and instantly to the shock of the ICUET. However, this reaction tends to be marginal from the fifth year (see
The goal of decomposing the variance of the forecast error is to understand its percentage contribution for each stage of the integration process. The part of variation of a system variable explained by another observed variable is analysed by the decomposition of Cholesky.
Induced growth changes in the free-trade area and the customs union are 2.36% and 1.66%, respectively. These results are similar to those of [
Period | S.E. | ICCIE | VCI | TX_D_INVEST | TPIBT |
---|---|---|---|---|---|
1 | 0.015 | 100 | 0 | 0 | 0 |
5 | 0.020 | 97.413 | 1.821 | 0.285 | 0.481 |
10 | 0.020 | 96.618 | 1.858 | 1.018 | 0.505 |
S.E.: Standard Error; Authors’ own calculation.
Period | S.E. | ICLE | ICUD | ICUE | ICUET | VCI | TX_D_INVEST | TPIBT |
---|---|---|---|---|---|---|---|---|
1 | 0.004 | 100 | 0 | 0 | 0 | 0 | 0 | 0 |
5 | 0.010 | 53.217 | 20.828 | 2.202 | 13.590 | 1.490 | 4.860 | 3.812 |
10 | 0.014 | 35.700 | 15.268 | 1.403 | 27.446 | 0.792 | 17.022 | 2.368 |
Cholesky ordering: ICLE ICUD ICUE ICUET VCI TX_D_INVEST TPIBT. Authors’ own calculation.
Period | S.E. | ICLE | ICUD | ICUE | ICUET | VCI | TX_D_INVEST | TPIBT |
---|---|---|---|---|---|---|---|---|
1 | 0.002 | 0.518 | 99.482 | 0 | 0 | 0 | 0 | 0 |
5 | 0.002 | 0.757 | 90.648 | 1.065 | 2.754 | 0.426 | 2.682 | 1.668 |
10 | 0.002 | 0.756 | 88.681 | 1.130 | 3.066 | 0.422 | 4.282 | 1.662 |
Cholesky ordering: ICLE ICUD ICUE ICUET VCI TX_D_INVEST TPIBT. Authors’ own calculation.
Period | S.E. | ICLE | ICUD | ICUE | ICUET | VCI | TX_D_INVEST | TPIBT |
---|---|---|---|---|---|---|---|---|
1 | 0.036 | 0.106 | 1.705 | 98.188 | 0 | 0 | 0 | 0 |
5 | 0.040 | 0.424 | 2.586 | 86.487 | 0.701 | 6.835 | 2.764 | 0.202 |
10 | 0.040 | 0.473 | 2.527 | 84.227 | 1.237 | 6.701 | 4.600 | 0.235 |
Cholesky ordering: ICLE ICUD ICUE ICUET VCI TX_D_INVEST TPIBT. Authors’ own calculation.
Period | S.E. | ICLE | ICUD | ICUE | ICUET | VCI | TX_D_INVEST | TPIBT |
---|---|---|---|---|---|---|---|---|
1 | 0.044 | 2.709 | 2.651 | 0.529 | 94.112 | 0 | 0 | 0 |
5 | 0.079 | 5.080 | 1.913 | 1.802 | 90.059 | 0.857 | 0.188 | 0.100 |
10 | 0.094 | 6.120 | 2.295 | 1.831 | 86.943 | 0.828 | 1.844 | 0.137 |
Cholesky ordering: ICLE ICUD ICUE ICUET VCI TX_D_INVEST TPIBT. Authors’ own calculation.
conducted in Central Africa. As for the economic union and total economic union, their contributions are respectively 0.23% and 0.14% in the long term and are specific to the WAEMU zone. The process of economic integration of the WAEMU zone does not respect the classical scheme as advocated by integration theorists [
The determination of the contribution of each step of the process of economic and monetary integration on the economic growth of the WAEMU was analysed by a model autoregressive vector specified on a panel from 2000 to 2013. The results show that the effect of free trade and customs union on growth are respectively 2.36% and 1.66%. The economic union and economic and monetary union are two stages of the integration process that have the least effect on growth, with contributions of 0.23% and 0.14% respectively. Economic integration explains 0.5% of the variation of the economic growth of the countries. The establishment of an effective free-trade zone that scrupulously respects the common external tariff set by WAEMU is more beneficial. The mobility of the factors of the production of goods and services has a greater effect on the economies of the countries.
In addition, the common market stimulates trade through the free movement of people, goods, and services to have inclusive and equitable growth, reducing inequalities. Its inclusion in the analysis in the case of data availability would improve the understanding of the contribution of the process of economic and monetary integration. This remains a prospect of future investigation.
The authors declare no conflicts of interest regarding the publication of this paper.
Ouedraogo, S. and Drabo, D. (2019) Dynamics of Integration and Economic Growth of the West African Economic and Monetary Union (WAEMU). Modern Economy, 10, 1121-1133. https://doi.org/10.4236/me.2019.104076