REAL CONVERGENCE IN WAEMU COUNTRIES

Full Text

Economic integration can be defined as a process that allows a given region to expand its economic space in order to create a single market. West African history shows that the first generation of economic integration initiatives was the Council of the Entente. After observing divergences and a dispersion of energy, the second generation established organizations such as ECOWAS.

However, the problems of coordination and harmonization of integration mechanisms observed at the level of a given subset were further amplified with multiple organizations. Moreover, a state's capacity to bear the costs of these different organizations was very limited. Therefore, it was necessary to determine the optimal space for economic integration and establish a single organization, which became ECOWAS.

Many economists have already examined the problem of integration (Bela Balassa 1993; Azam 1997). To be more effective in achieving their objectives, the member countries of the WAEMU, like developed economies, have decided to implement "signals" called "convergence criteria." These criteria are divided into first-order and second-order criteria. First-order criteria are those whose non-compliance leads to corrective measures and sometimes even sanctions. The primary criteria are four in number. They are worded as follows: (i) Ratio of the budget balance to nominal GDP (key criterion): it should be greater than or equal to 0% in 2002. As a key criterion, non-compliance will result in sanctions; (ii) Average annual inflation rate: it should be maintained at a maximum of 3% per year; (iii) Ratio of outstanding domestic and external debt to nominal GDP: it should not exceed 70% in 2005; (iv) Payment arrears: (a) domestic payment arrears: no accumulation of arrears during the current period; (b) external payment arrears: no accumulation of arrears during the current period.

Second-order criteria are considered indicative structural benchmarks that are rigorously monitored due to their crucial role in achieving internal and external balances within economies. Failure to meet these criteria does not, however, trigger corrective or remedial measures. Nevertheless, they may be the subject of economic policy recommendations. There are four of these criteria: (i) Ratio of the wage bill to tax revenue: it should not exceed 35%; (ii) Ratio of public investment financed by domestic resources to tax revenue: it should reach at least 20%; (iii) Ratio of the current account balance excluding public transfers to nominal GDP: it should be greater than or equal to -5%; (iv) Tax burden: it should be greater than or equal to 17%.

The concept of convergence criteria involves the notions of σ and β convergence. What do these two notions indicate or teach us? Studies on economic growth employ various definitions of real convergence (Quah, 1993). However, in addition to the criteria as presented, all these definitions revolve around two concepts: β convergence (or beta convergence) and σ convergence (or sigma convergence). β convergence (or beta convergence) refers to the process by which economies adjust, over time, towards the same growth path or towards a reference value. In growth theories, the idea that poor countries will catch up with rich countries if they achieve a higher growth rate leads to the use of the β-convergence test. Thus, β-convergence is observed in a sample of countries if there is a negative relationship between the growth rate of per capita income and the initial level of income; that is, if poor economies tend to grow faster than rich economies.

Sigma convergence (σ) measures the degree of convergence over time between several economies with respect to one or more indicators or criteria. Sigma convergence occurs when the dispersion of real per capita income decreases over time (Sala-i-Martin, 1995). Its analysis is based on studying the evolution of the dispersion of the time series considered. Convergence of the entire sample occurs when the dispersion decreases over time. The dispersion indicator used can be the variance or the standard deviation of the time series. It would be possible to speak of "perverse convergence" when, in the case of real convergence for example, the decrease in the standard deviation is the result of a decline in GDP per capita from the initially richer countries to the poorer countries. σ-convergence analysis indicates whether the real gross domestic products per capita of different countries within a zone have tended to converge or diverge over the analysis period. The term convergence, according to the Maastricht Treaty (1991), could be seen as the gradual reduction of differences in macroeconomic indicators between several states and required that these indicators fall within a range, meaning they must converge to within a constant without exceeding the standard proposed by the treaty. Several forms of convergence can be distinguished, the main ones being: nominal convergence, structural convergence, and real convergence. Therefore, analyzing the concept of convergence in the WAEMU zone is essential given the clear commitment of African authorities to integration and the development challenges they face. Our study thus aims to analyze real convergence in the WAEMU zone. The following specific objectives are linked to this general objective:

- Show the GDP trends of the WAEMU member states.

- Analyze school enrolment rates among the different WAEMU countries.

This paper will allow us to assess the evolution and impact of harmonization policies through the integration structure within the WAEMU zone.

Literature review

For some authors, the regionalization of trade provides an ideal framework for accelerating economic performance and reducing development gaps between unequally developed countries (Solow, 1956). Several authors have analyzed the effect of fiscal convergence on the synchronization of real business cycles through various transmission channels of monetary policy to the real sector (Clarida et al., 1999), basing their analysis on the interest rate channel. Indeed, according to these authors, fiscal convergence in a monetary union limits the divergence of real interest rates and facilitates the convergence of real variables such as consumption and investment. According to Rodrik (2011), some developing countries have recently experienced higher levels of development than developed countries. This trend is not limited to Asia, but is also observed in South America and even Africa. These findings seem to confirm the teleological theory developed by Géneau de Lamarlière and Staszak (2000), which posits that all economies are destined to converge; hence the possibility of economic catch-up with advanced capitalist economies. For Mishkin (2015), credit facilitates the convergence of budget deficits, leading to macroeconomic stability that increases banks' profitability expectations and thus encourages them to lend in all countries. Following the adoption of the new convergence criteria by the WAEMU member states, some authors have focused their analysis on these new criteria through retrospective studies. Diarra's 2016 study revealed a divergence in the basic budget balance and the average annual inflation rate. The research findings showed that the new budget balance criterion imposes fewer requirements on member states than the previous one. According to De Grauwe (2018), fiscal convergence enhances the central bank's credibility in the fight against inflation. Economic agents thus align their expectations with the rules issued by the central bank, resulting in more coordinated responses from different economies. Bamba (2025) analyzed the influence of institutional changes on the synchronization of business cycles in WAEMU countries from 2000 to 2020, focusing on fiscal convergence. His results indicate that reducing overall budget balance disparities leads to an increase in the correlation of real growth rates. Therefore, community authorities should implement a mechanism to facilitate compliance with the overall budget balance criterion. This mechanism could consist of a bonus for fiscally responsible countries, depending on the level of the observed deficit as well as the cost of the effort. In light of this educational situation, some studies have examined the relationship between current human capital and economic convergence in Africa (Garang and Erkekoglu, 2021; Ibourk and Elouaourt, 2023; d’Alfoul et al., 2024). From the above, the importance of education in the economic development of African countries is evident. It is also a specific factor and prerequisite for the convergence of African economies. However, the relevance of this intangible capital depends on its quality and therefore its capacity to meet the needs of these economies.

Methodology

The model to be estimated is a convergence equation by Barro (1991) supplemented by Philippe (1997). The dependent variable is the GDP per capita growth rate. The explanatory variables selected are, respectively, GDP per capita in 1995, the share of the population enrolled in secondary education (TSS) for the year 1995, and the average share of gross fixed capital formation (GFCF).

Other explanatory variables, such as the average share of public spending (PS) as a percentage of GDP and the share of foreign direct investment (FDI), are also taken into account in the model. The convergence equation used is identical to Philippe's, except for the observation period and the reference year. This is primarily due to a constraint on data availability for previous years concerning the eight countries of the Union. Thus, our model is as follows:

 

The pibh variable represents the growth rate of GDP per capita of country i between 1995 and 2025. This is our endogenous variable. Next, we have the explanatory variables represented by pibh, which is GDP per capita in 1995, tss, the share of the population of country i enrolled in secondary education in 1995, ifbcf, the average share of investment in GDP over the period, dppib, the average share of public spending as a percentage of GDP, and finally ide, the average share of foreign direct investment. The data used in this study are secondary in nature. They come primarily from the World Bank database and cover the period 1995–2025. Given the lack of sufficient data on Guinea-Bissau, the study will consider the WAEMU zone excluding Guinea-Bissau. This method was chosen due to the dual temporal and special dimensions of our sample for analysis. To analyze the state of actual convergence among WAEMU countries, a specific model was required. However, this step, while necessary, is not sufficient to achieve our objective. Therefore, we will perform specification tests on our variables. Indeed, the next step consists of determining the order of integration of the different variables. This is done using Augmented Dickey-Fuller (ADF) tests (1979). Next, we will perform the Johansen integration test to look for the existence of integration relationships between the integrated series. In addition, we will conduct Fisher's exact test to determine the presence or absence of specific individual effects (fixed or random effects), which will allow us to accurately adopt an estimation method.

After the estimations, we will verify the specification of our results by performing tests designed for this purpose. We will present the details of the diagnostic and validity tests of our model below.

Model results and their interpretation

The results of the stationarity tests are shown in Table 1.

 

Table 1: Stationarity Test Results on the Series

Unit Root Tests
Variables	LLC-Stat	IPS-Stat	ADF-Fisher	PP-Fisher	Prob	Order of integration
PIB	-2.760	-3.303	35.330	60.263	0.000	I(1)
IDE	-4.615	-6.971	66.804	182.717	0.000	I(1)
FBCF	-4.989	_	44.473	97.057	0.000	I(1)
DP	-4.561	-4.228	40.709	63.049	0.000	I(1)
TSS	-5.024	-4.032	37.378	46.700	0.000	I(0)

Source: World Bank data, 2024, Author's calculations

 

These stationarity tests are based on the Levin, Lin & Chu; Im, Pesaran and Shin; ADF-Fisher; and PP-Fisher tests. The variables GDP, FDI, GFCF, and DP show that all these variables are stationary in first differences except for TSS, which is stationary at levels. To determine the existence of a specific effect, we will perform several estimations: estimations of the fixed-effects model and the random-effects model.

 

Table 2. Results of the convergence equation estimation

GDP	Coef	Std.Err.	T-stat	P-value	[95% Conf	Interval]
FBCF	0.599	0.311	1.92	0.056	-0.0	1.216
IDE	12.865	1.227	10.48	0.000	10.438	15.292
DP	4.349	0.291	14.91	0.000	3.772	4.926
TSS	1.28	1.09	1.18	0.241	-8683	3.43
CONS	-6.74	1.65	-0.41	0.684	-3.94	2.59

 

Number of obs  =   144  between= 0.3562   F(4,134)   =  150.11

R-sq:    = 0.8175

Prob > F     =    0.0000                                      overall = 0.3572

Prob > F    =    0.0000                                       overall = 0.3572

Source: Authors' calculations

 

Having presented the various results from our estimations, we can now proceed to the interpretation. The results summarized in Table 1 allow us to conclude that all our variables are stationary in first differences. Since Table 1 shows the variables that exhibit a regularity process in past variables, we can therefore assume stability or invariance over time. To determine the existence of specific effects, we performed Fisher's exact test. If the null hypothesis is rejected, the model is either a fixed-effects or a random-effects model. The results of our test here allow us to conclude that fixed effects exist. This result is confirmed by the HAUSMAN test. Indeed, with a probability > Chi-square = 0.001, the null hypothesis, which states the absence of random effects, is accepted. In light of the statistical analyses provided after performing the various tests suggested by the panel data methodology, the retained model is the fixed-effects model, the results of which are recorded in Table 2. From this table, it appears that the regression coefficient R² has a value of 0.8175, or 81.75%, indicating that 81.75% of the changes in GDP are explained by our different variables.

Fisher's F-statistic, F(4,134) = 105.11, and a probability > F = 0.000, are also significant. Based on these results, all variables in our model, except for the secondary school enrolment rate (SER) and gross fixed capital formation (GFCF), are significant at the 5% level. After this statistical interpretation, we can easily move on to an economic interpretation. Economically, our estimates of the convergence equation show that the foreign direct investment (FDI) variable has a positive effect on GDP per capita in the WAEMU zone. Indeed, a 1-point increase in FDI leads to an improvement in the standard of living, as measured by GDP per capita of 12.86. This result remains consistent with theoretical analyses regarding the impact of FDI on economic growth. Like the previous variable, public spending (PS) has a positive impact on GDP per capita, and according to the values summarized in Table 2, a 1-point increase in PS results in a 4.34 increase in GDP. This result is consistent with economic theory, specifically Keynesian theory, which stipulates that an increase in government PS or government involvement in the economy will lead to an increase in the standard of living of economic agents through their income. As for gross fixed capital formation (GFCF), which approximates investment, it has a positive sign, a theoretically valid result, but remains insignificant for GDP per capita. The insignificance of this variable in the region could be explained by poor allocation or suboptimal choices of investment sectors and projects. It could also result from investments whose profitability is considered long-term or unprofitable, as was the case for the Ivorian economy with investments in the construction of sugar complexes. The secondary school enrolment rate (SER), commonly used in most studies, is an approximation of the labour force. Its sign is consistent with theoretical expectations, but it remains non-significant. Its coefficient of 1.28 means that a 1-point increase in the population with a secondary school education level will lead to a 1.28-point increase in GDP per capita. The non-significance of this variable in the WAEMU economies in general is justified both theoretically and empirically.

The age of the education system and training sectors in several countries within the region, as well as a mismatch between training and employment in these states, could explain the lack of significant improvement in well-being for this segment of the population. Our results, specifically the indicators for our various variables, allow us to support the conclusion that there is a general trend toward convergence with an improvement in living standards for different population groups within the WAEMU region. Despite this observation, we must nevertheless point out that this convergence remains driven by two leading countries (Côte d'Ivoire and Senegal). A real convergence exists among the WAEMU member states, especially in this decade. However, it could be described as relative, or even as a two-speed convergence driven by two states with sometimes significant disparities in living standards between them and the rest of the group. Given this observation, it seems important to formulate recommendations to make the state of convergence within the WAEMU zone more visible and consistent.

All these proposals, in light of the obtained results, are part of the plan to establish a broader, integrated, and stable African economic and monetary area. The idea here is to achieve the convergence of national economies. Recommendations include:

- giving greater consideration to structural variables in economic convergence objectives, given that these variables are generally delayed in their implementation and sometimes exhibit persistent disparities between member nations of the same sub-region;

- to ensure better monitoring of nominal convergence indicators (primarily economic policy convergence indicators) because, even though these are more effective than real convergence indicators, they have not yet facilitated the achievement of real convergence within the WAEMU zone;

- finally, to implement a regional solidarity policy that supports the poorest countries to enable them to improve their economic growth.

All these recommendations, to play their full role, must coexist with situations of political and social stability embodied by freedom of opinion and expression within democratically established political regimes. This perspective considers the future of the CFA zone in relation to the continent's short-, medium-, and long-term needs for performance and economic development. However, it would be worthwhile to extend this analysis further and examine additional factors accelerating economic convergence. Another equally important question is to study the mechanisms for meeting and transmitting the performance of convergence criteria to achieve economic development. In other words, how can the objectives of meeting convergence criteria and harmonizing economies be translated into objectives of economic growth and development?

Conclusion and recommendations

The West African Economic and Monetary Union (WAEMU) was created in a particularly difficult context following the devaluation of the CFA franc, as a continuation of the West African Economic and Monetary Union (WAEMU). Governing bodies and parliamentary and judicial oversight bodies were established. They were tasked with ensuring the harmonization of the legal environment for economic activity, the harmonization of the legal framework for public finances and tax legislation, and therefore the harmonization of economic and, above all, budgetary policies, accompanied by a dynamic program of sectoral policies within a unified economic space, combined with a strengthening of the macroeconomic framework. At the end of this study, we observe that the WAEMU countries have generally begun a process of convergence. From the above, we note that genuine convergence within the WAEMU countries could be an effective means of combating poverty and further strengthening the integration process in the region. This convergence will also help to address the economic and financial difficulties faced by the countries in the zone. To achieve this, the WAEMU countries must implement policies aimed at minimizing socio-political conflicts, which hinder the growth process. They must strengthen efforts to harmonize economic policies. They must further reassure foreign investors, whose effective and substantial participation could boost growth in WAEMU countries. However, for greater effectiveness, it will be necessary to theoretically examine the link between the variables that significantly impact GDP per capita growth, in order to better assess their impact on GDP growth and therefore on convergence. This trend is more noticeable with regard to real GDP per capita growth rates and much less pronounced with regard to structural variables.

About the authors

K. T. Bakayoko

Peleforo Gon Coulibaly University of Korhogo

Author for correspondence.
Email: kcoulibalye2014@gmail.com

Lecturer

Côte d'Ivoire, Côte d’Ivoire, Korhogo

K. C. N’Goran

Peleforo Gon Coulibaly University of Korhogo

Email: nkoffiagos2012@gmail.com

Senior Lecturer

Côte d'Ivoire, Côte d’Ivoire, Korhogo

References

Supplementary files