The impact of currency devaluation on non-oil exports in Africa

Aim: This study assesses how devaluation in official exchange rate and change in relative prices influenced non-oil export in African countries for the period of 30 years (1991-2020) in 11 African countries (Nigeria, Burkina Faso, Burundi, Eswatini, Mauritius, Mozambique, Niger, Rwanda, Sierra Leone, Tanzania and Uganda). Design/Research methods: This study utilized different estimators namely Mean Group, Pooled Mean Group as well as dynamic panel GMM methods. The major advantage of the MG estimator is that it is reliably efficient even in presence of weak cross-sectional dependence of the errors by estimating separate regressions to calculate coefficient means. Moreover, it applicability knows no bounds even when estimator for each individual country is weakly cross correlated. With a PMG estimator, a large scale individual panel heterogeneity in short-run responses is accommodated given homogenous long-run relations across countries. Findings: The results of the panel co-integration suggest a long-run equilibrating relation amongst the variables in the study. This was validated on the basis of absolute t-value of 5.0781under the t-bound. Our results for both MG and PMG estimators show significant negative devaluation and relative prices effects on non-oil exports in 11 African countries. The dynamic panel GMM results are robust and in agreement with the estimates of MG and PMG. From the results of cross-sectional analysis by country, results for countries revealed exchange rate devaluation had negative and significant impact on non-oil exports. Consequently, depreciation of the exchange rate has a short-run adverse effect on non-oil export due to high inelastic import dependence. Similarly, with exemption of Rwanda, and South Africa, the relative price effect was negatively significant for every other country in the study. Originality: The originality is based on fact that the paper establishes both static and dynamic responses of non-oil export to devaluation in official exchange rate, relative prices, and foreign capital from trading partners in 11 African countries. Limitations: It would be desirable to study 30 countries in Africa. We could not proceed with all countries due to inaccessibility of relevant data. Hence, caution should be taken in generalizing our findings.