Effects of external debt on economic growth in Sub-Saharan Africa : cointegration and causality evidence for Sierra Leone

Abstract

This prime objective of this study is to investigate the effect of external debt on economic growth in Sierra Leone using macroeconomic data from 1970 - 2001. In achieving it, this study estimates vector autoregression (VAR) models in the form of bivariate and multivariate (VAR) models. In the multivariate VAR framework, the model incorporates other relevant variables such as the terms of trade and gross capital formation since they are the possible variables through which indebtedness can work for or against economic growth.

Empirical results from the bivariate model establish a long-run statistical relationship between external debt and GDP and a causal (Granger) relationship between them. In addition, in the multivariate model, there is sufficient evidence of the existence of one cointegrating vector, implying a long-run relationship linking GDP with the debt stock via terms of trade and gross capital formation. We found that in Sierra Leone, increases in the debt stock had a positive effect on GDP, and the effect is mediated through the other macro variables of the system. This result is surprising based on the structure of the Sub-Saharan African economies where external debt had in the past have negative influence on their growth prospect. Nonetheless, the significant negative influence of the terms of trade and the positive influence of capital formation suggests a possible macroeconomic route through which external debt influenced GDP. The Granger causality test indicates a uni-directional relationship from economic growth to external debt. The forecast error variance decomposition analysis provides a dynamic picture of the interplay of the effects of shocks on these different variables, and produces results generally consistent with those of the Granger causality tests.