An empirical analysis of Ethiopia's import demand behavior: Johansen co-integration and Granger causality approaches

Abstract

This study examines the long run and the short run determinants of import demand for Ethiopia using a time series data for the period 1970/71-2010/11. Both the simple descriptive analysis and the Johansen's cointegration approach are employed to see the impact of real GDP, domestic price level, foreign exchange reserves and exchange rate on the import demand of the nation. This study differs from other similar studies in Ethiopia for it employs Johansen co integration approach, stationary series, more variables and more recent observations. The quantitative results from cointegration and error correction specifications show that imports of the country are sensitive to changes in domestic output level and foreign exchange reserves both in the long run and in the short run though their estimated elasticity coefficients are smaller in the later case; and domestic price level and exchange rate are found to be statistically insignificant. While only foreign exchange reserves Granger cause import in the short run, all variables jointly Granger causes import in the long run. The estimated Vector Error Correction Model of import is stable over the sample period that it can be used for a policy purpose. The lower short run income elasticity of import shows the room available for import substitution industrialization strategy in Ethiopia and the higher long run income elasticity provides an evidence in favor of product diversification. Devaluation can also be made effective by supplementing it with import restriction schemes. Keywords: Import Demand, Johansen Approach, Granger Causality Approach, Model stability, Forecast Performance, Ethiopia