Real exchange rate misalignment and agricultural trade: a panel analysis

Abstract

Literature focusing on real exchange rate (RER) misalignment and agricultural trade is inconclusive and far from convincing. Against this background, the study sought to examine the impact of RER misalignment on agricultural exports using a panel data set of 20 countries covering the period 1983-2006. The equilibrium RER is first estimated using the Dynamic Ordinary Least Squares technique and deviations from this equilibrium level are termed RER misalignment. The calculated misalignment measure is then inserted in the export model as a covariate. Pre-estimation checks suggested that export determinants follow a non-stationary process (a random walk with a drift) and are cointegrated. Using the Feasible Generalized Least Squares technique to control for heteroscedasticity and contemporaneous correlation across countries, the study indicates that RER misalignment has a detrimental effect on agricultural exports and the effect is non-linear. This result is robust to estimation by the instrumental variable technique to control for endogeneity and simultaneity of the exchange rate. Against this background, policymakers in these countries should work to eliminate currency misalignment through maintaining clean floats otherwise pegged rates should remain in tandem with the equilibrium level suggested by their long term fundamentals.