Using Vector Auto regression to Analyse the Determinant(s) of Inflation in the Gambia

Abstract

This research document examines the strength of the causal relationship between inflation in The Gambia and a set of key macroeconomic variables, viz: budget deficit as a percentage of GDP, broad money and nominal exchange rates, etc., all in the form of quarterly time series data. These data series were tested for both Unit Root and for the optimal Lag Number. Granger causality test was also conducted to verify the magnitude and direction of this causal link. Furthermore I tested the effect of a set of third variable by using Block Exogeneity test. Forecast Error Variance Decompositions as well as Impulse Response analysis were also done to further examine the relative strength of the above causal relationship. The results reveal that inflation is not caused directly, in the short run, by any of the expected variables, viz: fiscal slippages, monetary growth or exchange rate depreciation but by another monetary factor, inflation (inertia) expectation engendered, perhaps, by poor crop harvest and low revenue from tourism sector, with all their negative ramifications on foreign reserves as well as on the importation and consumption of essential food commodities in The Gambia. In the long run, however, nominal exchange rate depreciation to the domestic currency was found to be a stronger determinant of inflation compared to all the rest of the variables in the model.