The effect of government spending innovations on the Ethiopian economy

Purpose The purpose of this paper is to estimate the size of government spending components’ multipliers for the Ethiopian economy over the sample period of 2001Q1 up to 2017Q4. Design/methodology/approach The effects of government spending are analyzed by applying short-run contemporaneous restrictions for the identification of shocks in an SVAR in order to estimate multipliers for the small open economy. Accordingly, recursive identification scheme is used in this study. Findings From the impulse response functions, the authors found that aggregate government spending is less effective in stimulating the economy for the study period as evidenced by almost zero multipliers. This can be due to many structural and conjunctural factors that tend to lower the multiplier effects. At a disaggregate level, real GDP responds negatively to capital spending while its effect on recurrent spending is positive and insignificant on impact. The variation to real GDP is best explained by the variation in capital spending as compared to recurrent spending. Originality/value Though almost none in number, little research has been conducted in Ethiopia related to the effect of government spending shock on output. But this research deviates from the previous study by introducing a new methodology which is SVAR with cholesky decomposition. The previous study, however, used Bayesian VAR. Besides to that, using cholesky identification scheme, government spending is decomposed in to recurrent and capital spending to see the effect of government spending components on output and government spending multipliers are also computed both at an aggregate and disaggregate level.