Money and Output in Tanzania: A Test for Causality

Abstract

This study has investigated empirically the causal relationship between money and output in Tanzania for the period 1986 to 2018. A VECM was estimated in log-difference and in log-level and Granger causality test undertaken by using annual time series data to test a null hypothesis that money does not Granger cause output either way. The log-difference results rejected the null that money Granger cause output in favour of the alternative hypothesis that output Granger cause money and the effect was unidirectional. The results in this case suggest money supply is endogenous such that monetary policy cannot directly be used in stabilizing the economy over the short-run. Instead the government should rely on fiscal policy rather than monetary policy to attain macroeconomic objectives in general and price stability in particular. Robustness test results from estimation the log-level results suggested the causality was unidirectional from money to output, implying money is exogenously determined and could be controlled by the monetary authority to achieve macroeconomic objectives, price stability in particular. The differing results demand for further empirical tests.