Testing the Validity of Wagner’s Law of Government Size in Tanzania, 1966–2011: A Cointegration and Causality Analysis

Abstract

This paper presents an empirical analysis of the validity of the Wagner’s Law in the context of Tanzania for the period spanning 1966-2012. It analyses the long-run and causal relationship between public expenditure and economic growth. The Augmented Dickey Fuller test was first applied to analyse the stationarity properties of the data. With regard to this test all the variables were found to be non stationary at level, but stationary after the first difference. The Johansen test of co-integration was further deployed in which we found no co-integrating vectors or no long-run relationship among the variables. Besides, the study applied Granger causality test to test the validity of the Law as it requires existence of unidirectional causality running from GDP to public expenditure. By applying this test we noted no causality running from economic growth to public expenditure and vice versa for all six versions of the law analysed in this paper. Therefore, both Johansen co-integration and Granger causality tests revealed no empirical evidence to support validity of the Law in Tanzania. This implies that that public expenditure plays no significant role and cannot be used as a policy instruments in promoting economic growth in Tanzania.