Objectives: The relationship between government expenditure&government revenue has been an important topic in public finance. The purpose of this paper is to investigate causal relationship between these two important macroeconomic variables over the period of study 1963-2007 and to examine whether the dataset exhibit any structural breaks during the whole time span and such breaks have an impact on the direction of relationship between government expenditure&government revenue.
Methods: The present study involves use of battery of test for stationarity like ADF, PP and KPSS. The Engel-Granger cointegration method is used to test the long-run relationship between the variables concerned. The Vector Error Correction Model (VECM) is applied to examine whether the long-run relationship is stable along with causality test in the short run. Unresticted Vector Autoregressive Model (UVAR) is used for Granger Causality between the variables concerned in the long run. In order to test for a structural break in the data set we use the Chow test.
Results: The empirical results suggest that for Malaysia, fiscal neutrality hypothesis exists over the period of study 1963-2007 and government revenue&government expenditure have no long run equilibrium relationship. The dataset exhibits that possible structural changes have been occurred in the year 1983 and 1997 during the whole time span and consequently the relationship between revenue and expenditure has been undergoing a significant changes under different sub-periods i.e 1963-1980 and 1981-1997&1998-2007. The statistical results reveal an important fact that unidirectional causality is observed in the short-run during the sub-period 1963-1980. In the longrun during this sub-period, fiscal neutrality principle is observed. During the second sub-period 1981-1997, a long-run relationship exists in Malaysia but no-way causal relationship exists in the short-run during this sub-period. This subperiod 1981-1997 is marked by the presence of unidirectional causality running from expenditure to revenue in the long-run. The sub-period 1998-2007 is witnessed by the presence of unidirectional causality running from expenditure to revenue in the long-run.
Conclusion: The estimated causal relationship with historical dataset may not provide reliable guideline for making policy decisions. On the other hand, the estimated causal relationships under different sub-periods may provide reliable guidelines for decision making process.