Theorising government expenditure growth in Malaysia, 1961-1990 : an econometrics analysis of Wagner's Law, Keynesian relation and Peacock-Wiseman hypothesis
posted on 2014-12-15, 10:36authored byKu'Azam Tuan. Lonik
We applied cointegration, Granger-causality and Error Correction Mechanism [ECM] model to test for the Wagner's Law and Keynesian relation in an effort to explain the government expenditure growth in Malaysia for the period 1961-1990. We defined Wagner's Law following Musgrave [1969], Gupta [1967], Goffman [1968] and Mann's [1980] definitions. We modified Musgrave definition by excluding transfer payments from the total government expenditure to test the significance of transfer payments. We tested the Keynesian relation by reversing the Gupta's definition to see the effect of government expenditure on GNP. Following Diamond's [1977] interpretation of Peacock and Wiseman Hypothesis as a theory of structural break, we employed Perron's test for structural break to test for Peacock-Wiseman Hypothesis by considering the May 13, 1969 racial conflict as a form of social upheaval.;Following Nelson and Plosser [1982] and as a pre-requisite to cointegration, Granger-causality and ECM, we tested the data generating process to determine whether the time-series used in this study are generated by trend stationary [TS] or differenced stationary [DS] process. On discovering that the time-series are DS, we proceed by testing the unit root hypothesis using Dickey and Pantula [1985] procedure.;On Wagner's Law, we discovered that a) the variables from Musgrave, modified-Musgrave and Mann's definition are NOT cointegrated, b) using differenced variables, we find no Granger-causality to support Wagner's Law which is a sharp contrast when we tested the relationship using the level of the variables and c) ECM test confirmed our finding in (b). We obtained the same result as (b) when we tested the Keynesian relation. Using Perron's procedure, we cannot trace a structural break in total government expenditure, GNP and ratio of government expenditure in GNP to verify the Peacock-Wiseman Hypothesis.