CAUSALITY TESTING AND WAGNER’S LAW: THE CASE OF SRI LANKA

Abstract

This paper presents an empirical investigation into the validity of Wagner’s Law for Sri Lanka over the period 1959-2010. The research methodology employed includes testing for unit root, with the Augmented Dickey-Fuller (ADF) test, the use of a Vector Autoregression (VAR) model for the implementation of the Granger causality test, and cointegration tests according to Johansen-Juselious. The cointegration tests indicate that there is a long run relationship between public expenditure (TE) and Gross Domestic Product (GDP), and the ratio of total government (public) expenditure to gross domestic product (TE/GDP) and GDP (First and Six version of Wagner’s Law). Both eigenvalue and trace tests indicate that there is one cointegrating vector. Although the results reported herein do not reveal uniformity among the six versions of Wagner’s Law, the results show an apparent prevalence of the direction of causality from growth of GDP to public expenditure. For the first three versions of Wagner’s Law and the fifth version appear that Granger- causality runs one-way from GDP to TE, GDP to Total Consumption Expenditure (TCE), per capita gross domestic product (GDP/POP) to TE, and (GDP/POP) to per capita government expenditure (TE/POP), respectively. According to empirical findings of this study, it is possible to say that the growth of public expenditure in Sri Lanka is depended on and determined by economic growth as Wagner’s Law.