Exports and Economic Growth in Sri Lanka: A Cointeration Analysis And Causality Testing

Abstract

Given the increasing trend of globalization of the world economy, both developed and devel. Oping countries are actively moving towards free trade with other countries to achieve rapid growth and development. Sri Lankan economy is not exceptional involving in this debate. Following the economic liberalization in 1977, the export sector became the growth-oriented sector in the economy. This study analyzes the causal relationship between exports and economic growth in Sri Lanka under the post liberalization period. To investigate from an econometric perspective, this study applied Johansen and Juselius cointegration method- ology and Granger causality analysis. The annual time series data for the period 1977 to 2009 have been used and were drawn from various annual reports of Central Bank of Sri Lanka. The stationary properties of all the time series variables were tested using Aug- mented Dickey-Fuller (ADF) test, based on the Akaike information criterion (AIC) at both levels and first-differences. The results of unit root test indicate that all the variables were non-stationary at levels but they were stationary at first differences. Once the stationarity of the variables are established, these variables were included in the cointegration technique developed by Johanson and Juseliues to estimate the long run equilibrium relationship among the variables. According to the Johansen Maximum likelihood test, the computed Trace statistic, maximum Eigen statistic and their corresponding critical statistic indicates that the null hypothesis of “no cointegration” can be rejected at five percent level of significance.This result supports the presence of long run relationship among the variables. Moreover, the Granger causality test indicates while there was a significant causality from exports to economic growth, the causality from economic growth to exports was insignificant. These results support the existence of Export-Led Growth (ELG) hypothesis in Sri Lanka