Do Governance Indicators Matter for Economic Growth? The Case of Sri Lanka

Abstract

This study attempts to identify the impact of governance indicators on economic growth using time series data for Sri Lanka from 1996 to 2016 published by the World Bank. The Phillips–Perron (PP) unit root test confirmed that all the variables are integrated in order one and suggested the use of cointegration technique to identify the long-run relationship between the variables. All the lag length selection criteria except Schwarz Information Criterion (SIC) advocated the use of one lag as an optimal lag length for this study. Johansen cointegration method detected three cointegrating relationships among the variables. Further, this technique identified a significant and positive relationship between government effectiveness (GE) and gross domestic product per capita (GDPPC) in the long run. This result is in contrast to all the three traditional approaches, such as correlation test, scatter plot and ordinary least squared (OLS), in which they do not identify any clear relationship between them. Moreover, Johansen test found a negative and statistically significant link between political stability and absence of violence (PSAV) and GDPPC in the long run, while all three traditional approaches identified a positive correlation between them. The findings of this study indicate a negative association between rule of law (ROL) and GDPPC in the long run, which coincides with theory, some of the empirical studies and with findings of all three traditional approaches used in this study. Even though OLS did not identify a significant relationship between control of corruption (COC) and GDPPC, Johansen test, correlation test and scatter plot detected a significant and negative correlation between them in the long run as expected by the theoretical evidence. Granger’s causality test identified the bidirectional causality between GE and ROL and unidirectional causality between ROL and COC. However, relationship between governance variables and GDPPC vary based on the estimation methods. These findings suggest that the policymakers need to take considerable attention on the above when they formulate and implement policy to improve GE.