Essays on the impact of workers' remittances on key macroeconomic variables in Sri Lanka

Abstract

In recent years, researchers have increasingly paid great attention to investigate the impact of workers' remittances at the macro-level. Sri Lanka gains a huge income from these migrant workers' remittances every year. The migrant workers' remittances in Sri Lanka are identified as the 1st external source of income compared to others. However, researchers have not examined in detail how workers' remittances influence the economic development of Sri Lanka through the variables of consumption expenditure, human capital formation, financial development and economic growth at the aggregate level. In developing countries, these are the research topics of current interest among the scholars and academia. If the influence of workers' remittances on such variables is investigated at the aggregate level it will be easy to measure scientifically to what extent workers' remittances have a potential role to play in determining the economic development of Sri Lanka. Due to the lack of research in this area, this study is undertaken with the overall objective to test the impact of workers' remittances on Sri Lankan economy for the period of 1975-2017. In order to achieve the overall objective, this study established four self-contained essays. Each of them has a separate objective. In this study, analytical techniques such as exploratory data analysis (EDA) and inferential data analysis (IDA) were employed to test the objective of each essay. Further, Dynamic Factor Analysis was used to formulate an index for the variables of financial development (FD), an d human capital formation (HCF). The EDA used the tools of scatter plots and confidence ellipse with Kernel fit. As for the IDA, Unit root test, ARDL Bounds cointegration test, Pairwise Granger causality test, and Impulse Response Function (IRF ) analyses are included .

The test results of EDA show that workers' remittances have a positive relationship with the dependent variables given in each essay of this study. The unit root test results given in the 1st, 3rd and 4th essays of this study confirm that the variables are stationary at their 1st difference, while the 2nd essay of this study indicates that the variables are in mixed order. The ARDL Bounds cointegration test results indicate that workers' remittances have a long-run relationship with all the dependent variables given in each essay of this study. The diagnostic test results confirm that the estimated models given in each essay are robust. The estimated long-run coefficients of workers'remittances in each essay indicate that there is the positive relationship between workers' remittances an d the dependent variables. The estimated coefficient of error correction term illustrates that the response variable given in each essay of this study moves towards the equilibrium path, respectively. The Granger causality test results given in each essay of this study indicate that workers' remittances Granger causes all response variables, whereas the estimated coefficient of err or correction term given in each essay of this study explains that there are a long-run causal relationship among variables given in each essay of this study. The IRF test results indicate that a positive shock to workers' remittances has an immediate significant positive impact on the dependent variables given in each essay of this study. Therefore, the overall conclusion of this study is that workers' remittances in Sri Lanka have a vital role in determining economic development through the macroeconomic key variables of consumption expenditures, hum an capital formation, financial development, and economic growth at the aggregate level. The findings of this study, it is hoped, will provide the policymakers with valuable insights into the nature of the economic development of Sri Lanka. Hence, the findings of this study could be useful for policymakers when they design development policies.