The Relationship Between Energy Consumption and Sectoral Output in Bangladesh: An Empirical Analysis

Abstract

Listen

Translate article

The relationship between economic growth and energy is a widely researched topic in the world of economics and there are quite many studies on the relationship between energy and sectoral output; however, to the best of our knowledge there have been none in the context of Bangladesh. This paper examines the causal relationship between energy consumption and output growth of Bangladesh both at aggregate and sectoral level using the annual data from 1980 to 2014. For the purpose of our study we have used the GDP at constant local currency as an indicator of economic growth while energy use has been used as a proxy for energy consumption. These variables have been taken from World Development Indicator (WDI) 2015 published by World Bank and EViews 7.0 has been used to run the tests in this study. To test our data for stationarity, we have run two tests which are the Augmented Dickey-Fuller (ADF) test and the Phillips-Perron (PP) test. Then to test for cointegration and causality between the variables, we used the Johansen cointegration method and the Granger Causality tests respectively. We find the variables to be stationary after running both the Augmented Dickey-Fuller (ADF) test and Phillips-Perron (PP) test. Using the Johansen cointegration method, we find that all the variables are cointegrated. Granger Causality tests reveal that at the aggregate level there is unidirectional causality running from GDP to energy consumption. Similarly, at the sectoral levels, there is unidirectional causality running from agricultural, industrial and services sector output to energy consumption. However, Vector Error Correction Model (VECM) test observes no causality between energy consumption and output at the macro level and at the sectoral level in the short run. Thus we can conclude that GDP or sectoral output is not energy dependent in Bangladesh and our findings are also consistent with the conservation hypothesis. Moreover, our short run results are consistent with the neutrality hypothesis. This further strengthens our conclusion that the economic growth of Bangladesh is not energy dependent. These results have important implications for policy as energy conservation policies can be applied in Bangladesh without hurting growth. Furthermore, it is evident that implementing subsidies in the energy sector will not contribute in achieving growth and cost reflective pricing policy would be more effective in Bangladesh. Nevertheless, the government of Bangladesh should invest in efficient energy production and utilize the various energy sources available to Bangladesh.