the relationship between electricity consumption and economic growth in kenya: 1970-2004

Abstract

This study set out to evaluate the probable relationship between electricity consumption and economic growth in Kenya and to determine whether there is either short run or long run equilibrium relationship between these two variables. To achieve the above objectives both granger causality analysis and Error Correction Modelling (ECM) were applied to the data series for electricity consumption and economic growth. The time series data for both annual electricity consumption and GDP was obtained from government publications for the period 1970-2004. Prior to testing for causality, analysis of stationarity for both variables revealed that the raw data sets were non-stationary and only when differenced and subjected to the same tests became stationary. Further diagnostic tests for regression residuals including residual and stability tests established that both electricity consumption and economic growth data series exhibit normal distribution pattern. The series also did not show any serial correlation neither did they show autoregressive conditional heteroscedasticity effects nor white noise and had very stable regressions. Therefore, the subsequent regressions conform to the OLS assumption of consistency and efficiency. Estimation results from the granger causality analysis indicated that there is a bidirectional relationship running from electricity consumption to real GDP and vice versa. Thus an increase in electricity consumption would raise real GDP while improved economic growth would trigger higher electricity consumption.