Growth threshold-effect on renewable energy consumption in major oil-producing countries in sub-Saharan Africa: a dynamic panel threshold regression estimation

Abstract

Purpose This study aims to explore the relationship between the growth threshold effect on renewable energy consumption (REC) in the major oil-producing countries in sub-Saharan Africa (SSA) over the period 1990–2018. Design/methodology/approach This article used a dynamic panel threshold regression model introduced by Hansen (1996, 1999 and 2000) threshold (TR) models. The procedure is achieved using 5,000 bootstrapping replications and the grid search to obtain the asymptotic distribution and p-values. For the long-run relationship among our variables, the author followed the process in Pesaran et al. (1999) pooled mean group (PMG) for heterogeneous panels. Furthermore, for the robustness of our empirical results due to the sensitivity of the results to outliers, the author used the approach by Cook (1979) distance measure. The author applied quantile (QR) regression to explore the distribution of dependent variables following Bassett and Koenker (1982) and Koenker and Bassett (1978) approaches. Findings The results from the threshold effect test and threshold regression revealed a significant single threshold effect of growth level on REC. Furthermore, the result from the PMG estimation showed the growth of the variable, energy intensity, consumer prices and CO2 emissions play a significant role in REC in major oil-producing countries in SSA. The growth threshold estimation results indicated one significant threshold value of 1.013% at one period lagged of real growth. The outlier’s sensitivity detention greatly influenced our empirical results. Originality/value The article filled the literature gap by applying a combined measure that is robustness to detect outliers in the data, which none of the studies in the literature addresses hitherto. Further, the article extends the quantile regression to growth – REC literature.