Effects of energy consumption on per worker output: A study of Kenya and South Africa

The paper investigates the long-run cointegration relationship and energy elasticities for Kenya and South Africa over the periods 1978–2009 and 1971–2009, respectively, using the ARDL procedure developed by Pesaran et al. (2001) with recomputed critical bounds from Narayan (2005) and the Solow (1956) framework extended by Rao (2010). We also conduct the (Toda and Yamamoto (1995) test for Granger non-causality. The regression results show that short-run and long-run energy elasticities are 0.50 and 1.71, respectively for Kenya and 0.17 and 0.34, respectively for South Africa. The causality results indicate a unidirectional Granger causality running from capital per worker and energy per capita to output per worker for both countries. Moreover, in Kenya, we detect a strong unidirectional causality: (a) on output from joint consideration of capital stock and energy; and (b) on capital stock from joint consideration of energy and output. In South Africa, the joint causations are neutral. Hence, while energy and capital stock spurs growth in both countries, Kenya has a greater potential to harness growth and capital productivity via joint consideration of energy with capital and output, respectively.