Arouri, Mohamed El Hedi Ben Youssef, Adel M'henni, Hatem Rault, Christophe
Year of Publication:
IZA Discussion Papers No. 8296
Using a bootstrap panel analysis that allows for cross-country dependence, without requiring the use of pre-tests for a unit root, we study the causality links between energy use and employment for a sample of 16 African countries over the 1991-2010 period (according to availability of countries' data) in a panel Vector AutoRegressive model. Our results indicate that employment and energy use are strongly linked in Africa. Unidirectional causality from employment to energy use in Tunisia, Cameroun, Zambia and Ethiopia is found. A unidirectional causality from energy use to employment is found in DRC and Egypt. We found also bidirectional causality for Algeria, Benin, Kenya, Mozambique and Tanzania). However, our estimates did not indicate any causality in Big African players like South Africa, Nigeria, Morocco, Ghana and Senegal.