This paper makes a modest attempt to disentangle the causality between political and economic events within the context of the effects of the rate of economic growth on the probability of leaders' survival in Africa, variables that are significantly positively correlated in the data. However, the correlation cannot tell us whether it is the economic success that enhances the likelihood of survival, or whether giconomic growth is itself the outcome of the political events, through successful manipulation of the economy by the incumbent, or because economic growth is more rapid in the absence of the political turmoil that often precedes a change in leadership. Fluctuations in the world prices of African countries' exports provide an "instrument" that can help untangle the causality. Increases in commodity prices favor economic success, but are plausibly unaffected by political events in the countries themselves, so that the economic growth associated with commodity price booms is free from feedback from political events. Previous evidence that economic growth enhances political survival is confirmed with commodity-driven growth, thus providing (weak) evidence in favor of the economics to politics interpretation rather than the reverse.