Acemoglu, Johnson and Robinson have dramatically challenged the tendency of economists to confine their empirical search for the causes of economic growth to the recent past. They argue that the kind of institutions established by European colonialists, either protecting private property or extracting rents, resulted in the poorer parts of the pre-colonial world becoming some of the richest economies of today; while transforming some of the more prosperous parts of the non-European world of 1500 into the poorest economies today. This view has been further elaborated for Africa by Nunn, with reference to slave trading. Drawing on African and comparative economic historiography, the present paper endorses the importance of examining growth theories against long-term history: revealing relationships that recur because the situations are similar, as well as because of path dependence as such. But it also argues that the causal relationships involved are more differentiated than is recognised in AJR's formulations. By compressing different historical periods and paths, the 'reversal' thesis over-simplifies the causation. Relatively low labour productivity was a premise of the external slave trades; though the latter greatly reinforced the relative poverty of many Sub-Saharan economies. Again, it is important to distinguish settler and non-settler economies within colonial Africa itself. In the latter case it was in the interests of colonial regimes to support, rather than simply extract from, African economic enterprise. Finally, economic rent and economic growth have often been joint products, including in pre-colonial and colonial Africa; the kinds of institutions that favoured economic growth in certain historical contexts were not necessarily optimal for that purpose in others. AJR have done much to bring development economics and economic history together. The next step is a more flexible