In this paper, we create the beginnings of a database on indicators of poverty and deprivation in six African countries over the last 100 years, with the intention of explaining the long-term trend of policy. We argue that an important key to those trends is provided by the decision during the first decades of the twentieth century to allow, or not to allow, the occupation of agricultural land by
... [Show full abstract] European settlers. If allowed (as in the white settler-controlled economies of South Africa, Zimbabwe and Kenya, where Africans were squeezed out of income-earning opportunities by competition and prohibition until the 1960s at least) the legacy, we show, was a lack of bargaining power by Africans in the labour market, a real wage static at the subsistence level until late in the century, a highly unequal income distribution, and a small asset base from which to reduce poverty during the liberalisations of the 1980s and 1990s. If forbidden, as in the peasant-export economies of Ghana and Uganda (and to some extent Ethiopia), the legacy was a floor under the labour market, from the 1920s on, which could serve as a basis for poverty-reducing reforms in later years. Thus the potential for poverty reduction during the current 'poverty reduction wave' was indeed determined by the historical inheritance of institutions and policies, in particular settlement policies; but in a quite different way from that adumbrated by the recent analysis of Acemoglu et al. (2001). In their analysis, colonies with a high density of European occupation in colonial times persistently generate higher growth potential in later years; we argue the reverse. Copyright © 2008 John Wiley & Sons, Ltd.