In the early 1990s, Caldwell and others suggested a hypothesis that an African fertility decline would be characterised by declines in fertility at all ages and parities simultaneously, unlike that observed elsewhere in the developing world. Earlier research has documented the development of exceedingly long median birth intervals in South Africa, and suggested that the combination of political, economic and institutional factors associated with Africans’ lives under apartheid were responsible
for that pattern. At the time, tentative evidence from Zimbabwe suggested that similar features may be identifiable there, and that - in fact - the ‘South African’ pattern was simply a harbinger of demographic change elsewhere in the region. These ideas are explored, with particular emphasis placed on the measurement and analysis of patterns of childbearing in eight countries using DHS
data. Three distinct regional patterns of childbearing are identified, all of which are fundamentally different from those observed in a South East Asian country. This provides strong empirical support for the Caldwell hypothesis