Why should there be a historical difference between developed and developing world in the time profile of birth rates? The reasons are complex, and there appears to be no universally valid explanation for all countries. In Korea, fertility rates fell by 44 per cent between 1965 and 1982 as women married later and later. The 1982 level has continued to fall, but at a much slower rate because the average age of women at marriage is tending towards a limit. The potential for further fertility decline is thus constrained. In Costa Rica, fertility rates fell by 48 per cent between 1965 and 1987, and are expected to fall by a further 25 per cent by 2000. But there are signs that there is some lower limit to the desired family size.

This suggests that fertility rates will not decline continuously, as they did in the developed world. The experience with family planning varies markedly. In India, the total fertility rate has fallen continuously, but contraceptive practice varies according to female literacy and the vigour with which policy is pursued. Some insight into the reasons for the slower-than-expected decline in fertility in the developing world can be obtained by looking at the underlying factors involved in the African experience.

Clearly, the social, economic and cultural factors underlying decisions about family size are complex. They are also likely to vary from society to society. In general, however, the private benefit/cost ratio favours large families at low levels of education, high levels of rural dependence and low levels of state-provided social security. As these change, the ratio falls and favours smaller families. The obvious problem is that these factors are changing all too slowly relative to the factors reducing the death rate.