Fertility differentials as evidence of incipient fertility decline in a developing country

Abstract Age data from the 1960 and earlier censuses of Ghana allow the construction of child-woman ratios which appear to indicate the existence of a substantial urban-rural fertility differential. Plausible assumptions of urban-rural mortality differentials increase the apparent fertility differential. In this paper recently published data for Statistical Areas in the country's larger towns are used to demonstrate that one explanation for the fertility differential is almost certainly the enumeration of some females in the towns, while one or more of their surviving children were enumerated outside. Nevertheless, in 1960 the four largest towns exhibited birth levels which are likely to have been about 11% below those of the population in the surrounding regions. Roughly half the differential can be attributed to a general urban-rural differential and half to socio-economic differentials within the towns. It is shown that most fertility reduction within the towns may be explained by delayed female marriage, and that such delay is associated with extended education. It is also shown that amongst the higher socio-economic status groups a small part of the reduction can probably be attributed to the prevention of pregnancy within marriage, and that the making of such attempts is positively associated with extended education, urban birth, participation in first and monogamous marriages, Protestantism, and the holding of views about the harmful effect of high population growth rates on attempts to raise living standards. It is argued that these fertility differentials are evidence of some fertility decline among key groups in the population and that such declines are likely to become more widespread.