This report provides an overview of major fertility trends in sub-Saharan Africa in the second half of the 20th century. It also presents the proximate determinants (factors that have a direct mechanical effect on fertility levels) and the socioeconomic correlates of these trends.
Cohort and period fertility trends were constructed using World Fertility Survey (WFS) and Demographic and Health Surveys (DHS) data sets for 31 countries in sub-Saharan Africa. Cohort trends were derived from WFS and DHS data for some of the factors that affected fertility change: infertility, age at first marriage, education level, proportion Muslim, proportion Christian, proportion living in a polygynous union, and nutritional status. Period trends were derived for urbanization and income per capita from other sources.
Cohort fertility was higher among women born in 1950 than in those born in 1930 but tended to decline in women born in later years. Changes in cohort fertility levels were small, on average. The mean number of children ever born to a woman by age 40 increased from 5.9 in women born in 1930 to 6.2 in women born in 1950 and decreased to 5.6 in women born in 1970. In most cases, the increase in cohort fertility was apparently due to a decline in infertility and, to a lesser extent, decreasing age at first marriage, which was associated with the spread of monotheist religions in the first half of the 20th century. Nutritional status did not have any identifiable impact on cohort fertility.
Like cohort fertility, period fertility tended to rise from 1950 to 1975 and then fall until 2000 or later. On average, for the countries investigated, the total fertility rate at age 40 increased from 5.3 children per woman in 1950 to 6.2 in 1975, then declined to 4.9 in 2000. The decline in period fertility appeared to be due primarily to increasing contraceptive use and, to a lesser extent, rising age at first marriage and increasing urbanization. A regression model of the explanatory variables indicated that 37 percent of the decline was attributable to increased contraceptive use, 24 percent to decreased age at first marriage, and 16 percent to increased urbanization. These three variables correlated with level of education and, to a lesser extent, income per capita.
The dynamics of the fertility decline were different in urban and rural areas. On average for the countries investigated, the trends in urban and rural areas started to split in approximately 1960. The date of onset of the fertility decline varied greatly by region and country, ranging from the early 1960s in the first urban areas to the late 1990s in the last rural areas. A few rural communities had not started the transition at the time covered by the last available survey.
The speed of the fertility decline, approximately 1 child per decade, also varied markedly among countries, from 1.5 children per decade to less than 0.5 children per decade. In addition, a stall in fertility decline occurred in six of the countries investigated (Ghana, Kenya, Madagascar [urban areas], Nigeria, Rwanda-rural, Tanzania [rural areas]); in five of these countries, this stall occurred in 1995-2005.
The pattern of the fertility decline in sub-Saharan Africa did not appear to be very different from that of many other countries in the world. However, the fertility decline in sub-Saharan Africa seems to have been somewhat more influenced by changing nuptiality patterns than elsewhere, and its relationship with socioeconomic correlates was somewhat less influenced by income levels and trends than other countries.
The appendices present a detailed analysis of fertility trends by country, with information on trends in urban and rural areas, premarital and marital fertility, and periods of monotonic changes.
Key Words: Fertility transition, fertility decline, fertility increase, fertility stall, infertility, contraception, age at marriage, proximate determinants, socioeconomic correlates, sub-Saharan Africa