Birth intervals, postponement, and fertility decline in Africa: A new type of transition?

Abstract

We investigated birth-interval dynamics in 24 African countries using data from 76 Demographic and Health Surveys conducted since 1986. Controlling for selection bias in the birth-history data using the Brass-Juárez method and regression models produced almost identical results. Birth intervals have lengthened in every country examined. This analysis uncovered a distinctive and previously undocumented pattern of childbearing that is prevalent across sub-Saharan Africa. After allowing for time trends in birth-interval length, the lengthening of birth intervals in almost every country varies little by women's age or parity. Moreover, in several countries, birth intervals are now too long to be explicable by birth spacing contingent on the age of women's youngest child. Rather, women are postponing births for other reasons. These findings offer empirical support for the idea that the fertility transition in sub-Saharan Africa is following a different pattern from that observed elsewhere.

Contraceptive use and lengthening birth intervals in rural and urban Eastern Africa

“…The trend is most pronounced in Southern Africa and the largest increases have occurred in South Africa, where the median birth interval is thought to have risen to six years by the late 1990s. In Namibia, Zimbabwe, and Lesotho, intervals stand at approximately 45 months (Moultrie, Sayi, and Timaeus 2012). However, this trend is not restricted to Southern Africa: in Ghana the median birth interval has risen to 40 months.…”

“…In addition to the large literature that examines declines in total family size (Caldwell and Caldwell 2002;Caldwell, Orubuloye, and Caldwell 1992;Cleland, Onuoha, and Timaeus 1994;Cleland 2001;Cohen 1993;Garenne 2008;Garenne and Joseph 2002;Kirk and Pillet 1998), there is growing evidence of a more novel form of fertility change: the lengthening of birth intervals. It has been shown that birth intervals are now lengthening in a large number of populations across sub-Saharan Africa (Bongaarts and Casterline 2012;Moultrie, Sayi, and Timaeus 2012). During the 1960s and 1970s, median birth intervals across the region stood between 27 and 33 months, which is to be expected of populations where lengthy periods of breastfeeding are the norm.…”

“…A number of approaches have been proposed to examine trends in birth intervals that are designed to control for the effects of selectivity and censoring. These include the life-table method (Hobcraft and McDonald 1984;Hobcraft and Rodríguez 1992;Rodriguez and Hobcraft 1980), a paired-cohort truncation approach (Brass and Juárez 1983), and the regression method approach (Moultrie, Sayi, and Timaeus 2012;Towriss and Timaeus 2017). This study adopts the third, regression model approach, since it is more straightforward and flexible than the Brass and Juárez approach and is statistically more powerful than the life-table approach.…”

“…However, this trend is not restricted to Southern Africa: in Ghana the median birth interval has risen to 40 months. Moreover, lengthening intervals have not only been observed in countries where fertility decline is well established, but also characterize -although to a lesser extent -countries with higher total fertility and lower contraceptive prevalence (Moultrie, Sayi, and Timaeus 2012).…”

“…Most countries in the region experienced a 10%-50% decline in fertility between 1960-1964 and 2000-2004, compared with West and Central Africa where a number of populations experienced an increase in fertility in the same time frame (Tabutin and Schoumaker 2004). Since longer birth intervals are associated (although not strongly) with lower levels of fertility (Bongaarts and Casterline 2012;Moultrie, Sayi, and Timaeus 2012), it was necessary to carry out this study using data from populations that have witnessed a certain degree of fertility change. Second, the existing literature on birth interval lengthening has tended to focus to populations within Southern Africa; this study aims to contribute to this body of evidence through an examination of the phenomenon in Eastern Africa.…”