Approximately 20 years after the civil war in Mozambique, demographic characteristics of self-settled refugees of Mozambican origin in Agincourt are converging with those of their South African hosts. While the TFR in Mozambique itself has remained near 5 (24
), the Mozambican TFR in Agincourt was 2.6 in 2009, its lowest level to date. Both population groups now show similar fertility patterns, with a high proportion of first births in the 15–19 age range and delayed childbearing thereafter.
The findings of this study suggest adaptation of the Mozambican refugees in the AHDSS to the fertility patterns of their host community. Adaptation theory states that exposure to cultural norms and local costs of childbearing will lead migrants to change their fertility behaviour to converge with that of natives in the destination (25
). This appears to be the case, particularly through 2005 when the majority of the population of Mozambican women in Agincourt were former refugees. The fertility of more recent Mozambican migrants might additionally be suppressed due to the disruption caused by migration.
The adaptation of Mozambican refugees to the lower fertility regime in South Africa has important implications for many areas of sub-Saharan Africa hosting refugee populations. The adaptation of Mozambicans in South Africa is likely facilitated by a shared language and culture. Self-settled refugees are also probably more likely to be exposed to and adjust to the local norms of childbearing compared to refugees living in camps.
Access to contraception through the South African health system is a key component of the decrease in fertility of Mozambicans. Another important component is the improvement in socio-economic status partly attributable to access to education and host government social grants. Reducing the economic disadvantage of refugees and integrating refugees into local programmes and services encourages adaptation and can compensate for other factors that may otherwise increase the fertility of refugees such as poverty, lack of education, and lack of reproductive health services. Integration encourages adaptation and will likely benefit host communities by lowering the fertility of refugees.
Overall fertility decline in Agincourt over the past few decades has been driven primarily by the decline in fertility of Mozambican women. South African women's total fertility declined primarily in the early 1990s and has been wavering around 2.5 since 1995. Fertility decline has also been minimal for Mozambican women since 2002. With fertility decline stalling in both groups it remains to be seen if fertility will go below replacement level (2.5 in South Africa) as predicted by earlier research (26
). Further research is needed to determine the impact of factors such as infant mortality, changing marriage patterns, migration, and HIV on fertility in Agincourt and throughout South Africa.
Findings presented here suggest a few areas of future intervention that would be helpful in settings such as Agincourt. The pattern of childbearing in Agincourt shows that delaying first births could reduce overall fertility rates. Others have argued that family planning programmes in South Africa need to be reoriented to address the contraceptive needs of adolescents before first births (18
). Since contraception and family planning advice are provided largely by nurses working from primary health care facilities, strengthening the adolescentfriendly and responsiveness of clinic-based services is important. Programmes in Agincourt should pay special attention to Mozambican adolescents, whose reported contraceptive use is lower than that of South Africans. Increasing contraceptive use before age 20 will lower adolescent fertility and overall fertility rates. Furthermore, if programmes can successfully increase condom use, they may have the added benefit of reducing HIV transmission.
In other settings, increasing access to family planning and reproductive health programmes for all women has been shown to improve women's economic and health outcomes and to enhance economic growth (27
). However, the lingering effects of apartheid policies of differential development are evident in the low education and very high unemployment of women in Agincourt. Programmes that improve education and create job opportunities for all women, particularly Mozambican women, are needed to complement improvements in family planning and reproductive services in order to overcome endemic poverty in the area. Efforts to improve reproductive health services and improve the socio-economic status of women are likely to be synergistic, with each encouraging lower fertility and economic growth.
The primary limitations of our study are data driven. We do not have information on important variables such as prospective data on marriage, fertility desires, or detailed information on contraceptive use, to run models examining the proximate determinants of fertility.