Summary

This session has introduced the concept of proximate determinants of fertility as biological and behavioural factors which are themselves influenced by distal socio-economic, cultural and environmental factors.

It has introduced Bongaarts’ aggregate proximate determinants model to understand fertility differentials across populations and over time, by focusing interest on the links between each of the proximate fertility variables. Based on the model, level of fertility is determined by (a) marriage or sexual union (proxy of exposure to sexual intercourse), (b) postpartum infecundability, (c) contraceptive prevalence and their use-effectiveness, and (d) abortion. Other factors, such as pathological sterility, are also important in some populations.

We have extensively described the limitations of Bongaarts’ model related to pathological sterility, sexual activities and contraceptive prevalence. Furthermore, it is considerably difficult to make accurate estimation of abortion and use-effectiveness of contraception.

The model has been improved by Stover, especially by the use of sexual activity instead of marriage - making it more applicable in many countries outside of Asia (see Figure 11 on Page 26). Stover's model also takes proper account of sterility, which is an important and changing factor in many developing countries. However, Bongaarts' model is widely used.

Regardless of the limitations, the model is conceptually useful in the study of fertility and practically useful too. It allows the modelling of “what if” situations such as, ‘what if’ situations such as ‘what would happen to fertility (TFR) if contraceptive prevalence increased to x %’. It allows fertility comparisons of countries where fertility is changing by different combinations of factors, though the difficulty of getting reliable data on abortion, sexual activities, and consistent contraceptive use and its effectiveness is weakness. The model is useful to assess variations in fertility, but it is not perfect.