Endogenous childlessness and stages of development

Thomas Baudin, David de la Croix, Paula E. Gobbi 25 July 2017



Surprisingly, some developing countries have higher childlessness rates than developed countries. In Cameroon in 2005, for instance, 17.8% of women who completed their fecund lifecycle never had a child. Compare this to the US: in 2012, the number was 16.1%. The number of children born to women who completed their fertility lifecycle, however, was 3.94 in Cameroon and 1.98 in the US. In the region known as the African Infertility Belt, to which Cameroon belongs, it is not unusual for a woman to remain childless. In this region, up to 40% of women in some regions will be childless (Romaniuk 1980). Demographers have described this phenomenon, and suggested that the relationship between childlessness and development is U-shaped (Poston and Trent 1982).

Figure 1 shows the average number of children ever born (left panel) and childlessness rates (right panel) by education level, based on census data from 38 developing countries. The relationship is not the same along the intensive margin of fertility (the number of children that mothers have) and the extensive margin (the fraction of women who are mothers). In the left-hand panel, the intensive margin of fertility declines with education. In the right-hand panel, childlessness (the opposite of the extensive margin) increases only after nine years of education. Before nine years, childlessness is decreasing, consistent line with the U-shaped relationship.

Figure 1 Fertility and education (left panel) and childlessness and education (right panel)

Women are childless for different reasons in developing countries than in developed countries. Distinguishing these reasons is, however, not straightforward. Most datasets usually do not capture why a woman did or did not have children. In a new paper, we proposed a theoretical model that quantifies the reasons behind childlessness (Baudin et al. 2017).

Reasons for childlessness

There are as many reasons for remaining childless as there are women but, in our theoretical model, women remain childless for four distinct reasons.

  • Opportunity-driven childlessness: Some women remain childless because they face a high time cost of having children. Highly educated women earn high wages, and face a high opportunity cost when they are not at work (Gobbi 2013, Aaronson et al. 2014). This category also includes women who prefer not to have children because they are single. Opportunity-driven childlessness is the main reason for the high levels of childlessness in developed countries.
  • Natural sterility: This refers to the innate biological impossibility of having children, which does not depend on the level of education or wealth. It affects at least 2% of couples.
  • Poverty-driven childlessness: Low-educated women may remain childless, mainly because of untreated venereal diseases, pregnancy-related infections, malnutrition, or pollution.

For all the above reasons, there is some childlessness because of lack of an adequate partner. In a couple, poverty-driven childlessness is less likely because of the resources brought by the husband, and opportunity-driven childlessness is less likely because sharing childrearing activities is possible.

  • No newborn children survive: This can also be important in developing countries, in which health facilities are lacking, and infant mortality is high.

The responses of childlessness to development

In our research, we show that the expected effect of policies on fertility can be magnified, mitigated or even reversed by the way in which the extensive margin adjusts. To understand why, we consider four distinct changes associated with development: an increase in contraceptive availability, a reduction in infant mortality, an increase in female empowerment, and establishing universal primary education. All these changes are all advocated by international organisations to accelerate the demographic transition in poor countries.

Educating women about contraception to reduce unwanted births has a negative expected effect on fertility (Bongaarts 1994). Knowledgeable women are more attractive to potential mates, leading to an increase in the marriage rate. Some women who would have been single and childless become married mothers instead, leading to an increase in fertility along the extensive margin. Hence the reduction in the risk of having unplanned children can have a positive effect on fertility through the positive effect on marriage and motherhood rates. Accounting for the extensive margin of fertility therefore diminishes the expected negative effect that family planning policies have on fertility rates. Not accounting for this margin may lead planners to be over-optimistic about the effect of contraceptives on the demographic transition.

Empowering women increases the opportunity cost of childrearing time relative to working time, and so would be expected to decrease fertility. The negative effect is amplified when most childlessness is due to the opportunity cost. If many women suffer from poverty-driven childlessness though, the effect of female empowerment on the extensive margin of fertility may be positive, and therefore the overall impact on fertility diminishes. For example, empowering women in Argentina may lead to a stronger drop in fertility than expected by increasing opportunity-driven childlessness, while empowering women in Cameroon may lead to a lower drop in fertility than expected by reducing poverty-driven childlessness.

Finally, an increase in elementary education would be expected to decrease fertility. We find that fertility will decrease only in countries that are not much affected by poverty-driven childlessness. In many sub-Saharan countries, this is not the case. Poverty-driven childlessness also declines after an increase in education, and therefore the overall effect on fertility can end up being positive. Improving education unambiguously reduces fertility only when expected years of schooling are already above nine years.

Childlessness is not just low fertility

To conclude, let us stress that childlessness is not just a limit case of low fertility. It obeys to its own logic. Given its quantitative importance in some developing countries, ignoring it might lead to wrong predictions on how development affects fertility.


Aaronson D, F Lange and B Mazumder (2014), “Fertility Transitions along the Extensive and Intensive Margins”, American Economic Review, 104(11): 3701-24.

Baudin T, D de la Croix and P E Gobbi (2017), “Endogenous Childlessness and Stages of Development”, CEPR Discussion Paper 12071.

Bongaarts J, (1994), “Population Policy Options in the Developing World”, Science, New Series, 263(5148): 771-776.

Gobbi P E (2013), “A Model of Voluntary Childlessness”, Journal of Population Economics, 26(3): 963-982.

Poston D L and K Trent (1982), “International Variability in Childlessness: A Descriptive and Analytical Study”, Journal of Family Issues, 3(4): 473-91.

Romaniuk A (1980), “Increase in Natural Fertility During the Early Stages of Modernisation: Evidence from an African Case Study, Zaire”, Population Studies, 34(2): 293-310.



Topics:  Development Gender

Tags:  fertility, demographic transition, childlessness

Associate Professor, IESEG Business School

Professor at Université catholique de Louvain, Belgium

Postdoctoral Research Fellow FNRS, Université catholique de Louvain; CEPR Research Affiliate