Essays on Early Marriage Across Sub-Saharan Africa: An Economic Perspective

Foley, Sarah (2021) Essays on Early Marriage Across Sub-Saharan Africa: An Economic Perspective. Doctoral thesis, University of East Anglia.

[thumbnail of SFoley_Final Thesis.pdf]
Download (15MB) | Preview


Rates of early marriage, here defined as any legal or customary union involving a male or female below the age of 18, have declined significantly over the last decade. However, progress has not been equitable, and rates of early marriage remain extraordinarily high in parts of Sub-Saharan Africa. There is widespread consensus that early marriage causes a significant disruption in a child’s accumulation of human capital and has significantly negative intergenerational repercussions. Using a combination of secondary data from the Demographic and Health Surveys (DHS) and experimental data from the field, this thesis will empirically explore the phenomenon of early marriage across Sub-Saharan Africa.

Chapter 1 of this thesis explores associations between adolescent nuptiality and fertility patterns using the most recent Demographic and Health Survey datasets for thirty countries of Sub-Saharan Africa. Unlike previous studies examining marriage and fertility trends, we expand the adolescent period to more refined age categories to better capture age-specific variations in female sexual, marital, and reproductive trajectories. Overall, results suggest that among middle adolescents (ages 15 to 17), marriage-specific fertility levels are 4% higher than the fertility levels of women marrying as adults. However, variations between countries are substantial, and some results significantly contradict the assumptions of the aggregate fertility model. We speculate that some differences between countries are due to inconsistencies in domestic marriage laws. In addition to examining fertility levels, we also investigate precise marriage-birth sequences and find that most adolescent births result from marital conception. However, we find some evidence of middle adolescent premarital sexual activity that led to birth within the first eight months of marriage. The chapter concludes with a case study on Ghana, examining whether a domestic law criminalising underage marriage effectively reduced its incidence, and whether this law had an overall demographic effect. Using a multi-stage regression discontinuity design, results indicate that early marriage reduced by approximately 6% due to Ghana’s 1998 Children’s Act. This reduction simultaneously increased the age of women at first birth by delaying marriage. Furthermore, our results found an overall welfare-improving effect, with reduced reports of emotional, sexual, and physical domestic violence against women.

In Chapter 2, using lab-in-the-field experimental and comprehensive survey data, we examine whether age at first marriage affects the willingness of husbands and wives to cooperate to maximise household gains. Among the Bagisu of East Uganda, we find that women who marry older are more cooperative with their husbands. In a series of corresponding inter-household games, we conclude that female behaviour is not driven by the selection of more cooperative women into progressively later union, but by the marriage institutions’ effect. As an extension to the core analysis, we further examine the role of education and the cultural practice of Bridewealth on rates of cooperation. This chapter concludes by evaluating the linkages between the behaviour exhibited in our intrahousehold games and spousal behaviour in everyday lives. We find that pre-existing cooperative behaviours positively correlated to in-game contributions, particularly for husbands. Here, we add to the recent literature that focuses on correlates between behaviour in the lab and real-life behaviour.

In Chapter 3, using an original, multi-stage sampling strategy, we further investigate intrahousehold behaviour using a modified version of the Trust Game. Data gathered via an initial census allowed us to assign households to early and later marriage stratifications, based on the wife’s age at marriage. As in previous studies, our results appear inconsistent with the assumption of Pareto efficiency in household decision-making. We reject the unitary and collective household models and identify early marriage as a channel through which trust and reciprocity can affect low household efficiency levels. Specifically, we find that women married as a ‘child’ exhibit less trust to their husbands than women who marry as adults. In a series of interactions with education, we observe that the negative effect of early marriage on female trust is the same, even with increased levels of education. By employing a within-subject design to our lab-in-the-field experiment, we directly compare intrahousehold behaviour with stranger counterfactuals. We find weak evidence suggesting that women married under 18 trust men from other households less. Similarly, men who married a bride under 18 exhibit significantly less trust to women from different households. We do not, however, observe any significant behaviour from men in our intrahousehold treatment. Throughout, our results are robust to a wide variety of control variables, and we find evidence suggesting that lab behaviour roughly mirrors analogous real-life household behaviour

Item Type: Thesis (Doctoral)
Faculty \ School: Faculty of Social Sciences > School of Global Development (formerly School of International Development)
Depositing User: Kitty Laine
Date Deposited: 16 Nov 2022 12:48
Last Modified: 16 Nov 2022 12:49


Downloads per month over past year

Actions (login required)

View Item View Item