Polygyny and Child Growth: Evidence From Twenty-Six African Countries

ABSTRACT

Using household data from twenty-six African countries, this study examines the correlation between four measures of polygyny and child growth. External validity is added to existing small-sample evidence by investigating this correlation across many countries and by controlling for, as well as exploring, sources of heterogeneity at the regional, country, household, and maternal level. Household fixed-effects models indicate that the children of monogamous mothers have significantly greater height-for-age z-scores than children of polygynous mothers. Also, a low ranking in the hierarchy of mothers and the ratio of married women to men are negatively correlated with child height. The correlation varies widely across countries and is strongest for multigenerational polygynous households.

KEYWORDS:

• Polygyny
• child growth
• Sub-Saharan Africa
• demographic and health surveys
• heterogeneous effects

JEL Codes:

• C26
• D13
• O12

NOTES ON CONTRIBUTORS

Natascha Wagner is Assistant Professor at Erasmus University, Rotterdam. She is part of the Sustainable Development Group at the International Institute of Social Studies, The Hague. She completed her PhD at the Graduate Institute, Geneva. Her background is in microeconomics and econometrics with a focus on impact evaluations. She has implemented impact evaluations for the World Bank and the UNDP in Senegal and Côte d'Ivoire. Currently, she works on the long-term effects of quality improvements in public hospitals in the Philippines. Overall, her research covers social exclusion and health economics such as HIV/AIDS, child health, and female genital cutting.

Matthias Rieger is a Max Weber Fellow in the Department of Economics at the European University Institute (EUI) in Florence and an Academic Associate with the Development Impact Evaluation Initiative at the World Bank. In 2013, he completed his PhD in International Economics at the Graduate Institute of International and Development Studies in Geneva. He is a micro development economist, working on impact evaluations of development projects and the economics of civil war, as well as nutrition and health. Matthias has been involved in impact evaluations and household surveys in Morocco, the Central African Republic, Cambodia, and Burundi.

ACKNOWLEDGEMENTS

We received valuable comments from three anonymous referees and the associate editor. All remaining sins of omission or commission are our own.

Notes

1. The term “polygyny” refers to the form of polygamy where one man is married to multiple wives at the same time.

2. Based on George P. Murdock's Ethnographic Atlas (1967), Patrick J. Gray (1998) categorizes marital institutions in 1,231 societies over the period 1960–80. Only a minority of societies are found to be exclusively monogamous. Roughly 85 percent of the groups studied frequently or at least occasionally engage in polygynous unions – that is to say, it is not uncommon for males to be officially married to multiple wives.

3. Alternatively, we could also take the number of co-wives as a measure of the degree of polygyny within a household. For brevity, we do not report correlations between the number of co-wives and child growth. However, results are qualitatively similar to those presented in the section “Empirical Results.”

4. In the last measure, a small fraction of observations drop due to zeros in the denominator, describing a situation where the husband lives in another household.

5. An earlier version of this paper used variations in the sex composition of children born to the first wife to generate exogenous variations in the number of wives in order to estimate causal effects. However, due to data limitations, we decided to refrain from a causal analysis.

6. Supplementary Online Tables are available through the supplemental content tab on the publisher's website (http://dx.doi.org/10.1080/13545701.2014.927953).