This is our 14th blog post for our Job Market Paper Series blog for 2024-2025.
Abigail Stocker is a PhD candidate at the University of Illinois Urbana-Champaign. She is an applied microeconomist focusing on development, labor, and gender economics. You can find her JMP here.
Growing up in Nepal, Chakraman wanted to become a medical doctor. However, he had to give up his dream of going to medical school when he became a child groom at age 15. Instead of furthering his education, the financial resources that would have been invested in his studies were used to provide for his family (NPR, 2022). Roshni, from India, became a child bride at age 13 when she was married to a man over three times her own age. She experienced abuse from her husband and his family before escaping the marriage. Even after leaving, she struggled with depression and with the challenges of providing for her child (Global Citizen, 2024 ). Sadly, these types of stories are not uncommon, as millions of boys and girls around the world experience child marriage, often with little choice in the matter. This occurs despite the fact that child marriage is not legal in most countries and is even recognized as a human rights violation by the United Nations (United Nations, 1979).
Child Marriage in India, Indonesia, and Nepal
Recent studies have shown that child marriage, particularly for girls, is correlated with many negative outcomes, such as poor health, lower bargaining power, and lower educational attainment (Chakravarty, 2018; Jensen & Thornton, 2003; Parsons et al., 2015). However, the literature on male child marriage is almost non-existent, even though 6% of boys in the developing world marry before age 18 (UNICEF, 2023). Many factors, such as income, cultural norms, and marriage transfers, can influence the household’s decision to have sons and daughters marry. However, the ways that these factors impact child marriage incentives have not been fully characterized. In my paper, I investigate the question of how child marriage for both boys and girls responds to economic and social shocks.
My research focuses on India, Indonesia, and Nepal—countries with high rates of child marriage for both boys and girls. In my sample, between 1967 and 2010, 8% of boys and 38% of girls married before age 18. In these three countries, I can also study the role of marriage transfers, such as dowry (from bride to groom’s household, common in India and Nepal) and bride price (from groom to bride’s household, common in Indonesia). Marriage transfers could impact the financial incentives for households to participate in child marriage on both sides of the marriage market.
Data, Empirical Estimates, and Theoretical Model
Using individual-level data from IPUMS International and the Demographic and Health Surveys, I empirically estimate the impacts of a low rainfall or high temperature year (which reduces agricultural income) or a year with any conflict deaths on the probability of experiencing child marriage in the same year. I use rainfall and temperature data from the University of Delaware and conflict events and deaths from the Uppsala Conflict Data Program. In the analysis, I control for differences in child marriage rates across time and geography by including time and location fixed effects in the regressions.
To complement the empirical analysis and explain the mechanisms behind why child marriage responds to shocks, I develop a theoretical model, where pairs of households’ bargain over marriage transfer amounts. The model provides three main predictions: first, negative transitory income shocks will result in reductions in child marriage if the marriage transfer is partially influenced by cultural norms and cannot change directions (for example, a dowry payment would never be made in a market that practices bride price). Second, an increase in preferences for child marriage would cause an increase in child marriage rates, since households would be willing to sacrifice more of their income for child marriage to occur (and vice versa). Finally, the first two predictions will hold regardless of the direction of the marriage transfer (dowry or bride price). The existing literature, on the other hand, predicts that the response of child marriage to negative transitory income shocks is determined by the direction of the marriage transfer (Corno et al., 2020).
Negative Shocks Impact Child Marriage
My analysis reveals that negative income shocks—low rainfall or high temperatures—reduce child marriage rates for both boys and girls in the same year, as shown in Figure 1. These results hold across all three countries, suggesting that there is a positive relationship between transitory income shocks and child marriage that holds regardless of gender or the direction of the marriage transfer. This is likely because, when income falls, households are no longer able to afford the marriage transfer. Transfers become unaffordable because the transfer amount cannot fully adjust in response to a shock, since households that are in locations where dowry is practiced will never make a bride price payment, and vice versa.
I find that, when households are exposed to conflict-related violence (measured as any conflict deaths in the district in that year), there is a drop in the probability of child marriage for boys in Indonesia and an increase in the probability of child marriage for girls in the same year. This is shown in Figure 2. The effects of exposure to conflict seem to operate through channels other than income. Agricultural income is not reduced by exposure to conflict in my dataset, likely because these conflicts are small-scale. For boys, the likely channel is their recruitment into armed groups, which takes them off the marriage market. For girls, there is evidence that households use child marriage as a way to protect their daughters from experiencing physical or sexual violence (Sharma et al., 2015). If conflict increases the risk of experiencing violence, it makes sense that households would seek to protect their daughters by increasing their preferences for child marriage.
I also estimate the effects separately for different types of matches. There is reason to believe that child marriage might be worse for younger individuals, individuals entering matches with larger spousal age gaps, or individuals entering matches where the spouses are less compatible. I find that the effects of shocks on child marriage are smaller for these “worse” types of child marriage matches – children ages 12-14, entering matches where the spousal age gap is 5 years or more, or entering matches where the fertility preferences of the spouses are not aligned. This suggests that preferences for child marriage are stronger in cases where households choose to marry off their daughters even if it is quite harmful to the welfare of these children, so these households do not respond as much to shocks.
Policy Implications
My findings have implications for policies targeted at reducing child marriage. Specifically, if policymakers have multiple goals, such as simultaneously increasing income and decreasing child marriage, my results suggest that programs targeted at increasing household income should potentially be made conditional on children under age 18 in the household remaining unmarried. Further work is needed to understand exactly how this type of policy or conditionality could best be implemented.
Feature image generated using AI tools by the author.
