This is our fifth blog post for our Job Market Paper Series blog for 2024-2025.
Fatima Najeeb is a PhD candidate at the University of Maryland, College Park. Her research interests lie in applied microeconomics with a focus on development, family, and environmental economics. You can find her JMP here.
Climate change is not just a global threat; it is reshaping lives within family homes. In 2019, 4.5 billion people worldwide faced extreme weather events—a 12% increase since 2010 – with South Asia being the hardest hit (Doan et al., 2023). While much of the research on climate impacts has examined household-level outcomes, fewer studies have delved into how climate shocks might have uneven effects on the material well-being of people living under the same roof. In a perfectly egalitarian society, we might expect that all household members would experience the adverse effects of such shocks equally. However, when cultural norms disadvantage certain groups—such as women—we might think that these groups would bear a greater burden.
My job market paper examines how climate shocks reshape the material well-being of household members in unequal ways in rural Bangladesh. I use the concept of “resource shares”—the fraction of total household spending that each family member effectively controls—to capture individual well-being and intra-household inequality. Unlike traditional measures of per-capita consumption, which assume equality within households, my approach accounts for potential differences in resource allocation. Specifically, I ask: does exposure to a climate shock like a flood shift resources away from women (and children) towards men, deepening existing vulnerabilities? And if so, do these effects persist over time?
Identifying Underlying Effects on Resource Allocation within Households
As resource shares are not directly observed in multi-person households, I estimate them using a structural model based on collective household theory, which allows me to infer each member’s share of the total household budget without complete individual consumption data. I identify resource shares in this framework using variation in spending on private and assignable goods—goods consumed exclusively by known members—and total household expenditures, as in Dunbar et al. (2013). Building on their cross-sectional model, I extend this to a longitudinal framework to estimate how household resources are divided over time. As Browning et al. (2013) demonstrate, there is a monotonic correspondence between resource shares and bargaining power, highlighting the broader interpretation of intra-household resource distribution.
To see how these resource shares change with exposure to a climate shock, I use as-good-as-random variation in flood exposure across villages. I limit my sample to villages with similar ex-ante flood risk, so that the timing of the first flood during the study period (2012–2018) is effectively random. I create cohorts of villages based on their first flood year and focus on the 2014 cohort, which is the largest of these groups. By combining satellite-based flood data from the Global Flood Database with household-level data from the Bangladesh Integrated Household Survey (collected in 2011 before the flood, in 2015 six months after, and in 2018 four years later), my research design allows me to examine both immediate effects six months after the flood and longer-term impacts four years later.
Floods Shift Burden Towards Women, Benefit Towards Men
I find that flood exposure triggers a reallocation of household resources away from women—and, to a lesser extent, children—towards men within the household. Figure 1 shows the average resource shares of men and women, as predicted by the model, in flooded and unflooded households. Given the exogenous nature of the flood shock, resource allocations in unflooded households represent what would have happened in flooded households had the flood not occurred. Since resource transfers mainly occur between men and women, and the effects on children’s shares are smaller, children are excluded from this graph.
- Baseline: In my sample of villages with similar ex-ante flood probabilities—where households are also balanced in observable characteristics—men’s and women’s resource shares are balanced across flooded and unflooded households pre-flood.
- 6 Months Post-Flood: A flood sets off a redistribution in household resources. Women’s share of resources is 9 percentage points (pp) lower than in unflooded households (28% drop), while men’s share rises by 11 pp (33% higher) relative to counterparts in unflooded households.
- 4 Years Post-Flood: Women’s share remains 7 pp lower than those of women in unaffected households, while men’s share stays elevated by 12 pp, showing these shifts are not temporary.
This reallocation of resources away from women and towards men becomes stronger with the number of floods experienced.
Floods Push Women Closer to Poverty, Men Further from It
To connect these changes in resource shares to material well-being, I use the predicted resource shares and calculate each person’s “shadow budget”—their unobserved individual budget within the household. Six months post-flood, the daily shadow budget for a woman in a flooded household averages $3.33 (in 2017 PPP terms), compared to $6.59 for a man—a daily gap of $3.26. With the extreme poverty line at $2.15/day, these numbers reveal a greater vulnerability for women in flooded households. For context, in unflooded households, the difference between men’s and women’s shadow budgets is just $0.61. Four years later, this gap narrows only slightly to $3.18 in flooded households, while it remains negligible in unaffected ones. In both post-treatment periods, men in flooded households have higher shadow budgets than men in unflooded households. This disparity underscores how climate shocks increase women’s vulnerability, pushing them closer to poverty, while men in flooded households move further from it.
Why Do Women’s Resource Shares Decline?
The redistribution of household resources following a flood is closely linked to changes in women’s earnings relative to men’s. Six months after the flood, women’s relative earnings fall by approximately 4 pp (53%), significant at the 5% level, primarily driven by a decline in poultry farming. Four years later, this decline reduces but remains at 2 pp (24%). This decline occurs as women have limited options to substitute occupations. While both men and women are less likely to work in farming after the flood, men can shift to day labor, an option less available to women. This restricted adaptability for women, consistent with findings by Afridi et al. (2022), highlights the gendered economic vulnerabilities that climate shocks exacerbate.
The interpretation that women’s relative earnings drive the redistribution of household resources is strengthened by the observation that, as the flood’s impact on women’s relative earnings subsides, the effect on their resource shares also diminishes. Figure 2 illustrates this relationship. On the left, I plot the effects of the flood on women’s resource shares, varying treatment intensity—measured by the share of village area flooded—with 90% confidence intervals. On the right, I show the corresponding effect of the flood on women’s earnings relative to men’s. In the short term, the flood’s effect on women’s resource share remains stable across flood intensities, but in the long term, it lessens at higher intensities. A similar pattern appears in women’s relative earnings: as flood intensity increases, men’s earnings also decline in the longer run, reducing the earnings gap between men and women. Importantly, the structural model does not account for earnings or employment, so the alignment between effects on resource shares and reduced-form results on relative earnings reinforces this interpretation.
Policy Implications
This paper highlights a strong case for incorporating intra-household inequalities into climate adaptation policies and financing goals. Since changes in women’s relative earnings drive these redistributive effects within the household, improving women’s option to work through diverse employment opportunities and securing their livelihoods could help mitigate vulnerabilities to future shocks.
Feature Image from https://global-flood-database.cloudtostreet.ai/
