Droughts and Deluges: Effects of Climate Extremes on the Gender Gap in Labor Supply

Over the past three decades, extreme climate events have caused losses of worth USD 4.5 trillion. Using a panel of 151 countries (1995-2019), I examine how extreme climate conditions shape gender gap in labor force participation. Key results show that the gender gap in paid labor exhibits a U-shaped relationship with droughts and an inverted U-shaped relationship with extreme wet conditions. The drought pattern is primarily driven by gender gap in employment while wetness affects gender gap in participation through unemployment. These relationships vary with country characteristics. Countries with high disaster-displacement risk exhibit declining gender gaps in participation during excess wetness while moderate-risk economies experience expanded gaps during droughts. Furthermore, the drought U-shape is most pronounced in countries with low to moderate empowerment while the nonlinear wet responses is concentrated only in moderately empowered countries. Lastly, both droughts and excess wetness expands gender gap in countries with weak net resilience to climate shocks.

💡 Research Summary

This paper investigates how extreme climate events—specifically droughts and excessive precipitation (floods)—affect the gender gap in labor market outcomes across 151 countries over the period 1995‑2019. Using ILO‑modeled annual data on male and female labor force participation (LFP), employment rates, and unemployment rates, the author constructs two continuous climate variables based on the 12‑month Standardized Precipitation Evapotranspiration Index (SPEI‑12). Drought intensity is defined as the negative deviation of SPEI below the 10th percentile, while wet intensity captures the positive deviation above the 90th percentile. These measures are entered into a two‑way fixed‑effects (TWFE) regression that also includes country‑year controls (GDP growth, urbanization, FDI inflows) and lagged covariates to mitigate reverse causality. Standard errors are clustered at the country level.

The empirical specification includes linear and quadratic terms for both drought and wetness, allowing the identification of non‑linear (U‑shaped or inverted‑U) relationships. The results reveal a clear U‑shaped pattern for droughts: at low drought intensity the gender gap in LFP narrows (women’s participation rises relative to men), but as drought severity increases the gap widens dramatically. Conversely, excessive precipitation shows an inverted‑U shape: moderate wetness expands the gap (women’s employment rises faster than men’s), while extreme wetness reduces it as overall economic activity contracts.

Decomposition of the gap shows that droughts primarily operate through the employment component, whereas wetness influences the unemployment component. To explain these mechanisms, the author develops a collective‑bargaining household model with husband and wife allocating time among market work, home production, and leisure. The model incorporates three channels through which climate shocks affect labor supply: (1) an income effect via changes in non‑labor assets, (2) a substitution effect through wage changes and home‑production productivity, and (3) a social‑cost effect that raises the marginal cost of women’s market work (e.g., increased mobility constraints, caregiving burdens). Under drought, asset losses and wage declines generate a positive income effect (reducing labor supply for both genders), but the social‑cost channel for women dominates, leading to a sharp reduction in female labor supply and thus a widening gap. Under moderate wetness, temporary demand for reconstruction raises wages and creates employment opportunities, especially for women, expanding the gap; however, extreme wetness triggers widespread disruption that depresses wages for both sexes, narrowing the gap.

Heterogeneity analyses demonstrate that country characteristics condition these patterns. Nations with high disaster‑displacement risk experience a pronounced widening of the gender gap under wet conditions but only a modest drought response; those with moderate displacement risk show the opposite. The gender gap response to drought is strongest in countries with low to moderate women’s empowerment, while the inverted‑U response to wetness concentrates in moderately empowered societies. Moreover, countries with low ND‑GAIN resilience scores exhibit gap‑widening under both drought and excess wetness, indicating that overall climate resilience mitigates gendered labor market shocks.

Robustness checks include using alternative dependent variables (employment gap, unemployment gap), employing different SPEI time scales (3‑month, 6‑month), adding controls for secondary education enrollment gaps, and clustering standard errors by region and year. The findings remain stable across specifications.

In sum, the study provides compelling evidence that extreme climate events shape gender disparities in labor markets through complex, non‑linear pathways that are mediated by household economics, social norms, and national institutional contexts. Policy implications are threefold: (1) targeted social protection and public service provision in drought‑prone areas to alleviate women’s unpaid care burden; (2) gender‑sensitive employment programs in flood‑prone regions to curb rising female unemployment; and (3) integrated strategies that simultaneously boost women’s legal and economic empowerment and national climate resilience, thereby limiting the amplification of gender gaps under future climate extremes.