How poor households in developing countries live with climate shocks — floods, heat, and salinity — and the frictions that constrain adaptation and labor mobility, studied with field experiments, administrative records, and long-run survey data from Bangladesh.
Poor rural households smooth consumption through severe floods almost completely, yet three to seven years later the non-poor have rebuilt past their starting point while the poor remain 10 percent down in consumption and 21 points more likely to stay poor. The mechanism is financing: floods destroy the land value that collateralizes production credit.
Draft and full abstract coming soon.
Adaptation to heat is real but partial: neonatal mortality from moderate heat (30–40°C) has fallen to zero since the 1990s, tied to rural electrification, while mortality from extreme heat above 40°C shows no decline across four decades of birth cohorts.
Damage functions in climate economics assume that rising incomes and cooling technology steadily reduce the mortality cost of extreme heat. We test this using 239,563 births from eight rounds of the Bangladesh Demographic and Health Survey (1994–2022), merged with ERA5 daily-maximum temperature, and estimate how the heat–neonatal mortality relationship has shifted across birth cohorts. Adaptation is real but partial. Between 30 and 40°C, a positive mortality response appears only in the 1980s cohort and falls to essentially zero from the 1990s onward; a Sun–Abraham interaction-weighted estimator ties that flattening to the staggered rollout of Bangladesh's Rural Electrification Board program. Above 40°C, no adaptation occurred. The effect for the most recent cohort is large and precisely estimated — roughly 41 additional neonatal deaths per 1,000 live births per 10 extreme-heat days in the 28-day neonatal window — is carried by sustained heat waves rather than isolated hot days, and shows no decline across cohorts. Because adaptation is uneven across the temperature distribution, projections that attenuate the whole response with income understate the mortality cost of warming precisely where mid-century warming is concentrated: 815 to 2,540 excess neonatal deaths per year by 2060.
Son preference shortens birth intervals: mothers of two daughters conceive a third child sooner, raising the risk of neonatal death for that child by about 10 percentage points and implying roughly 35,000 excess neonatal deaths in Bangladesh over four decades.
Son preference raises child mortality through sex selection, discriminatory investment and early weaning. We document a fourth: birth spacing. Bangladeshi mothers of two daughters conceive a third child sooner, compressing the interval below the WHO-recommended 24 months. Instrumenting a sub-24-month third conception with the sex composition of the first two births, we estimate a 10.2 pp higher risk of third-child neonatal death in Matlab and Chakaria surveillance data and 11.4 pp in nine Bangladesh DHS rounds. The child harmed is the one the parents wanted, male with even probability. Nationally, this implies about 35,000 excess neonatal deaths over four decades.
Using census data on over one million farm households and a salinity discontinuity at the Bengal delta's hydrological boundary, this paper shows that salinity intrusion drives large reallocation from crops to aquaculture and a 52 percent rise in farm size, achieved through the rental market rather than land sales; religious diversity eases the transition and parcel fragmentation slows it.
This paper studies whether climate-induced shifts in land productivity can trigger structural adaptation in agriculture by overcoming long-standing frictions in land markets. I focus on coastal Bangladesh, where rising sea levels have increased dry-season soil salinity beyond the threshold at which conventional crops can survive. Using high-resolution census data on over one million farm households, I exploit a natural discontinuity in salinity at the hydrological boundary between the tidally active and mature regions of the Bengal delta. I implement a fuzzy spatial regression discontinuity design and an instrumental variables strategy to estimate the causal effect of high salinity on land use and farm structure. I find that salinity above 12 dS/m leads to a large reallocation of land from crop farming to aquaculture—by 32 to 50 decimals per household—and induces a 52 percent increase in average farm size and a 23 percent rise in farm size inequality. This transition is not driven by land sales, which remain stagnant, but by consolidation through the rental market: the share of households renting in land declines, while land rented per renter doubles. I show that this process is shaped by local frictions. In particular, land reallocation is significantly more responsive in villages with greater religious diversity (which improves cross-group contract enforceability) and less responsive in villages with high land parcel fragmentation (which raises transaction costs). These findings highlight the potential for market-mediated adaptation to climate change even in historically stagnant land markets, but also underscore the role of social and spatial frictions in shaping who adapts and how.