In this paper, I test a model in which visible contributions to religious institutions insure households through reputation with village lenders rather than divine protection. Reputation is effective only when lenders are able to lend. Such contributions should therefore rise after droughts, which hit poor households while sparing their richer lenders, and fall after floods, which hit the whole village. I test these predictions using a three-wave panel from rural Bangladesh. Droughts raise poor households’ contributions to mosques and temples, while time spent on early-morning religious rituals does not rise, and past donors are more likely to receive informal loans. These findings are consistent with religious giving as a way to build reputation with lenders. Contrary to the model, floods have no detectable effect on contributions.

Survey responses might be biased by enumerator characteristics, which could in turn affect impact evaluation estimates. This study demonstrates this in the context of intimate partner violence (IPV) in Bangladesh, by cross-randomizing the gender of the enumerator eliciting responses, and experimentally varying the rollout of an established anti-poverty program. It finds that respondents significantly under-report (0.51 sd) cases of IPV to males compared to females. Furthermore, the intervention significantly reduces (0.20 sd) incidence of IPV. However, this impact is only discernible when surveyed through females. If only male enumerators are used, the impact of the program is masked due to response bias.

Recent evidence shows that the inverse relationship between plot size and productivity is driven by an edge effect or by the over‐reporting of production by farmers on smaller plots. In this paper, we re‐examine the relationship between plot size and productivity in the context of Bangladesh. We also examine whether the effect of plot size is different across rice and non‐rice crops. We use panel data at the household level and estimate a fixed effects model. Results show that the effect of plot size on productivity is positive but not statistically significant. Further results show that there is no statistically significant effect of plot size on productivity for both rice and non‐rice crops. Our findings suggest that land redistribution, even if it leads to smaller plots, will not decrease farm production.

Increasing forcibly displaced populations worldwide are adversely affecting the poorest host communities' livelihoods. Livelihood programmes can reduce this tension by addressing host communities' skills and capital constraints. In this paper, we examine the effect of a customised version of BRAC's Ultra‐Poor Graduation (UPG) programme on the livelihoods of the host communities of Rohingya refugees. We find that the programme increases labour supply in self‐employment of working‐age men and women, household income, food expenditure and productive asset. Further, we find some weak evidence that the programme decreases the tension between hosts and Rohingya refugees.

Do Washington state’s targeted, public preschool programs help socially vulnerable students succeed in elementary school? In this study, we explore this question by examining the state’s Early Childhood Education and Assistance Program (ECEAP) and the Early Support for Infants and Toddlers (ESIT) program. Using statewide censuses of kindergarteners across multiple cohorts, we find that compared to white students, non-Asian students were more likely to participate in ECEAP, while ESIT participants were more likely to be male. On outcomes, we find ECEAP participants demonstrated significantly better kindergarten readiness compared to non-participants, but these effects diminished by 3rd grade. By contrast, ESIT participants performed relatively worse in kindergarten, but children who entered ESIT with higher initial skill levels and made greater progress during the program were better prepared for kindergarten. Although somewhat encouraging, these results do not speak to the causal impact of the programs because of how children were selected for participation.

Like other economic players, the novel pandemic severely hit small businesses—the larger source of growth and employment but also the most vulnerable sector—by disrupting national and international business networks, supply chains, and demand. To understand the evolving state of small enterprises during pre, par, and post-lockdown periods, BIGD in collaboration with Monash University, Australia conducted a survey on small enterprises, mostly light-engineering firms, and young workers across 18 districts in Bangladesh. The study finds that lockdown measures caused the majority of small enterprises to shut down, and during the early period of relaxing the lockdown, one-third of the enterprises were operating at limited capacity. Demand drop and the burden of fixed costs to run the businesses were the prominent reasons behind the drastic fall in profit. As a result, workers were losing jobs and the gender gap was widening, because female labour-intensive work (i.e. beauty parlour, tailoring) was affected harder. Other findings of concern include the emerging vulnerabilities for enterprises with lower endowments and poor access to government stimulus packages and other financial support. The study emphasises the importance of concrete targeting criteria and supports delivery platforms to assist more vulnerable enterprises. Finally, it highlights that the enterprises that received BRAC’s intensive training on occupational health and safety (OHS), along with business training and financial linkages, made double profits compared to their counterparts. However, the absolute amount of profit was substantially lower for both groups, compared to their pre-COVID profit, indicating the importance of scaling up such intensive training to create more resilient enterprises in such crises.

In this paper, I study how households share the burden of weather shocks between men and women. Using a three-wave panel from rural Bangladesh, I find that extreme excess monsoon rainfall reduces women’s paid employment and earnings, while men’s employment and earnings do not change. Women become more likely to report household work as their main occupation. Women’s livestock is not destroyed, but households sell fewer women-owned animals, while men’s animals continue to be sold. These findings suggest that women absorb a disproportionate share of weather shocks to the household, through a double burden of lost income and unpaid work.

How do graduation programs affect poor households’ access to government social safety nets? Using five waves of panel data from a randomized evaluation of BRAC’s Targeting the Ultra-Poor (TUP) program in rural Bangladesh, we find that TUP participation raises the probability of receiving safety-net benefits by 11.5 percentage points at the first major follow-up, an effect that persists in later waves. The gain comes at the extensive margin: the program brings previously excluded households into the government safety net for the first time. The mechanism does not appear to be informational, since about 80 percent of both treatment and control households already knew they were eligible. Instead, we explore a supply-side channel in which TUP integrates households into the community structures through which enrollment decisions are made. Our findings suggest that existing cost-benefit analyses of graduation programs undercount their social returns, and that coordinating graduation interventions with state safety nets could substantially improve targeting.