← Back to the MIDP Capstone Program
Every year, students at McCourt School’s Master in International Development Policy (MIDP) work with a client to produce an applied empirical report which answers a policy-relevant question. I provide some examples of previous projects below (including links to the reports and a policy brief summarizing the results).
2026
Corporal Punishment and Later Life Outcomes — Center for Global Development
Using four rounds of Young Lives panel data, students examined whether exposure to school-based corporal punishment at age 8 affected cognitive, psychosocial, and labor-market outcomes through age 22. Negative associations with cognitive performance (most consistent in Peru) and psychosocial outcomes (most consistent in Vietnam) persist seven years after exposure, suggesting that early developmental harm is durable rather than transitory.
Climate Change and Substance Use in Bangladesh — World Bank Group
Using a two-round household panel from Bangladesh linked to local weather data, students estimated how heat exposure shapes substance use and what channels drive the relationship. Hot-season substance use is 35% higher than in winter, the effect is concentrated among rural residents and women, and roughly a third of it operates through worsened sleep.
Drivers of Humanitarian Need in Conflict-Affected Eastern DRC — IMPACT Initiatives
Using IMPACT’s 2024 and 2025 Multi-Sector Needs Assessments alongside geolocated conflict data, students examined how humanitarian need evolved in conflict-affected eastern DRC following the simultaneous USAID withdrawal and rebel offensive in early 2025. Need rose across every province, with healthcare access, water access, and shelter conditions deteriorating the most. Health shocks are the strongest predictor of need in 2026 but not in 2025, when USAID programmes were still in operation — consistent with USAID programmes having insulated households from the consequences of shocks, an insulating role that disappeared with the funding.
Supporting Rural Women in the Lakhpati Didi Initiative — Indus Action
Students supported Indus Action in its evaluation of a programme to help rural women in Haryana access government livelihood schemes. The work included (i) designing and rolling out a baseline survey of 511 women; (ii) analysis and write-up of a baseline report; and (iii) a proposed evaluation design. The baseline shows that women across all livelihood groups know more schemes than they actually use, and that the poorest women face overlapping constraints in markets, agency, and institutional support that financial inclusion alone is unlikely to resolve.
Climate Vulnerability and the Allocation of Development Assistance — World Bank Group
Students constructed a harmonized sub-national panel dataset merging geocoded aid flows from 2001–2020 with country and sub-national measures of climate vulnerability and poverty. Preliminary results suggest that aid is more consistently associated with poverty than with climate vulnerability at both the national and sub-national level, implying that aid and climate risk may be linked only indirectly, through the socioeconomic dimensions of vulnerability itself. That said, the association between climate vulnerability and aid allocation has strengthened over time.
Mentors, Visits, and Learning: Foundational Literacy and Numeracy Reform in a South Indian State — Central Square Foundation
The team linked administrative visit records to multiple rounds of student assessments to test whether visit frequency predicts learning outcomes. They find that mentor visits to schools have risen substantially since the start of the state’s foundational learning reform, but urban and advantaged schools systematically receive more visits. But data limitations — a short follow-up window, non-random visit allocation, and a mismatch between observed teachers and assessed students — mean that the impact of mentoring on learning cannot (yet) be detected.
Report and Policy Brief Available on Request
Previous years
The Drivers and Returns to Migration — Mercy Corps
Using panel data from DRC, Pakistan, and Sri Lanka, students examined the effects of exogenous shocks on migration likelihood. Crop/livestock disease had the most significant effect, increasing the likelihood of migration by 11.2 percentage points.
The 2015 Gorkha Earthquake in Nepal — World Bank
Students analyzed exposure, vulnerability, and resilience patterns following the 2015 earthquake. Poorer households experienced proportionally greater asset losses and relied more on expenditure reduction strategies.
Inequality of Opportunity in Ethiopia — World Bank
Students examined trends in service access inequality between 2011-2016. Rural status is by far the largest driver of inequality, compared to income or education.
Can Sanitation Marketing Improve Latrine Usage and Reduce Incidence of Diarrhea? — iDE (Cambodia)
Students evaluated the impact of sanitation marketing in Cambodia. Latrine coverage increased by 17.8 percentage points while diarrhea decreased by 5.8 percentage points.
Household Resilience to Conflict in Nigeria — Mercy Corps
Using a triple-difference strategy, students examined determinants of resilience. Villages with higher social capital effectively mitigated negative conflict impacts on child malnutrition.
Ebola's Impact on Labor Markets in Sierra Leone — World Bank
Students studied employment effects in regions with varying Ebola exposure. Farming sector employment declined in heavily exposed regions, partially offset by increased non-formal self-employment.
Violence Prediction in Iraq — Mercy Corps
Students used machine learning to analyze violence predictors. Voting behavior and tribal chief connections emerged as the strongest predictive factors.
Private School Vouchers in India — ARK
Students evaluated a randomized low-fee private school voucher program in New Delhi. The program showed a negative impact on Hindi learning but a positive impact on English outcomes.