Aid at the margins: timing, scale, and targeting in refugee camps

Humanitarian assistance is a central feature of economic life in refugee camps. Yet, since 2023, shifts in donor commitments have sharply reduced aid budgets, leading to substantial changes in aid design, including delivery timing, transfer size, modality, and allocation rules. The effects of these changes are both practically relevant and theoretically ambiguous.

This thesis studies how variation in the timing, scale, modality, and allocation of humanitarian aid affects welfare, credit, prices, and social relations in refugee camp economies. Focusing on Kakuma refugee camp and the adjacent Kalobeyei settlement in Kenya, the analysis combines high-frequency household panel data, weekly market price data, and extensive qualitative interviews collected between 2022 and 2025.

The first analytical chapter examines the effects of aid distribution and delays in aid delivery. Exploiting high-frequency surveys with random interview timing, it shows that refugee households smooth consumption under regular aid cycles but experience sharp declines in food consumption and food security when transfers are delayed. Informal credit through local retailers partially mitigates these shocks, but at high cost, with prices for credit-based purchases around 17 percent higher.

The second chapter studies the short-run consequences of a large, unanticipated 20% cut in humanitarian assistance. Exploiting exogenous variation of interview dates around the cut-off, it shows that the aid reduction leads to substantial declines in calorie intake, dietary diversity, food security, and expenditures, alongside deteriorations in mental wellbeing and shifts in risk and time preferences. The shock also triggers a contraction in informal credit and food prices.

The final chapter examines the shift from universal aid provision to a vulnerabilitybased targeting approach. Drawing on qualitative evidence, it shows that targeting in a context of widespread poverty and aid dependency is associated with extensive exclusion errors, disruption of informal credit and mutual support systems, erosion of trust in humanitarian institutions, and contributes to insecurity, distress-driven migration, and economic contraction. In these contexts, the social and administrative costs of vulnerability targeting outweigh its perceived efficiency gains.

Overall, the findings suggest that different design choices along any of the margins studied here: timing, scale, modality, or targeting, have far-reaching economic consequences, and they can amplify welfare losses by simultaneously reducing income and dismantling key coping mechanisms. The thesis contributes to theories of intertemporal choice, general equilibrium, and informal credit, and highlights the importance of predictability and universality in highly aid-dependent settings.