The Elusive Quest for Aid in Uganda examines why, despite receiving over US$50 billion in development assistance since 1990, per capita income has stagnated and poverty remains above 20%. Using an original panel dataset that merges geolocated aid projects with high-resolution administrative, conflict, and survey records (1990–2023), we test three dominant explanations:

(i) donor fragmentation and weak alignment with national priorities;

(ii) elite capture and rent-seeking that divert resources from intended beneficiaries; and

(iii) external shocks—such as climate extremes and refugee inflows—that erode programme effectiveness.

Difference-in-differences estimates:

Difference-in-differences estimates show that districts receiving above-median per capita aid experienced 0.5–0.7 percentage points lower annual poverty reduction than under-aided districts, conditional on initial poverty and conflict exposure.

Instrumental-variable results:

Instrumental-variable results that exploit exogenous World Bank disbursement rules reveal that each additional US$100 per capita of aid is associated with a 3.1% increase in local government consumption but only a 0.08% increase in household consumption.

Narrative evidence:

Narrative evidence from 86 elite interviews and 12 focus group discussions suggests that fragmented donor procedures create parallel implementation structures, crowd out domestic tax effort, and facilitate rent extraction.

Political-economy model:

A calibrated political-economy model indicates that, under current institutional quality, even optimally allocated aid would close at most one-third of Uganda’s poverty gap by 2030.

Conclusion:

The findings imply that scaling aid without first strengthening domestic accountability mechanisms is unlikely to yield transformative impacts, and they provide a quantitative benchmark for the governance reforms required to make external finance truly effective.