Send message to Author

Paper short abstract

As OECD-DAC aid retreats, can the aid-effectiveness apparatus evaluate the donors replacing it? On 3,101 Qatari flows, allocation is sharply identified; impact is not identified at this design's power. Evaluation allocates legitimacy by measurability, through criteria emerging donors cannot meet.

Paper long abstract

With OECD aid retreating, aid is changing hands, but is it also changing outcomes? This paper puts that question to the evaluation apparatus rather than the donor. Results-based management and attribution-based impact evaluation were built around large DAC bureaucracies; applied to an emerging donor, what can they actually identify?

Using an original harmonised, project-level panel of Qatari foreign assistance (3,101 flows; USD 23.59 billion; 140 recipients; 2010–2024), the analysis models allocation as revealed preference, then tests separately the four assumptions attribution-based evaluation depends on: objective alignment, temporal cadence, spatial resolution, and contribution separability. Two-way fixed effects estimates span health, education and living-standard outcomes at national and, for the geocoded subset, subnational resolution, disciplined by future-aid placebos, Romano–Wolf correction and minimum-detectable-effect calculations.

Allocation is sharply identified and organised around political affinity rather than need; welfare impact is not identified, and the design could not have detected effects of the magnitude this donor's scale implies. The asymmetry is the finding: evaluation is treated as a governance technology that allocates legitimacy by measurability, an unevenly attainable criterion with distributive consequences. The paper's findings generalise, and the study specifies an evaluability screen and a decision rule assigning donors to attribution- or contribution-based assessment.