Researchers examined whether the sudden termination of the U.S. Agency for International Development (USAID) in early 2025 changed patterns of violence in African areas that had received U.S. assistance. Their study, published in the journal Science and led by University of Chicago data scientist Austin Wright, finds that abrupt withdrawal of aid corresponded to a measurable rise in conflict in places that had relied on those programs compared with places that had not.
Why aid can affect violence
Aid can reduce violence by creating jobs, supporting services and sustaining livelihoods, and thereby lowering incentives to join armed groups. But it can also generate conflict by creating valuable things—food distributions, clinics, infrastructure—that become objects of competition or control. Wright and his colleagues asked what happens when those valuable resources and the economic opportunities they provide vanish almost overnight.
Methods and evidence
The team mapped USAID disbursements at the state or provincial level before the agency shut down and matched that to detailed, event-level conflict data from the Armed Conflict Location and Event Data Project (ACLED). They compared violent events—armed clashes, protests, riots and attacks on civilians—during the ten months before and after the shutdown to see whether places that had received more USAID funding experienced different trends than places that had received less or none.
Findings
The analysis showed an increase in violence in locations that had previously received more aid once that assistance was cut. The researchers attribute this to the sudden loss of economic opportunities: wages, food programs and health services stopped, leaving people more vulnerable and more easily recruited by armed groups. At the same time, the assets and grievances that motivate fighting—territory, infrastructure, political power and ethnic tensions—did not disappear, creating conditions for more conflict. The observed rises included combat between armed groups, protests that turned violent, and deliberate attacks on civilians.
Concrete example
The authors cite protests at Kenya’s Kakuma refugee camp in July 2025, where roughly 300,000 refugees depended on food and services financed by U.S. assistance. After cuts, food distributions were sharply reduced and demonstrations escalated into rock-throwing, arson and at least one death—an episode the researchers say fits their broader pattern.
Exceptions and institutional buffers
The study found smaller effects in places with stronger institutional constraints on executives—settings where governments could not unilaterally redirect policy or declare emergency measures. In some countries, national budgets or program decisions helped fill gaps after the USAID cuts; the authors point to Nigeria’s supplementary health funding and South Africa’s efforts to cover HIV treatment shortfalls as examples.
Response and outside assessment
A U.S. State Department spokesperson, Tommy Pigott, criticized the report, saying it ‘‘fundamentally ignores what is actually happening in Africa’’ and arguing that the administration had refocused assistance on efficiency and partnership. Independent reviewers note that studying conflict is challenging and raise technical caveats such as contagion effects, where violent events spread across nearby areas. Still, at least one external statistician called the results convincing and unlikely to be overturned by those technical issues.
Implications
The authors warn that the increase in violence may have lasting consequences. Recent conflict is a strong predictor of future conflict, so a short-term surge in violence caused by abrupt aid withdrawal could be self-reinforcing and not quickly reversible even if assistance were restored. The study highlights how sudden changes in external support can reshape local incentives and stability, and it suggests that policymakers should weigh these risks when altering or ending aid programs.