Poor countries allied with the United States appear to get more aid when they act a little less friendly. When such a country skips or abstains on more UN General Assembly votes, it tends to receive more aid the following year: a one standard deviation rise in this “strategic nonresponse” goes with about $83 million more.
The study is by Anthony Luongo of Princeton University. It was published online in International Studies Quarterly on August 20, 2026, and is open access.
Pretending to drift away
Most research on foreign aid looks at what donors want. Luongo turns it around and asks how recipients bargain. A donor buying policy cooperation does not know what a recipient truly prefers, so it cannot tell how much aid is really needed.
In his game-theory model, a recipient sends a signal about its preferences before the donor makes an offer, and pays a price only if the signal is dishonest. When lying is cheap and the donor cares enough, recipients that already agree with the donor sometimes claim they do not, and collect aid they would not otherwise get. The model suggests this can keep working even when the game is repeated, though under narrower conditions.
He opens with South Africa, which abstained on the 2022 UN vote condemning Russia’s invasion of Ukraine and hosted Russia’s foreign minister while still cooperating with the West. The U.S. responded with a series of high-profile diplomatic visits.
Thirty years of UN votes and aid
The General Assembly suits this kind of bluff because its resolutions are nonbinding, so voting against one’s real views carries little cost. Luongo counted abstentions plus “strategic absences,” days a country missed a vote while voting on other resolutions, and linked them to the next year’s aid from the U.S., the World Bank and the IMF from 1992 to 2021.
The effect showed up only where the model predicts it: among U.S. allies poor enough to qualify for the World Bank’s International Development Association. A one standard deviation increase, about 7 to 8 more nonaligned votes a year, matched 0.3 percentage points of GDP in extra aid, against an average of 2.5 percent. For allies with the least similar alliance portfolios to the U.S., the gain rose to about $100 million. Non-allies and richer countries saw no such reward.
Pakistan, which met the conditions in every year studied, is the leading example. Its strategic nonresponse tracked U.S. aid inflows despite years of complaints in Washington about its “double game.”
Not just vote buying
Luongo checked whether the U.S. might simply be paying for votes. Nonresponse did not follow earlier aid increases, and high nonresponders were larger economies rather than the small, aid-dependent states typical of vote buying. Absences that look driven by capacity problems behaved differently, helping most the allies already closest to the U.S.
He notes the data cannot show whether any single abstention was honest. But the result questions a common practice: using UN voting records as a clean measure of where countries really stand.
Study Details:
- Title: Playing Hard to Get: Strategic Signaling in Aid Bargaining
- Authors: Anthony Luongo
- Journal: International Studies Quarterly
- Publication Date: August 20, 2026
- DOI: 10.1093/isq/sqag065