Flags of member nations fly outside the General Assembly building at the United Nations headquarters in New York in 2005. (Adam Rountree/AP)

President Trump and Nikki Haley, the U.S. ambassador to the United Nations, have argued that the United States should withhold foreign assistance to countries that do not support U.S. global objectives. This sort of leverage was the crux of Haley’s threat that the United States would be “taking names” ahead of the recent U.N. General Assembly (UNGA) vote on the U.S. Embassy’s move from Tel Aviv to Jerusalem.

At his Jan. 30 State of the Union speech, Trump made it clear that U.S. foreign assistance should “always serve American interests, and only go to America’s friends.”

But the administration’s fiscal 2019 budget request released this week takes a step back. The 2019 request restored funding to 24 of the 37 countries that were zeroed out in the fiscal 2018 request — a move that Congress rejected during appropriations. This could signal tacit recognition that the “friends only” policy on foreign assistance — while in line with Trump’s “America First” agenda — is fraught with problems, including identifying who America’s “friends” actually are.

Can foreign aid influence U.N. votes?

The friends-only policy isn’t new. President Ronald Reagan and Ambassador Jeane Kirkpatrick pursued a comparable policy in the 1980s, for similar reasons. During a 1983 Senate hearing, Kirkpatrick argued that the United States should communicate to nations that their votes, their attitudes and their actions inside the U.N. system inevitably must have consequences for their relations with the United States outside the U.N. system.” She contended that U.N. voting patterns “should also be one of the criteria we employ in deciding whether we will provide assistance, and what type of assistance and in what amount.”

Despite this tough talk, the administration never fully implemented Kirkpatrick’s approach. The Reagan administration pushed through overall reductions in foreign assistance, so even recipient countries that backed the United States at the General Assembly received less foreign aid than before. However, the threats may have achieved their desired effect of moderately increasing cooperation on U.N. votes that were particularly important to the United States.

But research has found some foreign assistance does in fact affect how countries vote at the General Assembly. In particular, one study suggests that U.S.-funded general budget support and grants incentivize countries to vote in line with the United States. Countries do vote with the United States for other reasons as well, including their position toward the U.S.-led world order. Overall, the association between foreign aid and General Assembly votes is relatively weak.

The United Nations isn’t that friendly

To better understand what “friendly” behavior at the United Nations looks like, I analyzed voting patterns during the past two presidential administrations. From 2001 to 2015 (the most recent year available in this data set by Michael A. Bailey, Anton Strezhnev and Erik Voeten), relatively few countries regularly aligned their votes with the United States. I grouped countries into three categories, each representing one-third of the observations: “hostile,” “ambivalent” and “friendly.”

Hostile countries demonstrated consistent opposition to the United States, voting with the United States less than 12 percent of the time on average. A subset of hostile countries, what we generally refer to as “strategic partners,” includes countries often politically hostile to the U.S. positions at the General Assembly but partnered with the United States in other domains, including military operations.

Despite averaging roughly $1.5 billion in U.S. foreign assistance annually, Egypt voted much like Iran at the UNGA in the mid-2000s. The massive amounts of U.S. foreign assistance to Egypt that followed the Camp David Accords initially may have bought some goodwill at the United Nations, but its influence clearly waned over time.

Ambivalent countries tend to vote with the United States between 12 percent and 22 percent of the time. Compared with Pakistan, for instance, the United States gives Bangladesh very little aid. Yet Bangladesh and Pakistan offer the United States similar levels of political support. Since 2001, Bangladesh voted with the United States an average of 12 percent of the time, compared with 11 percent for Pakistan.

U.S. support for Israel helps account for some of the low levels of support, as General Assembly votes are stacked against Israel. Muslim-majority countries tend to vote as a bloc on issues relating to Israel-Palestine, while the United States votes with Israel roughly 90 percent of the time. Since the end of the Cold War, Muslim countries have shifted toward a non-Western bloc in U.N. voting.

Threats are unlikely to affect many of these countries’ behavior, but promises of foreign aid might be an incentive to increase cooperation. However, the punitive focus of the administration’s statements seems to foreclose this option.

While traditional allies such as Israel and Canada are among the countries that vote with the United States the majority of the time, smaller countries such as Micronesia that thrive in the UNGA’s “one country — one vote” system also offer high levels of support for the United States.

More generally, friendly countries are those that support the United States’ positions at the UNGA on at least 1 out of every 4 votes. Such a low bar to clear the “friendly” threshold underscores how relatively few friends the United States has.

Can a friends-only approach to foreign assistance work?

The Trump administration’s plan seems to focus on rewarding friends, as defined by UNGA voting. But most of those friends either don’t need foreign assistance — Israel being the exception — or lack the political and tactical significance of strategic partners. This approach may help sway some smaller countries on particularly important votes, but it remains unclear what tangible benefits will result.

Even if the Trump administration follows through with a drastic shift in long-standing support for Pakistan, the change is unlikely to affect Pakistan’s behavior. And Trump has made clear his desire to continue aid for strategic partners such as Jordan and Egypt, where the United States needs cooperation on specific issue areas such as counterterrorism.

Over the long term, the effectiveness of Trump’s foreign policy won’t come from tallying how countries vote on nonbinding resolutions at a historically unfriendly institution. It will be measured by how the United States develops relationships with key allies and partners and the Trump administration’s ability to shape real-world events in the United States’ favor.

Jessica Trisko Darden is an assistant professor at American University’s School of International Service and Jeane Kirkpatrick fellow at the American Enterprise Institute. Follow her @triskodarden.