As the United States’ quagmire in Iran makes clear, wars are almost always much harder to stop than they are to start. Over the past 80 years, less than 20 percent of the roughly 300 wars tracked by the Uppsala Conflict Data Program ended in a clear victory by one side. A slightly smaller percentage ended in negotiated settlement; more than 60 percent had no formal ending. And the longer a conflict continues, the harder it becomes to broker a settlement. In wars fought between 1945 and 1989, another study found, mediation succeeded 37 percent of the time when it began within the first one to three months of fighting but only 19 percent of the time when negotiations began a year or more into a war.

The United States’ two-decade war in Afghanistan and decade-plus fighting wars in Iraq were no exception. In the wake of those conflicts, Washington has tried to avoid getting ensnared in so-called endless wars by avoiding deploying boots on the ground. In some cases, that means the United States relies primarily on airpower, as it has in Iran—an approach that comes with its own military challenges. But in many others, instead of participating directly, the United States is equipping close allies to fight shared threats and joint adversaries. Between 2010 and 2025, Washington more than tripled its annual security assistance for allies, from $15 billion to $46 billion, and increased its intelligence support and arm sales.

The United States has backed several of its partners’ wars in recent years. In 2015, Washington began coordinating U.S. intelligence and military support for the Saudi-led coalition fighting the Houthis in Yemen. Since 2022, it has supported Ukraine in its defense against Russia’s invasion. And the United States has aided Israel throughout the war in Gaza following Hamas’s attacks on October 7, 2023. In all three campaigns, Washington supplied or sold the overwhelming share of the military hardware, intelligence support, training, and diplomatic backing. Yet its efforts had mixed results. The United States’ support greatly enhanced the military prospects of allies, but in every case, Washington found itself unable to shape how these wars were fought, and more critically, when and how they might end.

Supporting an ally’s war effort is often in the United States’ interest. But if Washington wants to participate in conflicts from afar, it needs a better strategy for maintaining influence over the primary combatants. It needs to enter with a clear idea of the war’s aims and how to achieve them. It needs to secure its partner’s commitment to the same objectives, to be honest with that partner about how far U.S. backing will go, and to set up mechanisms to ensure that the partner’s military adheres to American standards of conduct in battle. Ultimately, continuing to support allies’ wars without a shared strategy and without clear conditions for aid will not deliver the United States the advantages it seeks. It will only lead to more tension with close allies and more conflicts that don’t end.

Together with my colleagues at the Brookings Institution, I spent the past year interviewing more than 30 senior American officials, reviewing hundreds of public statements and documents, and consulting foreign officials to put together a comprehensive picture of how Washington engaged its partners during recent conflicts abroad and how those American efforts were received. What we found was that the United States tends to encounter many of the same problems when it supports its allies’ wars, even when the specifics of each situation vary quite a lot.

At the start of a conflict, Washington has much more power to shape a partner’s decisions than it does later in the war. When allies come under attack, they feel stunned and insecure, and they tend to seek Washington’s support in the belief that they will need U.S. military power to win their fight. Yet the U.S. government consistently fails to use its leverage early on. By the time it tries to push its partners to change operational designs, adjust tactics to minimize civilian harm, or come to the negotiating table, its influence has already eroded.

This failure can partly be attributed to the enormous political and moral pressure Washington feels to offer unconditional support when a close ally is attacked, as Ukraine was in 2022 or Israel was in 2023. Demanding strategic concessions from a beleaguered ally can feel like a betrayal. But holding back defers crucial conversations about goals, methods, and the limits of American aid.

Wars are almost always much harder to stop than they are to start.

Consider Ukraine. A few months into the war, the Biden administration publicly tied Ukrainian battlefield gains in May 2022 to the objective of improving Ukraine’s negotiating position with Russia, implying that Washington might expect Kyiv to start discussing peace terms before it could regain all of its land. But it is not clear that Ukrainian leaders understood or accepted the American strategy. As one U.S. official recalled, “It’s not that we didn’t take the time to think through the mission’s objectives—we did.” But Washington “recognized from the start” that its objectives of ensuring that Ukraine remained a viable sovereign state, maintaining NATO unity, and avoiding a U.S.-Russian war “were perhaps not the same objectives Ukraine had.” As a result, U.S. officials felt reluctant to push Ukraine toward diplomatic negotiations, especially while the Ukrainian military was advancing and congressional support for providing security assistance remained high.

Hard conversations with allies were also missing in the early days of the conflicts in the Gaza Strip and Yemen. The United States did not demand in the weeks after October 7 that Israel develop a “day after” plan for governance in Gaza once the Israeli military had sufficiently degraded Hamas—or even agree on what........

© Foreign Affairs