Why can’t the United States end allies’ wars?

Introduction

In the 21st century, intrastate and interstate wars persist longer than expected—and longer than comparable wars did in the 20th century. Even when the initial stage of intense conflict subsides, or one side declares a ceasefire, a conclusive and definitive “war termination” is rare. Instead, frozen conflicts, stalemates, and periodic pauses in fighting have become the norm. Over the past 25 years, the United States has learned that the long wars it began in the 2000s were hard to end. The U.S. military withdrew from Afghanistan in 2021 after a painful exit and attempted a similar withdrawal from Iraq in 2011—though President Barack Obama sent U.S. troops back to fight ISIS in 2014 and 2015.

Partly because of the protracted nature of wars involving U.S. boots on the ground, U.S. national security strategy shifted in the 2010s toward building partner capacity, enabling partners to fight joint adversaries and threats without direct U.S. involvement. One would expect the United States to exert considerable influence over the duration of its partners’ wars, as well as over how those wars are conducted. After all, in wars fought by close U.S. security partners, the United States shares the same national security goals and typically provides an overwhelming share of the hardware, defense and intelligence equipment, training, and support that sustains these campaigns. Yet recent experience suggests otherwise.

This report addresses a central policy puzzle specific to the wars of U.S. security partners: Why is it so difficult for the United States to shape its partners’ decisions on war termination and to influence their operational design and conduct?

To answer this question, we examined four recent moments in partner wars: Afghanistan in the lead-up to the 2020 Doha agreement, the first few years of the Saudi Arabian-led coalition’s campaign in Yemen starting in 2015, Ukraine’s defensive war following Russia’s 2022 invasion, and Israel’s war in Gaza following the October 7, 2023, attacks. We conducted more than 30 interviews with senior officials at the National Security Council, the Departments of Defense and State, and the intelligence community who had been key decisionmakers at the time. These interviews were complemented by facilitated discussions among senior officials, extensive review of hundreds of public statements and documents, and consultations with foreign officials in partner capitals to understand how U.S. diplomatic messages were received. Asking participants the same set of questions across all four conflicts revealed recurring patterns, as well as instances in which interviewees offered differing accounts of the same high-level diplomatic and interagency meetings.

Although our research initially focused on four case studies, Afghanistan ultimately emerged as sui generis. Unlike the other cases, the United States had invaded the country and created the partner it was supporting militarily and politically (the Afghan National Security Forces and the Afghan national government). The United States under President Donald Trump then decided to end a war it had begun. Given these differences, as well as space constraints, this first report focuses on Ukraine and Gaza, drawing on Afghanistan and Yemen to illustrate broader trends.

Overall, we found similar patterns across U.S. partnered wars. When it comes to war termination, the United States unsurprisingly shares its partners’ interest in total victory. Yet the longer a war continues, the harder it becomes for the United States to influence how the war is waged and how long it lasts. Over time, U.S. officials come to favor termination through negotiated settlement as the option most likely to secure U.S. interests. As for influencing the operational design or conduct favored by the partner, the degree of U.S. influence waxed and waned at various points during each conflict.

Despite the inherent differences among the cases, we identified several additional consistent findings. First, the United States finds it difficult to agree on end goals with its partner at the outset of a war. In each case examined for this paper, early discussions between the United States and its partners were vague about expected political horizons, what the “day after” would look like, and the terms on which victory would be achieved. Many—if not most—of the 30 officials interviewed for this paper expressed regret over this lack of clarity in the initial diplomatic conversations, noting that it complicated subsequent diplomacy between the United States and its partners.

A second key finding across all four cases was that domestic political support for the partner’s war was initially high—measured by both general public and congressional opinion—but gradually diminished. Debates arise over the morality and legality of the decision to go to war and over the partner’s conduct in fighting it. Domestic opinion can shift regarding whether to continue supporting an enduring partner war. And while U.S. military, diplomatic, and intelligence officials are initially closely aligned with their counterparts in the partner government and military on strategy and operations, differences emerge over time. This divergence is in part inevitable since the United States does not exercise command and control over how the war is conducted. As the cases below reveal, however, U.S. decisionmakers’ frustration with their limited influence creates friction, eroding U.S. support for the campaign’s continuation and leading the United States to favor the war’s termination.

These findings point to a new way forward for U.S. security partners engaged in war—even a war that enjoys support at the highest levels of the U.S. government and among the American public. U.S. national security decisionmakers need to seize the maximum leverage available at the outset of a partnered conflict to negotiate, define, and clarify U.S. desired political objectives and end states up front, particularly since U.S. leverage over its partner tends to decline after the war’s initial stages. Over time, the partner is likely to remain far more committed to continuing the war than to ending it, at nearly any cost—and this divergence is likely to erode domestic and congressional support for continued U.S. assistance. U.S. security, intelligence, and diplomatic assistance should therefore be predicated on a complete, transparent agreement on the shared end state, along with a roadmap for concretizing U.S. expectations.

As U.S. leverage diminishes, the National Security Council and the president will need to maintain disciplined, coherent oversight over operations across all national security departments and agencies. When U.S. partners perceive cacophonous, inconsistent messaging across U.S. channels, it becomes easier for them to dismiss U.S. decisionmakers or play agencies off one another. And while every White House may find it easier to consult with the president through an exclusive senior circle on matters of partner war, this exclusivity risks marginalizing the interagency’s technical expertise and experience with partners—both of which are sources of U.S. leverage at each national security agency.

Finally, the United States cannot afford to press only its partner to end a war; it can and should design diplomatic strategies aimed directly at the adversary. Some officials may view direct diplomacy with an adversary as a betrayal of U.S. partners, but such diplomacy is necessary in partnered wars, where the adversary often regards the United States as a party to the conflict even if it has never fired a shot.