THERE OBVIOUSLY isn’t one reason why the United States has struggled to meet its evolving priorities in the Middle East; there are many. So to assign blame solely to U.S. security assistance for what ultimately are policy shortcomings is simplistic and misleading.

First, it should be acknowledged that the United States and its Arab partners have dissimilar threat perceptions and security priorities, which will inevitably impact the effectiveness of U.S. security assistance. Second, each Arab partner has a unique sociopolitical context and threat environment. And each has tackled its security challenges differently, depending on its own perceptions, priorities and resources.

With respect to Iran’s regional challenge, for example, it is no secret that the Arab partners do not view such a challenge similarly, nor do they agree on how to address it jointly. For some, Iran poses an existential threat. For others, it is an irritant but a manageable concern. Others even view Tehran’s regional policies with indifference. No amount of U.S. arms, funds and training can change these realities.

Political and cultural differences between the United States and its Arab partners help explain why these partners have utilized U.S. security assistance in ways that, from Washington’s point of view, seem unwise and ineffective. U.S. officials believe that many Arab partners insist on purchasing large and high-tech U.S. weapons that do not address pressing security needs and requirements or advance shared goals. Supersonic aircraft and advanced missile defenses, for example, are less relevant for Iran’s irregular warfare or terrorist insurgencies than attack helicopters, unmanned aerial vehicles and smaller combat search-and-rescue planes.

Given Iran’s weapons smuggling and other nefarious activities on land and at sea, it is mind-boggling how most Arab Gulf states have not made proper investments in their coast guards, special-operations forces and various local law-enforcement agencies. Instead of developing minesweeping capabilities to prevent the Iranians from mining the Strait of Hormuz, for example, some Arab Gulf states have been obsessing over the F-35, the latest generation of American fighter plane, even though they already possess considerable air power. It is true that some Arab partners have made smarter acquisitions lately. And it is also true that some partners increasingly realize the importance of formulating official and viable arms-acquisition strategies, and appreciate the value of military education and training. But this general preference for prestige and modernity, instead of effectiveness and integration, is not likely to go away anytime soon.

This is not just about a lack of military expertise or “wrong choices” by Arab partners concerning the types and quantities of weapons they acquire. Arab partners often make political decisions regarding certain arms sales that have nothing to do with national security. This is when factionalist considerations, which Washington has little understanding of and cannot control, come into play. For example, if an Arab leader buys the best weapons and authorizes the best training for one particular military service (for example, the air force) but deliberately ill equips another service (for example, the army) it is due to political dynamics that are tied to his or the ruling family’s perceptions of regime stability, or even survival.

Arab partners also are sometimes unwilling to cooperate with Washington on delicate security affairs. That’s perfectly understandable. Internal security is a profoundly sovereign decision that involves a country’s most intimate secrets. As close as relations between the United States and its Arab partners have been historically, there is a strong and legitimate inclination on the part of these partners to conceal various things about their politics and societies. Such Arab resistance to sharing sensitive information is also amplified whenever relations with Washington experience political turbulence, as was the case with most partners during the Obama presidency. U.S. security assistance is, at the end of the day, a function of bilateral ties. The healthier these are, the more effective U.S. security assistance will be. One would only have to look at America’s security-assistance experience with its NATO allies, which shows that the process of consultation and coordination on security and defense affairs is more organized, institutionalized and therefore effective than with Arab partners (although still not where it should be).

On the American side, the problems of U.S. security assistance to Arab partners are equally basic, and have little to do with security assistance per se. The source of all security-assistance ills is not the amount of funding, the quality of training, the speed of U.S. weapons delivery, or the type or quantity of arms that Washington provides. It is the often-broken U.S. policy toward the recipient country that profoundly undermines the entire enterprise. U.S. security assistance has become an end in itself rather than a means, often driving U.S. policy as opposed to the other way around. Every time a training program breaks down, or an arms transfer leads to unintended consequences, U.S. officials and military officers with security-assistance responsibilities are drilled by members of Congress and blamed for the failures. Of course, these administrators of security assistance are too courteous and loyal to their supervisors to state the obvious: things fell apart not because a U.S. general did something wrong or there wasn’t enough money, but because, more often than not, the policy toward the Arab partner was incoherent.

Consider Egypt. We finance a good bit of the country’s military requirements and support its economy, yet we tolerate its government’s political repression, which contributes to Islamist radicalization. In Syria, we said that Assad must go, but did the absolute bare minimum to make that happen. We armed and trained a handful of rebels so they could fight terrorism, which Assad helped create, yet we ceded the terrain to Iran and Russia, whose heavy-handed interventions ensured Assad’s survival. In Iraq, we helped build the country’s law-enforcement agencies after Saddam, in an attempt to help the Iraqis figure out their politics in relative peace and security. Yet we did nothing to limit Iran’s growing influence over Baghdad. Worse still, we tolerated the proliferation of IRGC-controlled Iraqi militias that recently used U.S. equipment to recapture Kirkuk. In Yemen, we politically and militarily enabled Saudi Arabia’s war against the Houthis, even though we knew it would be a major distraction from the fight against ISIS and Al Qaeda, and offer Iran influence in Yemen that didn’t exist before. Instead of fixing the policy, which should be the author of security assistance, we either have thrown more money at these problems, added more layers of bureaucracy, or overlitigated and politicized the process.

One immediate consequence of an unclear policy toward the recipient country is confusion in the U.S. government over who does what and when in the security-assistance process. The most obvious example of this tension is between the Department of State (DOS) and the Department of Defense (DOD). Bureaucratic infighting is commonplace in any government, and often attributed to rivalry or competition over power and resources. But that is not always the reason. The root cause can also be, especially in the context of security assistance, the lack of a viable policy toward the recipient country. The main reason why the United States defeated the Soviet Union in Afghanistan, for example, is not because the mujahideen of the past were braver or better warriors than today’s Arab rebels, or because we armed the Afghans with Stinger missiles (okay, that helped a lot). It’s because U.S. policy at the time—rollback of communism—and the strategy—containment—were clear and consistent, which made for effective security assistance.

Numerous congressional hearings have been held, and even more reforms enacted since 9/11, to try to improve DOD-DOS coordination on security assistance. But all the debates and tactical fixes have essentially missed the mark. There is no doubt that some improvement in interagency coordination has been made in recent years, and especially since the 2012 Benghazi incident, when terrorists attacked and burned the U.S. mission in Libya, killing Amb. Christopher Stevens and three other U.S. nationals. Unlike in the past, there is now legislation—Section 333 of the National Defense Authorization Act for Fiscal Year 2017, or FY17 NDAA, is the latest example—calling for synchronization and joint planning between DOS and DOD. A new steering committee has been established to oversee such joint planning. Further, the secretary of defense now must seek the concurrence of the secretary of state before launching some, but not all, security-assistance programs with partner nations (exceptions include Ukraine and counter-ISIS assistance).

However, optimal interagency coordination won’t happen absent a workable policy toward the country receiving such assistance. It also won’t happen if DOS continues to exhibit leadership deficiencies and organizational weaknesses. Today, a major opportunity for DOS to get more involved seems to present itself: both the Congress and DOD affirm that they have provided DOS with the necessary resources and authorities to lead this process (although DOS still assesses that section 333 of the FY17 NDAA consolidates DOD’s security-assistance “fiefdom” and that Foggy Bottom still lacks the necessary resources and authorities to take the lead as it did in the past). One would also be hard pressed to find a secretary of defense who values and champions interagency coordination more than James Mattis. Yet despite its absolutely crucial role in security assistance, many in the Pentagon and Congress believe that DOS does not appear to be ready yet to assume larger responsibilities in the process. That’s especially damaging for the future of the security-assistance enterprise.