For more than a decade now, U.S. soldiers have been laboring under a sad paradox: even though the United States enjoys unprecedented global military dominance that should cow enemies mightily, it has found itself in constant combat for longer than ever before in its history, and without much to show for it. Of the U.S. military actions in Kosovo, Afghanistan, Iraq, and Libya, only the first can be counted a success.

Assessing this record is a particularly crucial task now, with U.S. defense policy caught between powerful opposing pressures. Frustration with unending war and with strong fiscal constraints has pushed public opinion sharply toward retrenchment. At the same time, frightening challenges in three critical regions are demanding yet more action: Islamic extremists have seized large swaths of territory in Iraq and Syria, Russia has intervened in Ukraine, and China is flexing its muscles in East Asia. Washington faces a choice: Should it leave these endangered areas to their own fates, or double down on intervention to set them right?

In deciding when to expend blood and treasure abroad, U.S. policymakers should learn from the United States’ recent experience of war, but they must take care not to learn the lessons too well. Overconfidence from success can breed failure, and becoming gun-shy from failure can cause paralysis. For example, the U.S. military’s stunningly quick, cheap, and sweeping victory in the 1991 Gulf War raised policymakers’ expectations about what force could accomplish at low cost, causing them to underprepare and overreach when the United States invaded Iraq a second time. In a similar way, most American leaders have drawn too firm a lesson from the wars in Afghanistan and Iraq: never put boots on the ground. With the deployment of regular army units effectively ruled out, U.S. policymakers who still want to use force have been driven to options that involve airpower alone. That approach may make sense in places where the United States wants only to nudge a conflict in the right direction. But when the goal is to determine its outcome, airpower alone is insufficient.

Risky as pulling any lessons from recent experience may be, policymakers should start with the following. First, the United States should fight wars less frequently but more decisively, erring, when combat is necessary, on the side of committing too many forces rather than too few. Second, the country should avoid fighting in places where victory depends on controlling the politics of chaotic countries, since local politicians will rarely do what Americans want when that differs from their own aims. And third, Washington should give priority to first-order challenges, focusing its military planning on fighting wars with great powers and focusing its diplomacy on preventing them.


The United States’ slide into perpetual war has made it clear that presidents should resist the urge to unleash U.S. combat power. Unconventional wars are not easily resolved, and all types of wars tend to require much more effort to win than originally anticipated.

When presidents do decide to use force, however, they should resist the urge to minimize it. Overwhelming force may not guarantee success in a ground war, but underwhelming force invites failure. Although President George H. W. Bush did not make the mistake of using military power indecisively, all his successors have. In 1999, the Clinton administration, in conjunction with NATO, launched air strikes against Yugoslavia, assuming that a few days of bombing would compel Yugoslav President Slobodan Milosevic to allow NATO troops to administer Kosovo. But Milosevic refused, and so the United States and NATO found themselves in a prolonged campaign without an exit strategy. Only narrowly did Washington avert failure: by the time Milosevic conceded, months after the war had begun, the consensus among NATO members for continuing the campaign was on the verge of collapse. In 2003, President George W. Bush’s secretary of defense, Donald Rumsfeld, slashed the size of the force that military leaders had initially wanted for invading Iraq, dismissing the estimate of Eric Shinseki, then the army chief of staff, that it would take several hundred thousand troops to secure the country. The ensuing anarchy after the fall of Baghdad killed any chance of forestalling the Sunni insurgency.

President Barack Obama made a similar mistake in Afghanistan. In 2009, he ramped up U.S. efforts there, but not as much as his military advisers had recommended. After haggling over the number of additional troops to send, the White House settled on 30,000, rather than the 40,000 that Pentagon leaders had requested. The result was a campaign that had less shock effect than it could have had, took a long time to unfold, and never was able to put simultaneous pressure on all the Taliban’s strongholds. Limitation for limitation’s sake is false economy. If Obama wanted to minimize the cost of combat in Afghanistan, he should have chosen the high-end option for a troop surge. It’s hard to see how that course could have led to anything worse than what the smaller and slower commitment did.

Half measures are naturally tempting for politicians, whose instinct is to find a compromise between the level of military action foreign conflicts require and what skeptics at home will tolerate. They see the military’s recommendations to commit massive forces as overly cautious insurance policies, and they overlook the risk that doing less will lead to stalemate rather than success. At least until the troop surge of 2007, four years of war in Iraq had resulted in stalemate, and after more than a dozen years, Afghanistan has still not moved into the win column.

Advocates of the light-footprint approach contend that a massive U.S. presence alienates locals and undermines the purpose of intervention. That’s often true, but it does not make the case for light action. First, maximal force must not be confused with indiscriminate targeting and increased collateral damage; rather, it means engaging the largest possible portion of enemy forces and territory at the same time to deny the enemy the ability to adjust to the pressure. Second, where a large U.S. footprint would indeed be counterproductive, Washington should simply stay out altogether, since there’s every reason to believe that a smaller footprint would do no more than stretch out the same costs of full-scale combat on an installment plan while leading to the same end result.


Exercising restraint need not mean enshrining the Weinberger and Powell Doctrines -- the set of rules about when the United States should go to war put forth by Secretary of Defense Caspar Weinberger in the 1980s and added to by Chairman of the Joint Chiefs of Staff Colin Powell in the 1990s. These guidelines advise limiting military intervention to cases in which it is the last resort, enjoys widespread public support, would further a vital national interest, would entail overwhelming force, and meets other restrictive conditions.

Nor does restraint mean neglecting the many tools short of conventional combat that the United States can and should still use: diplomacy, economic aid, covert action, special operations forces, military-training assistance, intelligence sharing, and arms supplies. Few conflicts in the post–Cold War era are worth the cost of massive U.S. military action, even if successful. But when the goal is not to resolve local conflicts decisively but only to boost client governments or eliminate particular plotters, then it may be appropriate for Washington to undertake limited actions. When they fail, the costs are as limited as the instruments.

Special operations forces, in particular, have assumed a much greater role than they used to have, and they now represent the primary military instrument in the fight against terrorism. They offer a way to act short of putting regular boots on the ground, and, as with airpower, they can be used and then withdrawn without admitting defeat. As a result, some strategists have promoted giving more and more missions to these forces, especially as deployments of regular forces are scaled back. Yet loading too many responsibilities onto special operations forces risks stretching them thin or diluting their quality by expanding their ranks. Moreover, special operations forces cannot function effectively or for very long unless they are supported by local logistical infrastructure for intelligence, supply, and transportation -- and this requires visible bases and extra personnel, which end up raising the profile that these forces were supposed to reduce.

With the ground wars in Afghanistan and Iraq having left a bad taste in their mouths, U.S. policymakers might also turn naturally to airpower. Bombing does work better now than it used to, thanks to advances in precision-guided munitions, which allow faraway targets to be destroyed with little risk and low collateral damage. Airpower can serve a purely punitive function, as when U.S. forces bombed Libya in 1986 in retaliation for Muammar al-Qaddafi’s attack on a Berlin disco. It can even work to tilt the balance between local forces, something the 2011 U.S. and NATO campaign against Qaddafi succeeded in doing. But the eventual outcome in Libya, with militias battling for control of the country, has proved terrible. Airpower can destroy things, but it cannot guarantee who will inhabit the presidential palace after the bombs have stopped falling.

Airpower on its own holds little promise for achieving larger strategic ends. Consider the idea of conducting U.S. air strikes on Iran’s nuclear infrastructure. Those who advocate that course offer no convincing reason to believe that the temporary benefit of delaying Tehran’s development of a nuclear weapon would outweigh the costs of an attack: Iran would probably work even harder to acquire a nuclear arsenal, collaborate more with North Korea, and retaliate against U.S. targets through its terrorist proxies. Air strikes would surely inflame anti-American sentiment among Iranians and the rest of the world’s Muslims, aiding the recruitment efforts of the Islamic State of Iraq and al-Sham (isis, or the Islamic State), al Qaeda, Hezbollah, and their ilk. The decisive way to keep Iran from going nuclear -- invading and occupying the country -- is out of the question. But air strikes represent an indecisive option, and the record of indecisive wars does not recommend that course. Better to dissuade Iran from building a bomb through negotiations and the threat of heightened sanctions, and then shift to nuclear deterrence if that effort fails.

As the political scientist Robert Pape has shown, airpower has worked when it has supported combat on the ground. But used by itself, it doesn’t force governments to surrender (with the 1999 campaign in Yugoslavia counting as the one possible exception). In any case, when it comes to defeating shadowy revolutionary movements during civil wars, airpower on its own can accomplish little beyond attrition around the edges. And messy civil wars are so far the most common challenge in the twenty-first century, and the type of conflict least easily settled by overwhelming conventional military power.


Military intervention in an unstable country can work if it reinforces the development of a stable political system -- but that depends on the aims and actions of the country’s politicians. U.S. counterinsurgency strategy has recognized this truth in principle, but not always in practice. Too often, those implementing it have confused administrative accomplishments for political progress and overestimated Washington’s ability to guide the politics of a foreign country. The United States has encouraged democratization, only to see it have the perverse effect of empowering venal leaders and dividing, rather than unifying, a country. And U.S. counterinsurgency strategy has not succeeded at getting host governments to mobilize rural populations to actively resist insurgents. 

Campaigning in 2000, George W. Bush swore off nation building as a U.S. military mission. As president, however, he launched two wars that turned into the most ambitious nation-building projects the world had seen since the Vietnam War. Despite well over a decade of hard work, both now look like failures. The lessons seem obvious: policymakers shouldn’t overestimate the United States’ ability to create viable political systems, and they should be wary of launching a counterinsurgency campaign in a country where the government is too weak or corrupt to do the job itself. But these lessons are not as straightforward as they sound. After all, the U.S. government never set out to undertake a long campaign in either Afghanistan or Iraq, yet it wound up doing so anyway. Rarely do wars go exactly as planned.

Nowhere has that proved truer than in Iraq. The 2003 invasion was a self-inflicted wound, unnecessary from the start, but its backers in Washington thought it would end in an easy, conclusive victory, like that in 1991. When the conquest spiraled into disaster, policymakers found their strategy floundering as they struggled to democratize the country and as the U.S. Army endeavored to relearn the lore of counterinsurgency it had banished from its consciousness after Vietnam. As in South Vietnam, counterinsurgency in Iraq after the 2007 U.S. troop surge worked well enough to improve security for a while, but the government ultimately proved incapable of generating national unity.

Because democratization unleashed the social divisions that Saddam Hussein’s brutality had suppressed, it enabled bitter political conflict to explode and paralyzed the government. (Both Iraq and Afghanistan held elections that failed to result in the formation of a government.) Tutelage by U.S. proconsuls cannot fix these problems any more easily than U.S. politicians can solve partisan gridlock at home. The fortunes of military intervention too often fall hostage to political intrigues that the United States cannot control.

Hawks counter such a conclusion by arguing that the Obama administration could have solidified the tentative progress Iraq began making after the surge (and averted the onslaught of isis) if it had gotten the Iraqi government to agree to a residual U.S. military presence beyond 2011. The problem with this argument is that Iraqi Prime Minister Nouri al-Maliki opposed letting U.S. forces remain in Iraq precisely because he was planning to go his own way. And there is no evidence that American exhortation or occupation could have mended Iraq’s fissures. Even if some U.S. troops had stayed, they would likely have exercised little control and yet still been blamed for what went wrong. And any pacifying effect they might have had would likely have disappeared when they did eventually come home.

Unlike the campaign in Iraq, the one in Afghanistan started as a legitimate war of self-defense: the Taliban government refused to extradite al Qaeda leaders after 9/11. But like the Iraq war, the liberation of Afghanistan morphed into something else. When the government in Kabul, which had been installed in 2002 with U.S. help, proved corrupt and incompetent, the Taliban rebounded and the war became a struggle over control of the population.

As part of that struggle, the U.S. military adopted a sequential strategy called “clear, hold, and build.” U.S. forces are adept at clearing -- regular combat, in other words -- but they had to depend on Afghan forces to hold the cleared territory. To hold the territory, the Afghans needed to build a durable, honest, and responsive government that offered an attractive alternative to the Taliban. But the Afghans, under President Hamid Karzai, failed to do this. U.S. soldiers have performed well when it comes to providing village authorities with material assistance, but they simply aren’t equipped to enlist villagers in civic organizations, link their demands to effective government programs, or give them an incentive to get off the fence and fight when the Taliban come calling. Foreign soldiers can hardly be expected to succeed at those tasks if the host government fails to do them or, worse, victimizes the locals.

When working with client governments that have their own agendas, outsiders can rarely reshape political practices in ways that permanently undercut the appeal of rebels. As the military expert Stephen Biddle has argued, U.S. counterinsurgency doctrine has suffered from the illusion that host governments will share the aims of their U.S. supporters. Washington cannot suppress kleptocrats in those governments without walking back the very democratization U.S. intervention was supposed to promote. Not every counterinsurgency campaign is guaranteed to fail, but the most successful ones rely on indigenous, rather than foreign, forces and take some ten to 20 years -- longer than even the very patient American public is willing to endure.

Given the obstacles, any attempt to pacify a well-organized rebellion involves making a very risky bet. Even if U.S. policymakers come to recognize this truth, however, the problem will remain, because they rarely enter such entanglements deliberately. More often, political leaders act because they focus on the danger of failing to do something rather than the danger of doing something and failing. Realistically, then, U.S. military leaders will sometimes find themselves saddled with missions they consider wrong-headed. Thus, the U.S. Army and the Marine Corps cannot afford to repeat the mistake they made after Vietnam, when they deliberately did not plan for intervention in messy civil wars. Once extricated from Afghanistan and Iraq, those services will need to retain at least a nucleus of capability for counter-insurgency, in case civilian authorities throw them into such a morass again.


Looking at the mistakes of the past dozen years, it’s easier to draw lessons about which military commitments Washington should avoid than which it should embrace. When determining what to plan for militarily, policymakers should ask two main questions. First, how important are the interests at stake? Second, how effective can military power be at protecting these interests?

The answers to those two questions suggest that it’s time for the United States to refocus its priorities on planning for conventional interstate wars. The United States’ top priority should be the defense of long-standing allies in Europe and Asia. This task became largely passé as the Cold War gave way to a long holiday from great-power conflict, but recent events have ended that holiday. Compared with the chaotic internal wars of the Middle East, the threats in Europe and East Asia are more suited to the application of conventional force. In these regions, the United States’ unparalleled conventional strength can act as a powerful deterrent. The United States has a successful record of preparing for and preventing conventional wars, and this should be its focus once again. But that will mean being ready to put plenty of boots on the ground.

Unlike Europe and much of the rest of Asia after the Cold War, the Korean Peninsula never stopped being a uniquely dangerous place, because the regime in Pyongyang has regularly engaged in reckless provocations. Defending South Korea against an attack by North Korea, however, would be as manageable as conventional wars get: the front to cover is short, U.S. and South Korean forces are modern and proficient, and their North Korean counterparts are daunting in number but threadbare in quality and especially vulnerable to airpower. Although North Korea’s nuclear weapons complicate matters, its nascent nuclear stockpile remains too small to counter a U.S. retaliation and thus could not dominate a contest in escalation, which would be the prime concern in wartime. The Kim regime still knows that any use of nuclear weapons would guarantee its demise at Washington’s hands. The weapons North Korea possesses can deter the United States and South Korea from invading the North, but they could not enable the North to conquer the South.

If North Korea ever comes to act like just a normally nasty state, the United States could consider withdrawing its ground forces from the peninsula, limiting its presence to air forces that could act immediately, and planning to reintroduce ground forces only in the event of war. This was a bad idea when U.S. President Jimmy Carter proposed it in the 1970s, when the military balance was less favorable for South Korea, but it is a bad idea whose time has almost come. Today, the South grossly outweighs the North in terms of conventional military technology, economic resources, and population, even as it spends only half as much of its GDP on defense as the United States does. Yet the time for the United States to reduce its subsidy for South Korea’s security is not quite ripe: the current regime in Pyongyang is so reckless that any scaling back of the U.S. presence might be misperceived as a sign of weakness, increasing the risk of a North Korean miscalculation. So the task on the peninsula remains about the same as it has been for over 60 years.


Even before it can get unstuck from the Middle East, the United States needs to deal with even higher priorities than the Korean Peninsula. For a quarter century, Washington had the luxury of concentrating on second- and third-order challenges: rogue states, medium-sized wars, terrorists, peacekeeping operations, and humanitarian relief. But the time has come to focus again on first-order dangers. Russia is back, and China is coming. These prospective opponents could not just hurt U.S. allies; they could inflict epochal damage on the United States itself. The main strategic task has therefore become a diplomatic one: find a modus vivendi with each country that stops the drift toward collision. If political accommodation fails, however, it will fall to the U.S. military to deter or defend.

In its beginning, the crackup in Ukraine was caused hardly more by Russian President Vladimir Putin’s aggression than by unthinking Western provocations, including unbridled NATO expansion, the humiliating dismissal of Russia as a great power, and the EU’s efforts to convince Kiev to cut its ties to Moscow. That is water over the dam, and Russia’s actions in Ukraine have complicated the options for accommodation. But there is no constituency in the West for going to war over Ukraine, and so ratcheting down the new cold war in Europe may require a political compromise that involves more autonomy for eastern Ukraine and a neutral foreign policy for Kiev. That said, Russia’s actions have resurrected concern for protecting NATO’s own members, especially the new ones. If Russia’s probes continue, the pressure to deploy NATO forces in Poland and the Baltic countries could become irresistible. Then, the conflict with Moscow would sharpen rather than ease.

Fortunately for the West, if the need for deterrence and defense in Europe does become a serious issue again, the mission will be far easier than it was during the Cold War. The balance of both military power and geographic vulnerability, which seemed perilous to NATO back then, now overwhelmingly favors the alliance. No longer can Russian generals dream of a blitzkrieg to the English Channel. Even without the risk of nuclear escalation, Putin should know that attacking a NATO country would be suicidal. The Russian economy is also in no shape to fund a truly serious military buildup.

Although the U.S. Army, faced with a future of diminishing missions, might see a lifeline in revived demands for it to play a greater role in NATO, Washington should let its European allies take care of any needed increase in military capabilities on the continent themselves. Moscow is nowhere near as wild and crazy as Pyongyang, and most NATO allies currently spend less than half of what the United States does on defense relative to GDP. Washington can remain fully committed to the alliance without getting dragged into expensive posturing by critics who whine about U.S. credibility.

China poses a bigger potential threat. The country has made new territorial claims throughout the waters of East Asia, especially to islands over which Japan also asserts ownership. Here, however, shifting the burden to the United States’ rich ally in the neighborhood would be a bad choice. For historical reasons, Japan still evokes intense antipathy in China -- and also in other countries in the region. If Japan were to step up as a normal great power, the unsettling effect in the region would not be worth whatever money the United States would have saved by reducing its military role there.

U.S. political leaders have not yet made clear choices about exactly what circumstances would warrant a war with China, but U.S. military leaders have been busy thinking about how to conduct such a war. The Pentagon’s “air-sea battle” concept is aimed at fighting China, and it emphasizes the advantages of cutting-edge technologies. This is good news for the U.S. Navy and the U.S. Air Force, which would lead a fight against China, but given how astoundingly expensive such high-tech weapons are, it is bad news for budget cutters. Washington should concentrate on defusing the growing conflict with China, but if events raise the priority of deterring Beijing, hopes for cutting defense spending will be frustrated.

The Pentagon’s plans also make cyberwarfare a priority, but that poses problems, too. As has all modern society, the U.S. military has become utterly dependent on a complex system of computer networks. Since the world has never seen a great-power war in the information age, however, there has been no test showing what unrecognized vulnerabilities this dependence may conceal -- and thus the United States cannot be confident that its superiority in traditional combat power will hold up against innovative cyberattacks. The recent era of long wars against backward enemies can tell policymakers little about how to protect against high-tech surprises. 


Choosing a strategy requires making judgments about what objectives are worth risking the lives of American soldiers and foreign civilians. Yet in Afghanistan and Iraq, U.S. officials underestimated the total costs, in blood and treasure, which kept ballooning after the ventures began. Many of the benefits they anticipated failed to arrive or weren’t that valuable to begin with -- a particularly grave error, since costs become less acceptable the further the benefits get from promoting strict national security. The most tragic overreach, attacking Iraq in 2003, actually damaged U.S. security by multiplying enemies in the Muslim world. All these miscalculations flowed from the heady ambitions of both Republicans and Democrats who wanted the United States to not just reverse local injustices but also enforce order worldwide, often at the point of a gun. They acted surprised when their seemingly generous objectives were met with fierce pushback. 

In the recent era of permanent war, the United States has learned hard lessons about the ambitious use of U.S. military power for secondary purposes. The United States’ overbearing military power encouraged civilian leaders to focus more on the desired benefits of military action than on its potential costs. Global primacy still gives the United States more room for maneuver than it had during the Cold War, but as tensions with Russia and China mount, that room becomes less than it was during the past quarter century. Now the United States needs to temper the ambitions unleashed by its post–Cold War dominance, not only in reaction to the setbacks it has experienced in small wars but also to prepare for bigger wars for bigger stakes against bigger powers.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • RICHARD K. BETTS is Director of the Saltzman Institute of War and Peace Studies at Columbia University, an Adjunct Senior Fellow at the Council on Foreign Relations, and the author of American Force: Dangers, Delusions, and Dilemmas in National Security.

  • More By Richard K. Betts