What national interest does the United States have in Afghanistan? According to recent polls, more and more Americans doubt there is any--or at least enough to warrant escalation of the war. This flagging support partly reflects the inadequate job the Obama administration has done explaining its goals and strategy in Afghanistan to a skeptical public. But it also reflects the underappreciated fact that succeeding in Afghanistan and defeating America's enemies there, as important as that is in its own right, is even more so for its effects in shaping Pakistan's future. That is the bigger prize. Put it this way: If we achieve the best case outcome in Afghanistan--a relatively secure and democratic country free of al Qaeda and its allies--but fail to prevent the worst case outcome in Pakistan--a failed or Talibanized state with nuclear weapons--we've still lost.

Of course, the importance of Pakistan has not gone unnoticed. It has spawned the shorthand "Af-Pak"--the idea being that the principal U.S. mission is to defeat al Qaeda in Afghanistan, and that requires stabilizing Pakistan. In truth, it's the other way around: It is Pakistan that is the swing state of South Asia, whose success could propel the region's development and whose failure could mire it in even worse instability. Shaping a positive future for Pakistan requires a new approach to the problem--and central to that is not abandoning the war in Afghanistan, but winning it.

Pakistan is not only a first-order foreign policy challenge for the United States--the nexus of terrorism, nuclear weapons, Islamic extremism, democracy promotion, and the geopolitical intersection of South and Central Asia and the Middle East--it's also the hardest to shape for the better. For all of the U.S. assistance to Pakistan and involvement in its internal affairs over the last decade, our influence remains limited and indirect. We don't have troops on the ground or the leverage to define outcomes directly. We don't have, for security reasons, much opportunity to move around the country, which is essential to conducting broad-based diplomacy and monitoring development programs. We don't have the trust of the Pakistani people, who view America--not unfairly, given recent history--as a fickle ally that only wants a transactional relationship and will abandon Pakistan as soon as it is convenient. And most important, we don't necessarily have common interests with the dominant actor inside Pakistan--the military--which supports U.S. enemies and is hostile to U.S. goals, especially in Afghanistan.

Indeed, the American vision for South Asia is at odds, even in conflict, with that of Pakistan's military. We envision an Afghanistan in which a legitimate, democratic state is capable of defeating the insurgency and fostering opportunity for its people; the Pakistani military views Afghanistan as a source of strategic depth against India and backs Taliban fighters to achieve it. The United States envisions a global strategic partnership with a rising India and an enduring peace between New Delhi and Islamabad; Pakistan's military views India as an existential threat to be countered asymmetrically with terrorists based in and backed by Pakistan. Finally, the United States seeks a comprehensive relationship with a civilian-led, democratic Pakistan to fight extremism and expand justice and prosperity for its people; elements of the Pakistani security establishment undermine exactly these goals by meddling in politics and sponsoring terrorism.

This is in some respects a zero-sum game. The realization of U.S. goals in South Asia would necessarily come at the expense of the Pakistani military's interests, as it currently defines them--for it would undercut the military's political legitimacy and privileged place in Pakistani society. The United States thus finds itself in the unenviable position of having major strategic interests in the future of Pakistan, while the internal actor with the most influence over that future is a military establishment with little incentive to cooperate fully with U.S. objectives. We can neither compel the Pakistani military to decisively change its behavior nor impose change directly ourselves.

How, then, do we further U.S. interests in Pakistan and South Asia?

For too long, the U.S. approach to Pakistan has swung sharply between two extreme assumptions: One, that the Pakistani military is really a U.S. ally with which we have common interests and to which we should defer to do the right thing; the other, that Pakistan's military leaders are hostile to U.S. goals, but that we can directly compel them to work against their own interests. The former overstates the enlightened self-interest of our Pakistani "partners"; the latter our ability to change their behavior.

A better approach would deal with Pakistan as it is, not as we might want it to be. This means accepting that those calling the shots in Pakistan are not our natural allies. Nor can we change their worldview directly through our actions and appeals or top-down dictates. Rather, we must work from the bottom up, creating incentives that could lead Pakistan's leaders to make new choices. Call it the indirect approach. The focus is less on sweeping initiatives to change Pakistani thinking in one master stroke (like an elusive Kashmir deal) than on incremental steps to create new realities, new facts on the ground, to which Pakistan's leaders would be forced to adapt.

This is as much a problem of psychology as policy. Few nations are as paranoid as Pakistan--where it is a widespread belief that India is behind every setback, that the United States plans to seize Pakistan's nuclear arsenal, that America is out to undermine Pakistani democracy, and that Washington can't be trusted. This paranoia explains many of Pakistan's policies: why America is treated as both ally and adversary in national security planning, for instance, or why Pakistan sponsors the Taliban in Afghanistan as a hedge against both New Delhi and Washington.

Some of this paranoia is baseless; some less so, especially considering past U.S. support for Pakistani dictators. But this misses the point: Pakistani paranoia is real in the minds of the country's leaders, so we must deal with it as a fact whether we like it or not. We should not dismiss Pakistani insecurity or try to disabuse the country's leadership of it through rational discourse. Rather, we should take Pakistan's paranoia as the point of departure for our policies and pursue incremental but sustained actions to create new facts on the ground that might lead Islamabad to alter its strategic calculus. This would be a slow, systematic, and evolutionary--not revolutionary--approach to changing the strategic context of Pakistani decision-making and so nudging Pakistan in a direction more favorable to American interests in South and Central Asia.

What would the elements of such a strategy be?

First and most important would be defeating Pakistan's terrorist proxies in Afghanistan--and, as much as possible, in Pakistan. Despite new and welcome Pakistani pressure on militants fighting within the tribal regions, there is still little evidence that Pakistan's military has broken its longstanding Faustian bargain with Taliban fighters in Afghanistan, viewing their fight instead as the ideal path to Pakistan's own security. The best chance to end Pakistani support for terrorism in Afghanistan--and India--is to demonstrate in a visceral way that its allies are losing and that if Islamabad does not stop backing them, it will end up with little influence in Afghanistan and less security as a result. Only once the facts on the ground change might Pakistani policies follow suit.

This suggests an opposite way of thinking about the "Af-Pak" challenge from the one the Obama administration has presented. We should not assume that Pakistani cooperation holds the key to progress in Afghanistan, and that the former is a prerequisite for the latter. To the contrary, the United States and its Afghan and NATO allies can make significant gains in Afghanistan without the Pakistani military choosing to fight cross-border militancy rather than support it--a choice it likely won't make anyway.

What's more, the benefit of defeating the Pakistani-backed enemies of Afghanistan goes beyond its value as a prerequisite for Afghanistan's stabilization and development, as important as that is. Defeating the Taliban would help secure the greater U.S. goal: fostering a realization among the Pakistani military that their terrorist allies, washing in waves over the Pakistani border, are breaking against a hardened Afghan state, and that to continue backing violent extremists will only leave Pakistan isolated, insecure, and weakened. Ending Afghanistan's status as a playing field for Pakistan's military-intelligence forces will also weaken the security establishment's influence in Pakistani politics, advancing the U.S. objective--shared by most Pakistanis--of strengthening civilian rule.

With President Obama's change of strategy and leadership in Afghanistan, the United States and its allies could now accomplish these goals. What remains is to put the necessary resources behind General Stanley McChrystal's new counterinsurgency plan, which is the best chance to achieve success as the United States should define it: a representative, self-sustaining Afghan state that can defeat the insurgency and al Qaeda without a permanent Western military presence. Beyond surging U.S. troops to protect the Afghan population and strengthening the country's governing institutions and security forces, a controversial yet critical component of this strategy is continuing to target and eliminate high-value Taliban and al Qaeda leaders through drone strikes in western Pakistan, which on balance has been an effective counterterrorism tool.

A second step: While helping our Afghan partners to grind down their Pakistani-backed enemies, we can also seriously engage with Islamabad over its security concerns in Afghanistan. We may not always view Pakistan's professed insecurities as legitimate, and many of them surely are not, but it is counterproductive to downplay or dismiss them for that reason. Instead, as Marin Strmecki has suggested, the United States, together with its Afghan allies, should seek "to draw out from Pakistani military and intelligence leaders what are their strategic concerns and how these might be addressed in a manner consistent with a strong and stable Afghanistan."

This might include establishing a transparent system of agreements to govern how regional actors like India and Pakistan exercise influence in Afghanistan. Another, simpler step would be closer consultation with Pakistan to take into account more of its concerns on issues of Afghan reconstruction, from building infrastructure to training security forces. Pakistani leaders must accept and reconcile themselves to the fact that regional powers--especially India--will remain important players in Afghanistan. But the most practical way to get Pakistan to accept this unpalatable fact, and to stop defining its national security as the destabilization of its neighbors, is to demonstrate that diplomatic avenues exist for it to peacefully shape the nature of others' influence in Afghanistan.

A third step is to help Islamabad and New Delhi normalize their fraught relationship--a massive undertaking, to be sure, but one that civilian leaders in both countries have been increasingly outspoken about dealing with urgently. The best way for the United States to help is not a high-profile attempt to mediate an elusive deal on Kashmir. This would only alienate both countries. Indeed, the best thing America can do is stay out of the way. After all, the greatest progress in Indo-Pakistani relations to date occurred in secret talks from 2004-07, which were enabled by the strong ties that Washington enjoyed independently with leaders in each country.

A better approach is to focus on small steps that help Pakistan and India overcome obstacles to peace by achieving incremental progress where their interests converge. This could include greater U.S. intelligence sharing with India and Pakistan bilaterally to counter Pakistani terrorists who threaten both societies. It could include increased sharing of technology to facilitate the transit of goods and people in ways that enhance border security. It could include more assistance for positive regional initiatives like a planned pipeline to bring natural gas from Turkmenistan to Pakistan and India via Afghanistan, and a framework for managing precious Himalayan water resources. And it could include support for Indo-Pakistani cooperation in building roads, railways, and other regional infrastructure that deepen economic interdependence in South Asia. Establishing new patterns of cooperation in these ways, rather than trying to force an unlikely ideological convergence, is the most realistic way to build confidence between these old rivals.

A fourth way to create new incentives that could lead Pakistan to alter its security policies is through an approach of "more for more" on military relations and supply. For decades, U.S. security assistance to Pakistan has been schizophrenic: At times, we write blank checks to the general staff; at other times, we renege on our commitments to supply the Pakistani armed forces with coveted hardware, even going as far as severing military-to-military relations altogether in the 1990s and imposing sanctions. After several years of playing Santa Claus under the Bush administration, there are now calls to revert to being Scrooge, including by denying Pakistan access to further supplies of its beloved F-16. This is no way to deal with a paranoid country that doubts America's commitment to its security. And it is certainly no way to convince Pakistan that its main security threat is not India, but the insurgency within its own borders, and that it should entirely reorient its military to conduct counterinsurgency, not just conventional warfare.

Because Pakistan's military leaders doubt that such a transformation serves their interests, they have to be goaded into it. That means addressing their insecurities head on and bolstering America's credibility as a reliable supplier of military training, education, and hardware to Pakistan. We should continue to supply Pakistan's general staff with the weapons platforms it covets, but we should do that as a way of taking off the table their argument that Pakistan lacks adequate defenses. The goal should be to ensure that Pakistan's military views America--not Saudi Arabia, China, or the Taliban--as its ally of choice. And we should use this conventional military assistance and training not just as reassurance but as pressure on Pakistan's military to adopt reforms and accept an expanded train-and-equip program that prepare it for counterinsurgency warfare. This approach would also help to inculcate in the Pakistani officer corps an understanding of the military's proper role in a democracy as a way of strengthening civilian government in Islamabad.

More paranoid and skeptical of the United States than even the Pakistani security establishment is the Pakistani public, as America's low public approval ratings show in poll after poll. That is why a fifth and final element of shaping a favorable Pakistani strategic outlook is supporting opportunity for the Pakistani people and helping them to strengthen the political and economic institutions that too often fail them. This is not simply a liberal ideal but a hardheaded national interest: Neither Pakistan's officer corps nor its civilian leadership will cooperate openly and systematically with the United States if prevailing Pakistani opinion rates Osama bin Laden higher than Barack Obama.

We won't change this dynamic through huge public diplomacy efforts to convince Pakistanis that America really is their friend; rather, we can win this support bit by bit, from the bottom up, through actions that tangibly benefit Pakistanis. It is worth recalling that the greatest outpouring of Pakistani support for America in recent memory occurred during the U.S. relief effort following the 2005 earthquake. That is why sustained U.S. reconstruction and development assistance for the several million refugees who have fled fighting in northwest Pakistan could make such a major contribution now, particularly given the Pakistani government's own shortcomings in offering assistance. The Obama administration deserves credit on this score: It has pledged more than $300 million in assistance.

More strategically, Washington can commit itself to the same scale and intensity of investment in Pakistan's economic development and civilian governance as it has to funding and equipping its armed forces. For too long, U.S. assistance efforts have been plagued by bureaucratic ineptitude, squandered by Pakistani corruption, and hamstrung by insecurity. But there is a model for U.S. development assistance that has succeeded under the hardest of all possible circumstances in Pakistan, in its remote and dangerous Federally Administered Tribal Areas (FATA). Under USAID's Office of Transition Initiatives (OTI), a small team of U.S. nationals working with hundreds of Pakistani colleagues have succeeded precisely by departing from USAID's traditional development playbook. OTI employs Pakistanis and buys its materials from local sources. It creates democratic mechanisms for local ownership of all its assistance programs. And it rigorously measures success in terms of outcomes, not outputs, using multiple means of accountability. Special Envoy Richard Holbrooke was reportedly so taken with OTI's work during an early visit to Pakistan that he expressed shock that its budget wasn't many times its current size. That would be a good start--to be followed by applying this small office's approach to development in the FATA throughout Pakistan.

Many of Pakistan's pathologies--its lack of territorial control, poor governance and development, weak rule of law, and overfunding of its military at the expense of other public priorities--are a function of the country's failure to build institutions that deliver for the Pakistani people. This creates a vicious circle in which U.S. security interests demand a partnership with Pakistan's military, which is far more capable than the country's civilian institutions, but the very act of our working around these civilian institutions only weakens them further. It also serves to exacerbate Pakistan's mounting Taliban insurgency and make it more difficult to defeat. That's why the United States can begin building a long-term partnership with Pakistan to strengthen the civilian institutions of its democratic state.

In this endeavor, as in all of the others listed above, the ultimate goal is to reorient Pakistan's views of its national security. This change will not be brought about through U.S. pleading or persuasion; Pakistan's civilian and military leaders must choose it themselves, and America can increase the likelihood that they will do so by creating new incentives, new realities, new facts on the ground that militate in favor of changing Pakistani attitudes. This is an indirect effort, but it can be effective.

Pakistan is a paranoid country that conceives of its national security in ways that are destructive to its neighbors, to us, and to itself. The United States should deal with Pakistan as it is, but we should not accept that it has to remain that way. Shaping a more positive future for Pakistan--the nuclear armed, insurgent-plagued, swing state of South Asia--is possible, necessary, and extremely urgent. It is what is most at stake in the "Af-Pak" conflict. And it is why the idea of losing in Afghanistan should be unthinkable.

Christian Brose is senior editor of Foreign Policy. Daniel Twining is senior fellow for Asia at the German Marshall Fund. Both served on the State Department's policy planning staff during the Bush administration.

Next Page