The United States has failed to get South Asia right.

In India, the U.S. was caught off guard by New Delhi's refusal to revise legislation that would have permitted American firms to bid on projects in the immense nuclear market. This was followed by an Indian decision to exclude two American companies from the $10 billion competition for a multi-role combat aircraft. Both developments were crushing disappointments to those who had expected these deals to be the capstone of a new strategic partnership.

In Pakistan, the United States tracked down and killed Osama bin Laden without help from the government. Despite repeated denials, this "non-NATO" ally had been hosting Osama bin Laden for years in a small city notable for its military installations. The jury is out on Islamabad's exact role, but either of the two likely scenarios—a benign inability to capture, or active protection—casts doubt on the value of a decade of almost unconditional American inducements and support.

In Afghanistan, almost ten years after vanquishing the Taliban, there is still confusion about strategy. Should we continue with our counterinsurgency efforts, or move on to a more limited counterterrorism strategy? We still have no idea what role Pakistan will play in Afghanistan's future, let alone India, which already has a large economic role there. Reducing our assistance to Pakistan, as announced last week, may put additional pressure on Islamabad to perform, but it is just another isolated measure with few prospects of having any long-term effect.

There are several reasons why American policies towards India, Pakistan, and Afghanistan have gone awry. One is that the United States lacks a clear conceptual understanding of what it calls "Af-Pak." Additionally, Washington is sub-optimally organized to think strategically and coherently about the area. Both of these insufficiencies are rooted in a wrong "theory of the region" which has led the United States into serial blunders.

For many decades, American policy towards India and Pakistan was derived from a Cold War framework. India was incorrectly seen as a Soviet ally just as Pakistan's reliability as an American ally was misjudged. This was a costly mistake because it not only neglected our overlapping interests with India, it ignored Pakistan's ability to exploit US tolerance as it covertly built nuclear weapons and nurtured a terrorist network that now poses a major threat to itself, India and the world.

Then, even as our Cold War imperatives started to wind down, we failed to prevent both countries from entering into a nuclear arms race and never confronted the one country—China—that was singularly responsible for Pakistan's proliferation. The United States was thereafter unable to stop Islamabad from turning into the world's nuclear ATM machine. At the same time, even while creating an exception for Israel, America dogmatically argued that the universal and treaty-bound approach to nonproliferation was the only way and scoffed at Indian efforts to manage proliferation regionally.

Finally, after 9/11 and the onset of the global war on terror, we hived off Pakistan from India and tried to de-hyphenate the two states, treating them as if they had no relations with each other except for the occasional crisis. This ignored a variety of historical, cultural and geostrategic imperatives that do tie the two states together, and it intensified our inability to take coherent decisions regarding the South Asian region.

These perceptional failures were compounded by faulty government organization. For example, the institutional setup in the military commands and the Defense Department perpetuate the India-Pakistan divide; the State Department is fragmented between the office of the Special Representative for Af-Pak and its South Asia bureau; and the White House has different reporting and decision lines for India and Pakistan.

The rise of India as a major power, the decline and possibly failure of Pakistan, increasing Chinese influence, and an unstable Afghanistan where we are entangled in a costly war cannot be managed without major organizational reform—including the creation of a new combatant command for South Asia and the Indian Ocean.

Organizational reform is a necessary but not sufficient adjustment. The United States also needs to correct course on three fronts.

More broadly, in the case of India, Washington must moderate expectations: New Delhi will not evolve into its new ally in Asia, like Japan. Our alliance with Pakistan will continue to stimulate Indian defense acquisitions from other suppliers—including Russia and Europe—as New Delhi will never want to rely on us to service their American equipment in case of a new conflict with Pakistan. The same reasoning applies to the 2008 nuclear cooperation deal: it improved relations, but did not make India an ally. New Delhi has a deep commitment to strategic autonomy, as indicated by its insistent use of the moderating prefix "natural" to describe its US relationship. In the end, India got what it needed from Washington, including recognition of its nuclear weapons program and support for its permanent membership on the United Nations' Security Council, at little or no cost.

America's Pakistan policy too needs to move from wishful thinking to a more strategic use of both carrots and sticks. American officials—civilian and military—persistently fool themselves by subscribing to the Pakistani definition of the relationship as suffering from a "trust deficit." Trust will come only if we (and Pakistan) can verify that the agreements entered into are being fulfilled. And it is not only Pakistan that has engaged in subterfuge; for example, by insisting on carrying out drone attacks in the Federally Administered Tribal Areas and by privileging the army as its main interlocutor in the country, the United States has often intensified the asymmetry in civil-military relations and undermined the progress toward full democracy. On the other hand, sheer trust without verification led to chronic Pakistani non-compliance, and from Pakistan's military point of view, an American failure to deliver as promised. It is already an asymmetrical relationship. The new US-Indian strategic alliance (even though still more symbolic than real), the failure of the Pakistan aid program to show results, and unilateral American military actions threaten to rupture the relationship entirely.

America also needs a fundamental course correction in Afghanistan. This policy now is based on the wrong assumption that this is a nation-building project that can succeed within the framework of a counterinsurgency strategy. This narrow view has neglected Afghanistan's broader geopolitical context, including the vital roles of Pakistan, Iran, Russia, China and India. This is why we are now reaping the harvest of an insistence on doing it alone. At the same time, we can no longer ignore that any acceptable solution for Afghanistan depends on a secure and supportive Pakistan, which, in turn, also depends on a stable South Asia and the normalization of India-Pakistan relations.

India is a friend, but not an ally. Pakistan is an ally, but not a friend. Afghanistan is everyone's problem. To pursue its interests in these three states, America needs to approach the region holistically, both in conceptual and organizational terms. With a new crew manning key positions on the ship of state, this is the best opportunity for a course correction.