ON JANUARY 13, a U.S. missile strike on the Pakistani village of Damadola, intended to kill Al-Qaeda's deputy leader Ayman al-Zawahiri, missed its target but killed at least 17 other people, probably including Al-Qaeda members but certainly including local women and children. If it had succeeded, this would have been a notable coup in the struggle against Al-Qaeda. Instead, this violation of Pakistani territory has humiliated the administration of President Pervez Musharraf and compromised his government's assistance to the United States.

The Damadola incident illustrates a central dilemma in the War on Terror. It seems not only necessary but also just that the United States should retain the right to strike against acknowledged terrorists in those areas of the world where states cannot or will not take action.

There are, however, two central problems with this approach. First, most of the countries where large-scale terrorist activity is occurring are like Pakistan, where governments do control the greater part of their territory--just not all of it. They are not failed or even failing states, but functioning states suffering from certain weaknesses. Moreover, while their governments are allied with Washington, their populations are largely unfriendly to the United States.

The second, even more obvious point is that the United States and its allies cannot in fact invade and control such territories themselves. In the case of Pakistan, the United States will never have any choice but to work with and through a Pakistani government--almost any Pakistani government--if it wishes to exert any wider control over the fight against terrorism and extremism in Pakistan. This is especially true of the indirectly administered tribal areas that border Afghanistan.

But the importance of Pakistan goes beyond developments in Afghanistan or its possession of nuclear weapons. With some 160 million people, Pakistan has a population more than half the size of the entire Arab world. Many Pakistanis are tempted by radical Islamism. And of all the major Muslim states, Pakistan has some of the deepest underlying ethnic fissures and social problems, leading to a long-term threat of disintegration.¹

The danger of tactics like the Damadola attack is that a U.S. strategy that has the unintended consequence of weakening the Pakistani government's moral authority over its own population (and most especially its soldiers and police) may be utterly self-defeating. Even if such attacks succeed in killing leading Al-Qaeda figures, if their effect is also in the end to bring down the Musharraf government and bring to power a much less helpful or even weaker one, then the United States will have traded a limited tactical victory for a very serious strategic defeat.

The difficulty that Pakistan presents to U.S. policymakers, however, goes beyond that of tactics. This country challenges some dominant and paradigmatic American ways of seeing the world and thinking about history, democracy and progress--ideas that are common both to the Bush Administration and most of the Democratic opposition.

Pakistanand "Democracy"

UNLIKE MOST of the Arab world, Pakistan has had several prolonged experiences of democratically elected parliamentary governments in the past, from 1947 to 1958, 1971 to 1977 and 1988 to 1999. None of these democratically elected governments succeeded in lifting the country out of mass poverty, and some were economically disastrous. All civilian governments have been guilty of corruption, election-rigging and the imprisonment or murder of political opponents, in some cases to a worse degree than the military administrations that followed.

All of these periods involved serious unrest in some or all of Pakistan. The losing parties in elections have, as a rule, denounced these elections as rigged (which admittedly they often were) and encouraged the military to seize power. When the army eventually did so, in all three cases this was initially at least welcomed by a large majority of the population, utterly fed up with the experience of "democratic rule."

Given these three previous experiences, to argue that if formal democracy were to be reintroduced in Pakistan tomorrow it would be radically different and better, one must be able to present credible evidence that something fundamental about Pakistan has changed radically for the better since the 1990s.

It is true that the economy has grown very well in recent years, thanks to a mixture of write-offs of debt since 9/11, increased flows of Western aid and wise economic policies on the part of the Musharraf government. In 2005 GDP grew by more than 8 percent, on top of growth of 6.1 percent in 2004. This is the highest growth rate in Asia after China and compares to average growth rates of only around 3 percent in the 1990s. Nonetheless, given the continual rise in the population, present growth rates will take decades to lift a majority of the population out of poverty and create large, stable and progressive middle classes. And as in Pakistan in the 1960s, high economic growth that is very unevenly distributed can actually increase political unrest. Finally, as history demonstrates, there is usually a considerable lag time between economic change and the development of political structures that reflect that change.

Lack of political progress in recent decades has been generally attributed in the West, and by Pakistani liberals, to the military's repeated seizures of power. There is an element of truth to this, but it is also true that those interventions have usually occurred because the civilian political order has already broken down.

For democracy is representative not only of the people, but of all those classes, groups and institutions through which the popular will is refracted until it eventually finds some kind of distorted reflection in elected institutions. In other words, democracy usually reflects not so much "the people" or "the electorate" as the distribution of social, economic, cultural and political power within a given society. And if that power is held by groups like most of the Pakistani rural and urban elites, then the resulting democracy is not going to be a force for good governance, economic progress, respect for human rights, fair elections or even orderly transitions of elected government.

The nature of Pakistani society, and the weakness of real democratic development, is shown above all by the lack of real, modern, mass political parties. Without such parties, democracy is bound to be more or less a sham or facade for oligarchic rule--just as it has been in so much of Latin America. In Pakistan the only true national political parties are those of some of the Islamists. The parties routinely described in the Western media as "democratic" are in fact congeries of landlords, clan chieftains and urban bosses vowing more-or-less temporary allegiance to some national leader like Benazir Bhutto. As noted, all have been more than willing to adopt highly undemocratic methods when necessary. Most Pakistanis have fully accepted the form of democracy but are still far from truly accepting the content. The same set of attitudes prevails to a considerable extent with regard to the law, or at least laws derived ultimately from British rule and not sanctioned by religion or by tribal or communal tradition.

It would be quite wrong to see these features of Pakistan as reflecting simply the absence of "modern" values of democracy and the law. Rather, they also stem from the continued presence of traditions of overriding loyalty to family, clan and religion (often in a local form, which is contrary to the precepts of orthodox Islam as well as the Pakistani legal code) and to the rules of behavior that these loyalties enjoin. Similarly, "corruption" in Pakistan, as in so much of the world, is not the kind of viral infection instinctively portrayed by much of Western analysis. Insofar as it is entwined with patronage and family allegiance, it is an integral part of the system as a whole. Corruption cannot therefore be "cured." Rather, as in South Korea and other societies, it can only be changed organically into less destructive forms of patronage.

A genuine, even passionate belief in law and democracy in Pakistan therefore co-exists with a belief that these institutions are like the ropes around a boxing ring. They may help to limit the area of conflict, but they do not in themselves govern what goes on inside the ring.

That includes violence. During a visit to Sind in 1989, a member of a great local landowning and political family told me:

"This is a difficult country. If neighboring landowners see that you are weakening, there are always a lot of people to take your place, and they will hit your interests in various ways, like bringing lawsuits to seize your land or your water. If you can't protect yourself, your followers and tenants will ask how you can protect them. A semblance of strength must be maintained, or you're finished. The trick is to show your armed strength without getting involved in endless blood-feuds. . . . Such rivalries between families and clans are also conducted in the law courts, but the ultimate decision always lies with physical force."

This is a description straight out of the 15th-century English Paston letters--or from the old American world of the Hatfields and the McCoys. And the speaker, by the way, was no rural thug, but a senior official of a European-based bank, as highly educated and cultivated as were no doubt the Earl of Warwick or the Duke of Lancaster in their day. And as the example of the Hatfields and the McCoys suggests, these patterns, though old and deep, are not permanent or intrinsic. They can be changed radically, and Pakistan can be turned into a modern democracy--but this will take a long time and immense effort.

As to the Pakistani army, it sees itself as fulfilling the role of referee within the political boxing ring--but a referee, it must be said, with a strong personal interest in the outcome of many of the fights and a strong tendency to make up the rules as he goes along. This referee also has a long record of either joining in the fight on one side or another, or clubbing both boxers to the ground and taking the prize himself. This does not encourage respect for democratic boxing as a profession.

On the other hand, since the civilian boxers have not just boxed but often gouged, bitten, knifed, kicked, bribed the referee, corrupted the jury and encouraged their supporters in the audience to climb into the ring with baseball bats and knuckle-dusters, I must confess that I have never been able to summon up the paroxysms of righteous anger about military rule that are expected of right-thinking Western observers.

Nor can I feel too outraged about the military's diversion of state patronage into its own pockets, especially since a good deal of this does not stick with the generals but is circulated to the lower ranks, helping to keep them disciplined and loyal. Of course, if the civilian politicians loot the state, then the men with guns are going to take their share. But because the military have been schooled in a loyalty to the country that is missing from most of the population and are a disciplined and hierarchical force, their corruption tends to be somewhat more regulated and limited than the Pakistani norm.

Finally, and most strikingly, Pakistan--and indeed South Asia and much of Latin America--demonstrates the frequent irrelevance of democracy even in an area where we instinctively think that it makes all the difference, namely human rights. The truth is that the overwhelming majority of human rights abuses in Pakistan take place not on the orders of the state, but as a result of state weakness. They stem from a mixture of freelance brutality and exploitation by policemen, working either for themselves or local elites; actions by local landlords and bosses; and punishments by local communities of real or perceived infringements of their moral code.

This kind of behavior has continued unchanged whether democratic or military governments are in power. Western human rights organizations, given their obsession with state oppression and the protection of small numbers of "dissidents", find it extremely hard to recognize this basic fact, which is known to every inhabitant of a Pakistani or Mexican village.

This is also entirely characteristic of much of neighboring democratic India, as starkly revealed in a recent book about police and criminals in Bombay by Suketu Mehta called Maximum City (2005). Once again, this is a subject on which the U.S. media tend to be silent, as for example in the total lack of American reporting of an incident in the Indian state of Orissa on January 2, in which police shot dead twelve tribal people protesting at the seizure of their land. If this had happened in Pakistan or China, it would have been cited as another consequence of "dictatorship." But since it occurred in India, the world's largest democracy, it could not be fitted into any convenient U.S. paradigm and, therefore, in a certain sense could not be seen at all.

The Limits of Realism

BUT IF Pakistan reveals the grave limitations of the Bush Administration's pseudo-idealistic "democratization" strategy when it comes to the War on Terror, it is also true that classical realism is seriously challenged when it comes to Pakistan. For in one regard, the democratizers do have it right. They recognize that in the long term, addressing the threats to the United States from Pakistan means changing the internal nature of Pakistani society.

Classical realism--with its obsession with the relative strength of states, and also its instinctive tendency to see states as basically unchanging--has always had a problem recognizing the profound importance of internal factors. By far the greatest danger from Pakistan to U.S. security--and even to that of India, though most Indians naturally cannot be expected to see this--comes not from Pakistani strength but Pakistani weakness.

In the short to medium term, the threats to the existing state from Islamist militancy have been greatly exaggerated. Electoral support for the Islamist parties is concentrated among the Pashtuns of the North-West Frontier Province and Baluchistan, and makes up less than 15 percent of the Pakistani population. The army has shown again and again that it can control mass unrest in Pakistan as long as it does not spread to northern Punjab, from where most of the army's own soldiers are recruited. Even if Musharraf is--God forbid--killed or the army is forced by mass protest to replace him, the result will almost certainly not be an Islamist takeover, but as in the past some form of transition to limited democracy, managed by the military.

Then again, nuclear weapons are forever; and so will be the potential threat from Pakistan to the United States. I'd bet a large sum of money on Pakistan surviving for the next ten years. I would hesitate to bet on it surviving for the next fifty, if current demographic, social, economic, cultural and above all ecological trends continue, and if U.S. actions in the War on Terror continue to inflame the Pakistani population. If Pakistan does one day go the way of Somalia or other African states, then Al-Qaeda or its descendants will have tremendous opportunities to fish in the wreckage.

The traditional tools of realism, from direct U.S. military or economic pressure to the use of local U.S. allies (in this case, India) to "balance" or "contain" Pakistan, are of course not without value. After 9/11, this kind of U.S. threat was critical in forcing the Musharraf government first to cooperate in the war against the Taliban and Al-Qaeda, and later seriously to reduce Pakistani support for Islamist insurgents fighting Indian rule over Kashmir. Without this pressure, Musharraf himself might never have agreed to cooperate--despite his genuinely secular and national-liberal attitude to basic issues; and it would certainly have been quite impossible for him to get much of the rest of the military high command and Pakistani establishment to agree.

On the other hand, U.S. policymakers need to keep two things firmly in mind when it comes to trying to use India against Pakistan. The first is that India cannot control Pakistani internal developments any more than the United States can. In the end, only Pakistanis can rule Pakistan. The second is that very few Pakistanis, and virtually no Pakistani officers, are simply going to accept Indian hegemony in South Asia and trust in a hegemonic India to guarantee their security. They would regard this as not just treason but insanity.

It is indeed absolutely necessary to help convince the Pakistani military and civilian establishment that a continuation of the military competition with their much larger and richer neighbor threatens to ruin their state in the same way that the Soviet Union was ruined. But this should be done not through an overt U.S. security tilt towards India, but by working with both countries to help them find a settlement to the Kashmir conflict and to reduce tensions between them.

This need for balance applies especially to the Indian and Pakistani nuclear deterrents. From this point of view, the Bush Administration's relaxation of restrictions on India has been a very negative step. We should not forget that the nuclear arms race in South Asia was initiated by India, not Pakistan, and that to ask Pakistan to forego nuclear arms when India has them is impossible.

The problem was that, because Pakistan was in a greatly inferior position to India, and under greater U.S. pressure, its search for nuclear technology necessarily led it into deals with North Korea and other "rogue states" and increased the dangers of nuclear proliferation. To prevent Pakistan from again engaging in nuclear smuggling, it is of course necessary to face Islamabad with serious threats, but it is also necessary to work hard for a nuclear weapons freeze and confidence-building measures in South Asia.

U.S. policymakers therefore need to recognize two basic truths about dealing with Pakistan; and unfortunately, both of them, for different reasons, are extremely difficult for contemporary Americans to accept. The first is that the problems presented by Pakistan are not amenable to solution, but only to long-term management and gradual amelioration. The second is that, while the only policies appropriate to dealing with Pakistan are "realist" ones, this needs to be "developmental realism", focused on using U.S. assistance to help build up the country over a long period.

Developmental Realism

DEVELOPMENTAL realism draws upon U.S. strategies in East and Southeast Asia from 1950 to the 1970s, when very large sums in U.S. aid were directed to states from South Korea to Indonesia so as to strengthen their resistance to communist subversion. U.S. trade policy was also tailored to promote economic development by giving preferred access to U.S. markets.

Certainly, a good deal of corruption attended these efforts. But contrary to the perception of many Americans, this strategy was not only very successful overall in preventing the spread of communism but laid the basis for the development of democracy in the region. Moreover, in a country like Malaysia, the boost given to the local economy has helped Malays resist the lures of radical Islamism.

With regard to Pakistan, the biggest single focus of U.S. aid should be the improvement of Pakistan's water infrastructure, especially in the area of conservation and reducing the present appalling degree of waste. Water shortages present the greatest future threat to the viability of Pakistan as a state and a society. If present demographic and ecological trends continue, then in a few decades Pakistan will have 250 million people living in a country much of which will be as dry as the Sahara desert. In particular, given the dependence of much of Pakistan's agriculture on river-based irrigation, the melting of the Himalayan glaciers could face populations downstream with a literally existential challenge.

The second focus of U.S. aid should be helping to provide jobs. Improving Pakistan's educational system is important, but if this only produces unemployed and embittered graduates, the effect will be only to increase Islamist radicalism. And since the ultimate reason for U.S. aid to Pakistan is not charitable but political, it must bring visible benefits to ordinary Pakistanis.

Finally, U.S. aid should help turn Pakistan into a transport route for goods and energy between India and Europe via Afghanistan and Central Asia. This is also essential if Afghanistan itself is to be stabilized and its population given any real alternative to the heroin trade. This U.S. strategy should form part of a genuine and direct offer to Iran to help integrate that country into the regional and world economies in return for Iran's abandonment of its nuclear weapons program. Encouraging Iranian and Central Asian links to India via Pakistan should be made central to U.S. attempts to diminish Indian-Pakistani tension and to give both countries a critical stake in regional peace and stability. Helping Pakistan will also require U.S. responsiveness to Pakistani needs in the area of exports to the United States and the wider regulation of international trade.

These may seem wholly quixotic recommendations, given the present state of the U.S. budget deficit and the attitudes of the U.S. Congress. However, even radical improvements in Pakistan's economy would require only a tiny fraction of the U.S. funds thrown away on the Iraq War. And, as Stephen Cohen has rightfully observed, "The Pakistani people must see tangible evidence that the government's tilt in favor of the United States brings significant benefits to all socio-economic strata. Most aid is invisible to the average Pakistani, who cares little about debt relief or balance of payments problems."

The United States was capable of such enlightened realist generosity during the Cold War, and from Western Europe to South Korea to Thailand this often had spectacularly successful results in helping countries to resist communism. And if the threat from Islamist extremism and terrorism to U.S. interests and U.S. lives is not comparable to that of communism, then what has all the fuss been about these past five years?

So far, U.S. assistance has been frankly inadequate. By the end of 2006, Pakistan will have received about $3.4 billion in U.S. aid since 9/11, which sounds like a lot but is, in fact, very small in comparison to Pakistan's needs and the size of its population, and given that almost half of this aid is not for economic development but is security related.

"Democracy", in the shapes that it has been adopted in Pakistan so far, is not going to transform the country. Equally, direct U.S. intervention is out of the question. The United States will always have to try to shape the policies of Pakistani governments from the outside. The United States, therefore, will always require a Pakistani government that is not only responsive to U.S. wishes, but also strong enough to carry them out. Pakistan is likely to remain a kind of client state of the American empire, sometimes cooperative and sometimes recalcitrant, and U.S. policy will have to be calibrated accordingly.

The good news is that this is a problem that has faced empires throughout their history, so that we have a huge corpus of experience to draw on; the bad news is that as the Damadola incident has once again reminded us, the management of client states with unruly populations is an extremely delicate and complex task, and one that has often broken down in disaster.

Two of the world's most spectacular modern instances of empires mismanaging this kind of relationship happened almost simultaneously in the late 1970s: the fall of the American-backed shah of Iran to one Islamist revolt, and the crumbling of the Soviet-backed communist regime in Afghanistan in the face of another. The disastrous consequences of the Iranian revolution haunt American strategy to this day. The Soviet military intervention to save its Afghan clients was even more catastrophic, contributing both to the fall of the Soviet Union itself and to the rise of Al-Qaeda. It is essential that we do everything possible to prevent the United States in Pakistan repeating similar mistakes on an even greater scale.

¹Valuable recent works on Pakistan include: Stephen Philip Cohen, The Idea of Pakistan (2004); Hassan Abbas, Pakistan's Drift into Extremism: Allah, the Army and America's War on Terror (2004); Hussain Haqqani, Pakistan: Between Mosque and Military (2005); Mary Anne Weaver, Pakistan: In the Shadow of Jihad and Afghanistan (2002); and Kathy Gannon, I Is for Infidel: From Holy War to Holy Terror; 18 Years inside Afghanistan (2005).

Anatol Lieven is a senior fellow at the New America Foundation. His next book, Through Humility to Victory: Ethical Realism and the Great Capitalist Peace, co-authored with John Hulsman, is forthcoming this fall. It is based on their essay "The Ethics of Realism", which appeared in The National Interest in the summer of 2005.

Essay Types: Essay