BOTH NEOCONSERVATIVES and liberals have overestimated the extent to which one nation, even a superpower with United Nations support, can re-engineer regimes. Neoconservatives believe forced democratization is possible; liberals believe in the transformative power of foreign aid, debt relief, trade concessions and support for reformers. The tragic reality is that both approaches to long-distance, large-scale social engineering have failed in most cases.

One cannot stop genocide without bringing home body bags (as the Dutch found in Srebrenica). Brutal international realities often require following a "second-worst" course to avoid the "first-worst" one-our choices are not always between the best and the second best. In short, ask not what international order you desire, but what international order you can achieve.

Dimitri Simes made the same point in the last issue of The National Interest:

The idea that foreign policy decisions require hard-headed analysis and difficult choices is apparently offensive to those who believe that international affairs is a morality play and the United States is a global hegemon entitled, and indeed obliged, to right what they see as global wrongs. But being realistic about American options and dilemmas is not an obstacle to morality. . . .

Principled realism has moral foundations all its own. It avoids squandering lives and scarce resources (including political will, not just economic assets and military might) in the pursuit of illusory goals, the delays that result from unrealistically ambitious goals (for instance, opposing a unilateral withdrawal by Israel from Gaza because a negotiated one is preferred) and promises that cannot be realized (such as turning Middle Eastern nations into "shining democracies" by "flipping Iraq")-all of which in turn avoid losing credibility abroad and at home.

THE DEMISE of democratization as a rationale for U.S. foreign policy is all too evident. However, one can only now discern which leitmotif is to replace it. Using the principle of the "primacy of life", I propose what might be termed a "security first" approach.

Security here refers to conditions-both domestic and international-under which most people, most of the time, are able to go about their lives, venture onto the street, work, study and participate in public life (politics included) without acute fear of being killed or injured. It does not mean a utopia in which no crime or violence exists, and indeed to pursue "full-fledged" security puts us on a slippery slope toward a police state. Hence, this approach does not favor curtailing basic freedoms for marginal security gains in London, Madrid or New York.

A security first approach is centered on the assumption, increasingly borne out by events not only in Iraq but also in places like Russia, that security is a prerequisite for successful democratization. Among many Americans, there is a tendency to view security as antithetical to individual and civil rights and to warn that in the quest for security a nation may end up a dictatorship. These are indeed valid concerns. However, one should not overlook the primacy of the right to life. To not be killed, maimed or tortured are basic human rights, enumerated in the UN Universal Declaration of Human Rights. Also, Life precedes Liberty (and the Pursuit of Happiness) in the American lineup of core values.

Much of ethics deals with ranking two goods rather than determining which is right versus which is wrong. Lives and legal-political rights are two such goods. One is naturally inclined to hold that the government can protect both, and there is truth there: The better life is protected, the stronger the support for non-security rights. But, when the two do conflict, which wins? In a study, "How Liberty is Lost", I found that the failure to provide for basic public needs, especially security-not the gradual erosion of legal-political rights in the name of security, as is so often assumed-is why democracies fall. The Weimar Republic is a case in point; another is Russia since 1990.

This indicates that basic security is more urgent than other rights in dealing with failing states, rogue nations and genocide. In addition, the chaotic and violent international system sometimes produces conditions even more brutal than the internal conditions of most states. Therefore, this dictum applies with special force to attempts to form a stable global order. So moral arguments and empirical evidence support the same proposition: In circumstances where political right and security come into conflict-a common condition-the right to basic security must prevail.

THE SECURITY first paradigm has real implications for foreign policy. Take the approach to rogue states. In 2003, the UN lifted economic sanctions on Libya when the country accepted responsibility for the Pan Am 103 bombing and agreed to pay $2.7 billion in damages. Three months later, Muammar Qaddafi announced the voluntary dismantling of his nation's nascent program to build nuclear weapons and other WMDs. Libya's centrifuges and mustard gas tanks were loaded on an American ship and removed.

But human rights groups-along with a number of neoconservative figures-have complained that little progress, if any, has been made in terms of democratization. Thus, many demanded the continuation of sanctions in place of "rewards" to Libya for what were, by any objective standard, major contributions to global security.

The response to Libya is of considerable importance, given that the international community is keen for other nations to follow the same course. The security first approach takes the position that when a country has met international security standards, the world community will cease efforts to bring down the government, give it full access to various international institutions, lift sanctions and encourage investment. This, of course, does not rule out democracy promotion through educational and cultural means, but it makes compliance with basic security standards the principal factor in determining unencumbered access to the international community.

The current approach-expecting rogue nations to change their regimes and failing to recognize accomplishments short of that-makes it unlikely a country will drop its WMD programs. Both Iran and North Korea are reported to have sought non-aggression treaties or security guarantees from the West as part of a deproliferation deal. There is no way to determine a priori whether the countries made these offers in good faith or merely to gain time to advance further their nuclear programs. However, one can hardly expect them to consider a deal that will put the governing elites in play.

A deal that eliminates external challenges, in exchange for deproliferation and ending support for terrorism, is less bitter than it seems. It would not entail giving up the West's liberal soul in exchange for security. Although governments like those in Iran and in North Korea are swimming upstream against the current of democratization-it will happen at its own pace. In many parts of the world, overt U.S. support for the forces of change in these nations undermines rather than helps them. Moreover, the United States, and its allies, can topple Saddam or the Taliban, but it cannot secure rights and democracy in countries with little preparation for such polities. Hence, there is little to be lost and much to be gained by providing security guarantees and other international rewards in exchange for vigorous and verifiable deproliferation and the end of harboring, financing and equipping terrorists.

This approach should also guide relations with Russia. Our first priority ought to be ensuring that loose nuclear materials and nuclear arms do not fall into the hands of terrorists. Advancing this goal is quite challenging given that Russia has some features of a failing state. The national government is unable to consistently implement its domestic policies. Corruption is widespread, and individual entrepreneurs, generals and crime lords wheel and deal in an atmosphere reminiscent of the Wild West. This is of great concern, as most of the fissile material from which terrorists could make nuclear weapons is in Russia. Particularly worrisome are accounts, such as one issued in a 2005 NATO parliamentary report, of theft, or loss of highly enriched uranium and plutonium. There have also been reports that Russia could not account for scores of "suitcase-sized" nuclear bombs with yields up to one kiloton.

Nevertheless, in its dealings with Russia, the United States has put much more emphasis on democratization and human rights than on nuclear materials and arms. Indeed, the Nunn-Lugar program, which addresses nuclear security, has not had a budget increase since September 11. Putin may agree to much tighter controls on Russia's nuclear materials and arms if the proper incentives are offered; such controls add to his own security and have only small political costs. In contrast, he is unlikely to reverse the numerous undemocratic measures he has introduced. The reason is elementary: Such moves might well endanger his own power and regime.

Proponents of a tougher line toward Russia argue that the United States can increase pressure on Moscow without compromising U.S. national security. The fact is that even superpowers have limited leverage, and hence, the use of this leverage must be subject to prioritization. Here we see another major example of how wanton pursuit of democratization undermines security: ours, theirs and that of the world.

SETTING PRIORITIES, however, does seem quite difficult for the U.S. government and, more importantly, for the pundits who argue against having to choose. A security first paradigm acknowledges that, particularly in states newly liberated from tyrannical rule, it is often necessary to work with the established tribal, ethnic or religious groups-even if these groups rule in an illiberal manner-as long as the new leadership maintains basic security and does not preclude long-term, gradual democratization.

A brief review of post-invasion Afghanistan and Iraq highlights the stark differences between a democratization agenda and the security first approach. For advocates of democratization, U.S. efforts in Iraq and Afghanistan are on course. In both countries, U.S. military offensives toppled oppressive, authoritarian governments. In Afghanistan, an assembly composed of prominent Afghanis selected one of their own, Hamid Karzai, as president, a selection later validated by elections. His election, as well as the 2005 parliamentary elections, was verified by the UN as equitable and free of major fraud. As many as three-quarters of Afghanistan's ten million eligible voters turned out in 2004, despite threats of violence. A new constitution includes support for human rights, with a mandate that at least two women from each province serve in Afghanistan's lower parliamentary house.

In Iraq, the American head of the Coalition Provisional Authority was quickly replaced by an Iraqi Interim Prime Minister, Iyad Allawi, in 2003, who in turn was replaced by an elected prime minister. A constitution won popular support, and then the people elected a parliament. As in Afghanistan, the threat of insurgent disruption did little to deter voters. Even many disaffected Sunnis visited the polls.

Viewed through the lens of a security first approach, however, the postwar developments in both nations read rather differently. By mid-2006, the Iraqi insurgency continued in full force and was steadily on the rise in Afghanistan. Sectarian conflicts in Iraq and strife among provincial warlords in Afghanistan were also on the rise. Large segments of both populations were increasingly opposed to the U.S. presence and alienated from the democratic institutions, in large part due to the persistent failure to provide security. In Afghanistan, President Karzai barely dared leave Kabul, and only traveled with heavily-armed American bodyguards within Kabul itself. Rights guaranteed in the constitution mean little in the countryside. Sharia law is increasingly enforced in Iraq and never ceased to be the norm in Afghanistan. In short, the democratization scenario did not pan out.

Under a security first scenario, occupying forces might have to make the second-worst decision to leave many elements of the old regime in place, and then slowly work to convert them, while allowing considerable time for new forces to grow. This is, of course, the "hidden lesson" from the U.S. and Allied experience in postwar Nazi Germany. In Afghanistan this would have meant working with the warlords and the heads of the main ethnic groups, and allowing them to have extensive local leeway as long as they maintained security and did not engage in war with one another. In Iraq, this would have meant giving much greater license on local issues to Kurdish and Shi‘a leaders, and leaving some of the Ba‘ath Party members in power in the Sunni areas. In mixed regions, especially Baghdad, neighborhoods might have been segregated as they were in Belfast, Beirut and Jerusalem, until a more effective national government could take root. This would entail leaving local militias intact-incorporating them as far as possible into official structures-and limiting U.S. intervention to preventing clashes among them and impositions of religious or ideological codes by violent, extralegal means. The leaders of Basra, for instance, would be free to ban alcohol and fine those who do not obey, and even enact dress codes for women, but not to bomb liquor stores or assassinate Sunnis.

Deferring to ethnic groups, upon which the security first approach builds, is supported by the trend of ethnic groups gaining autonomy vis-à-vis national governments. In contrast, forced nation building, in the name of democratization, bucks recent history and people's preferences.

Nowhere is this more apparent than in the difficulties Americans have faced in coping with the role of religion in public life, especially in the Muslim world. The West tends to associate progress with secularization, but confronting religious authorities, beliefs and allegiances head-on invites failure. A security first agenda is less concerned with changing long-held norms and focuses attention on ensuring that religious expression is non-violent.

The fact is that most believers of all religions, including Islam, are moderate and non-violent, though often illiberal. In Iraq, the Grand Ayatollah Ali al-Sistani, by far the most influential Shi‘a religious authority, has repeatedly condemned assassinations, kidnappings and other forms of terror, supported free elections and the rule of law-but has not endorsed women's full equality with men or an unrestricted marketplace of ideas. Call him, and others like him, illiberal moderates. Over time, different religious and cultural communities might embrace a more liberal agenda. But they do not at present, and that should not disqualify them as allies in working to promote and extend security.

Finally, a security first approach also provides guidance on armed humanitarian intervention. The primacy of life means that the international community must intervene wherever populations suffer from genocide, defined in 1948 by the UN as widespread violence "committed with intent to destroy, in whole or in part, a national, ethnical, racial or religious group." However, armed interventions are not justified to protect a democratic regime from internal challenges of the kind faced by Haiti in 1994, when President Clinton sent 20,000 troops to reinstate Jean-Bertrand Aristide. Such challenges are common and even superpowers cannot realistically deal with all or even most of them-making decisions about when and when not to intervene inconsistent and hard to justify. One of the tragic features of international reality is that these matters must often be left to the people involved; painful domestic struggles are the growing pains of incipient democracies.

Moreover, under the primacy of life principle, any intervention must limit itself to security purposes. Troops should not be used for nation-building or "reconstruction." It is an odd notion that when one nation intervenes in another to stop a genocide, the rescuing nation "owes" political and economic aid. It is like arguing that a lifeguard who saves a drowning man owes him free swimming lessons. Indeed, the entanglements that followed in Kosovo, Bosnia and scores of other countries contribute to many nations' reluctance to stop genocides in places like Rwanda and Sudan. A primacy of life foreign policy avoids this pitfall.

However, the security first approach in no way precludes democratization by non-lethal means, such as sharing information, education, training, public diplomacy and financial resources. The United States already has a huge array of such public and private programs. These programs are difficult to evaluate, but in combination they may gradually improve the political cultures of the beneficiary nations.

A SECURITY first agenda is always difficult to accept if one still hopes that, with only a little extra effort (a few extra thousand troops here, tighter sanctions there, a tiny bit of aid more), the United States can fundamentally reshape states around the world. But as events over the last two years have dramatically shown, even superpowers are limited in their ability to change international realities. Getting security right is itself a very tall order. At the same time, focusing first on security-building does not mean ignoring democratization by peaceful means, as long as one realizes that democracy cannot be rushed and must be largely homegrown. This is the kind of realism that is faithful to our core values without risking our safety.

Amitai Etzioni is a professor of sociology and international relations at The George Washington University and author of Security First: For a Muscular, Moral Foreign Policy(Yale: Spring 2007).