Too often, the Beltway conventional wisdom emerges without careful scrutiny, before the hard questions have been asked. We are convinced this is the case with the thesis that the United States must reform failed states, represented most recently in Brent Scowcroft and Sandy Berger's article, "In the Wake of War" (Fall 2005).

The conventional wisdom asserts that failed states pose a grave danger to the United States, and the threat has grown more acute in the post-9/11 environment. In Scowcroft and Berger's formulation, "action to stabilize and rebuild states emerging from conflict is not 'foreign policy as social work', a favorite quip of the 1990s. It is equally a national security priority."

But failed states should be a priority only to the extent that such states pose a threat. The lesson of Afghanistan--that failed states can represent threats--should not be extrapolated onto failed states writ large. Although there has been precious little attempt (none by Scowcroft and Berger) to define what "failed state" means, a few states regularly appear at the top of any list: Côte d'Ivoire, Liberia, Haiti, Sierra Leone--are these the states where Scowcroft and Berger propose we "get serious about nation building"? If not these states, which? And why? To advocate nation building without a better sense of the criteria governing when and where to intervene sounds to us like a recipe for squandering national power to no good end.

Further, the American record in nation building is abysmal, and our failed attempts to reform failed states have been extremely costly. America has poured well over $23 billion into its Balkan interventions since the early 1990s, but all that money and effort has not reconciled the many differences between the various religious and ethnic groups in that region. Was intervening in the Balkans a good use of American power and focus, or would those resources have been better trained on the emerging threat posed by Islamic terrorism? The legacy of nation-building failures need not be catalogued on these pages.

It could be argued that we have failed so often because we have lacked determination; as Francis Fukuyama, a passionate proponent of nation building, points out, nation building "has been most successful . . . [in places] where U.S. forces have remained for generations. We should not get involved to begin with if we are not willing to pay those high costs." We agree, but we would humbly submit that proposals to "remain for generations" in strategic backwaters like Haiti should be viewed with extreme wariness.

Even Stephen Krasner, the director of policy planning at the State Department and another fervent believer in nation building, pointed out flatly in 2003 that "[t]he simple fact is that we do not know how to do democracy-building." This from a man now in a top post in the State Department, the same institution which Scowcroft and Berger propose should take the lead on democracy building.

Scowcroft and Berger write that "[t]he military will always have the main responsibility for establishing and maintaining public order and security in a post-combat setting." Indeed. But they provide no estimates of what this may entail, or the strains it could place on the military, other than to propose that DOD develop "the right mix and number of troops to provide for sustained operations" and that DOD "apply existing and emerging technologies to support stabilization operations " (They concede that "[t]his means increasing the number of active duty personnel . . . ") This is as close as the authors come to explaining the costs of the new policy. Thankfully, the Defense Science Board's Summer Study pinned some numbers to the abstractions.

DSB pointed out that "stabilization of disordered societies, with ambitious goals involving lasting cultural change, may require 20 troops per 1000 indigenous people." It is hard to imagine the United States having lower goals than these in a failed state. Extrapolating these numbers onto any number of failed or failing countries yields startling figures. We would need 162,000 troops in Haiti. Zimbabwe? 255,000 troops. What about Iraq? 522,000 troops.

The 20 troops per 1000 indigenous figure has long existed in the literature on stabilization and reconstruction missions. We hope that it makes its way to the forefront of our national security discussions, because ultimately the debate over reforming failed states revolves around a question of costs and benefits. By inferring from Afghanistan, and applying the lesson that all failed states constitute similar threats, the proponents of nation building dramatically overstate the benefits that will derive from these missions; by skirting the crucial question of what military and financial assets will be required to support these missions, and by avoiding the prickly issue of our limited success at reforming failed states in the past, Berger and Scowcroft underrepresent the costs. The debate would be better served by a more thorough accounting.

Christopher Preble and Justin Logan
Cato Institute

Ted Galen Carpenter and Charles V. Pena open their article ("Rethinking Non-Proliferation", The National Interest, Summer 2005) by telling us that Iran and North Korea feel they have much to fear from the United States. Elaborating on both the general scope of American military operations in the post-Cold War world and the specific threats these two nations perceive from the United States, Carpenter and Pena write: " . . . it is hardly surprising that Pyongyang and Tehran concluded that they might be next on Washington's hit list unless they could effectively deter an attack." One might say then that Iran and North Korea possess, oh, "legitimate security reasons" for seeking nuclear arsenals.

Turn the page, however, and Carpenter and Pena quickly resort to familiar caricatures. States like Iran and North Korea are "aggressive", "unstable", "erratic", even "unsavory" . . . and hence must be denied the nuclear prize. Even if they perceive it as essential to deterring an American regime change. And even if we, despite the most powerful conventional military force in the world, insist that it is vital to our own security.

Why can't a single "unsavory" regime be permitted to possess a single nuclear warhead? Because, say Carpenter and Pena, it then might "blackmail its neighbors." Perhaps it means that such a state might try to coerce another state by threatening a nuclear first strike. ("Evacuate the West Bank and East Jerusalem by next Thursday or else.") But all eight nuclear weapon states already possess the capability to make such threats, which we might call "nuclear compellance." And yet, with the possible exception of a few actual battlefield situations (for example, the Chinese let U.S. Marines withdraw from Chosin Reservoir after President Truman publicly considered employing nuclear weapons in 1950, the Viet Minh let French forces escape from Dien Bien Phu after Secretary of State Dulles offered Paris three Mark-21 tactical nuclear weapons in 1954), it is difficult to identify any historical instances where any of them have actually done so. Why would the leaders of an "unsavory" regime do so when the final result for them would likely be both personal and national suicide? Are these leaders more prone to the ultimate irrational act than were Stalin or Khrushchev or Mao? We worried for many decades about many scenarios regarding their nuclear capabilities, but not very much about "blackmail."

Perhaps instead it means that such a state might use its nuclear capability to persuade another not to do something. ("Don't try to change our regime or else.") That's nuclear deterrence. And that's precisely what we have long done, and that Carpenter and Pena advocate we continue to do, with our own nuclear capabilities. Why is it "nuclear blackmail" in one case but legitimate geopolitical behavior in the other case? Carpenter and Pena do not say.

But they do say that "nuclear arsenals in the hands of stable, democratic, status quo powers are not an inherent threat to peace and stability." (Is it even necessary to point out that a state which recently launched a unilateral, preventative, illegal, regime-changing invasion in defiance of both the UN Security Council and almost universal world opinion might not merit the adjective "status quo"?) Washington therefore should be pleased to welcome these states into the oak-paneled drawing rooms of the ultra-exclusive nuclear club. And then Carpenter and Pena lay out their own vision of a desirable nuclear future, and take the nuclear double standard to its logical consequence. One dozen, two dozen, perhaps five dozen nuclear weapon states . . . as long as they are states we like, or states like us.

Perhaps we should distribute Carpenter and Pena's article in bazaars, boulangeries, and barracks around the world. Some states are rational, sober, and righteous . . . and hence can be trusted with nuclear arsenals. Others are far too volatile, radical, and unpredictable to be allowed the same . . . even when such weapons are their only possible deterrent against the great hyperpower of the post-Cold War world.

Who will decide? Who will render subjective, ad hoc, case-by-case verdicts on whether certain leaders or certain peoples can be trusted with nuclear weapons? Who will serve as prosecutor, judge, jury, and enforcer? Why detached, dispassionate, disinterested observers, of course, like Rumsfeld and Rice, or Cheney and Rove, or Bush and Blair. Those in whose hands, it turns out, the nuclear arsenals already reside.