How, then, can the ongoing civil war be terminated without ruinous escalation?

There are options. James Dobbins of rand has proposed a regional diplomatic campaign to induce Iraq's neighbors to use their influence with their Iraqi clients to compel compromise on a power-sharing deal. Given the Sunnis' dependence on outside backers for money and supplies, and the growing Shi‘a links with Iran, an agreement by neighboring states to sever this support unless their clients compromise could have real traction. Of course, this means offering neighbors such as Iran and Syria inducements that would make this worth their while; inducements sufficient to do the job could be expensive indeed, in many ways. And even if Iran and Syria cooperate, someone would still have to police the deal. But regional diplomacy could at least provide some real bargaining leverage, which we lack today.

The United States could also begin to use its own military policy in Iraq as a tool for settlement, rather than merely as a quick ticket home for U.S. troops. This would require the United States to make our presence, and our assistance, conditional on the parties' bargaining behavior: Those who compromise must be rewarded with security guarantees, but those who refuse must be threatened with military sanction. Today, U.S. military policy is independent of Iraqis' bargaining behavior: It is disconnected from our diplomatic strategy. In an ongoing war, military power is a potentially powerful lever, yet today the United States has left military power off the table as a reward for cooperation or a punishment for obduracy. Of course, military force is a blunt instrument: to use it as a bargaining tool could strain U.S. diplomacy beyond its capacity. An American willingness to realign militarily could destroy all sides' confidence in U.S. guarantees if not handled deftly. And any American promise to remain in a potentially dangerous Iraq could well prove beyond the tolerance of American voters.

But unless some new source of leverage is found-and quickly-our chance for even an intermediate outcome in Iraq could evaporate. The best-case outcome at this point is for a negotiated ceasefire in which the Sunni insurgency-not just the elected Sunni political leadership in Baghdad, but the insurgents and their armed leadership in the field-agrees to peace in exchange for concessions that would surely have to include a broad-based amnesty and a role for former insurgents in the government security apparatus, among other requirements. Such a deal would require U.S. troops to oversee its terms. But it also requires at least some level of Iraqi willingness to set aside the desire for revenge in exchange for the hope of peaceful coexistence. With every passing day, however, this reservoir of tolerance is being drawn down as the sectarian body count rises. At some point, it will surely be exhausted and the chance for a negotiated settlement will be lost in an uncontrolled escalatory spiral of violence and retribution. This point of no return does not appear to have been reached yet. Although the death toll rises, it also periodically falls-Iraqis still appear to be able to draw back from the precipice and restrain their combatants. But this will not last forever. And it may not last very much longer.

This brings me back to metrics. The analysis above implies a very different set than those most common in today's debate. Rather than Iraqi battalions trained or hours of electricity in Baghdad, the real measures of success and failure in Iraq are threefold. First, how close are the parties to achieving a power sharing deal and associated ceasefire? Second, how willing is the American public to accept a sustained peacekeeping role sufficient to police any deal the parties may reach? And third, how rapidly is the sectarian death toll rising?

Iraq today is a race between progress toward a settlement and acceleration of inter-communal tensions fueled by sectarian killing. Success requires that a settlement precede the loss of tolerance; defeat will occur if killing outpaces compromise. And to obtain the former rather than the latter will almost certainly require that Americans be willing to accept a long-term role in policing any ceasefire.

For now, the trends in these metrics are not promising: Compromise has been slow and grudging; while the death toll occasionally falls, the overall trend is sharply upward; and Americans are displaying diminishing tolerance for the U.S. troop presence in Iraq. Time is thus not on our side. Current U.S. policy is not yielding an aggressive pace of communal compromise in Baghdad; we risk letting the war slip out of control if we cannot find a means of accelerating the deal-making, and soon. And the longer the fighting goes on and the more Americans die without intercommunal accommodation or a ceasefire, the slimmer the political prospects for a significant long-term American troop presence. If a truce comes soon, trends in U.S. support for Iraqi deployments might reverse; if not, they surely will not. We still have a chance, but this window will not stay open forever. And this implies that we must aggressively seek out new forms of leverage to move this process along soon-before it is too late.

Stephen Biddle is senior fellow for defense policy at the Council on Foreign Relations.

Forget Failure, Let's Avoid Catastrophe

Peter Charles Choharis
If "success" in Iraq means that the war's benefits outweigh the sacrifice of the American and Iraqi people, then it is no longer possible even to conceive of success in Iraq, let alone achieve it.

More than 22,000 Americans have died or been wounded, and the financial cost has passed $300 billion. Iraqi civilians are being slaughtered by the thousands each month-often by sadistic death squads that torture their victims first-while thousands more are being driven from their homes. Billions of dollars remain unaccounted for, even as such basics as fuel, clean water and electricity remain in short supply. Regionally, American influence is at its nadir; while our ability to meet other global interests-including waging war against terrorism-is also at a low.

It wasn't supposed to be this way. Before the war started, the Bush Administration fostered inflated expectations about what victory would bring. America would create a stable and unified Iraq with a representative, constitutional government that respected ethnic, religious and women's rights, which could repel foreign terrorists as well as interference from neighboring countries, and which would enjoy a growing, independent, free-market economy.

Three and one-half years after the American invasion, Iraq is none of these. Despite elections, the government has little power to stop the sectarian violence, end the insurgency or protect its borders. Corruption is rampant, the economy depends on huge American subsidies and fundamental political questions remain unresolved.

Nonetheless, despite the Bush Administration's failure to create a peaceful and prosperous Iraq, we cannot simply walk away and thereby expose ourselves to new threats throughout the world. The challenge going forward for U.S. policy-makers, then, is to salvage Iraq in a way that will enable us to protect our regional and global security interests-even as political support for the war continues to decline and our military continues to suffer.

The question on the table should be: What can the United States hope to achieve within a time frame and at a cost that is acceptable to both Americans and Iraqis?

Chasing objective measures of improvement-like electricity production, hospital construction and fuel prices-will never lead to success, since these can be shattered by a single ied. In order to achieve lasting progress, Iraq's political process must be made to work. To prove to the Iraqi people that their elected leaders-not militias and death squads-hold the key to their future, the Iraqi Parliament must achieve five political milestones in the next 18 months (if not sooner) in order for the country to survive.

The first is federalism. The Iraqi people must decide whether to cast their fate with the central government or to look to more local authorities to protect them. The central government's inability to stop the incessant violence, to provide many basic services, to act independently of the American and other coalition forces and to rebuild the economy has created doubts about its abilities to govern. This doubt, coupled with a deep anxiety about the widespread violence, has created a crisis of confidence that has denied Iraqi Prime Minister Maliki's government a mandate to take necessary, tough measures.

The Iraqi constitution provides a peaceful way for Iraqis to decide where to place their faith and loyalty. Section Five of the constitution permits Iraqis to form "governates" and larger "regions." Governates are supposed to enjoy "broad administrative and financial authorities", while regions are even more autonomous and have "the right to exercise executive, legislative and judicial authority."

One thing Iraqis have proven themselves good at is going to the polls. Having to vote whether to join a semi-autonomous region or to stick with current political arrangements would resolve a lot of uncertainty-as well as provide a new outlet for vying for power and settling disputes.

The Iraqi Parliament is deeply divided over federalism. Sunni and secular parties, along with lawmakers aligned with the Shi‘a cleric Moqtada Sadr, oppose federalism, while the powerful Shi‘a Supreme Council for Islamic Revolution party backs federalism. Opponents fear that the Shi‘a will quickly form a nine-province region in the south-dooming the idea of national unity and depriving the rest of the country of southern oil revenues.