MOST DISCUSSIONS about newly liberated states, such as Iraq, Kosovo or Afghanistan, start with what outsiders consider a preferred end product-a multi-ethnic, united, democratic, rights-respecting nation-state and one that is receptive to U.S. security interests. The question is then asked: How can the United States and its allies bring about these desiderata in these countries? Often, despite considerable human and economic costs caused by such overly ambitious designs, foreign powers continue to persist in their pursuit of utopian goals.
Discourse on Iraq in particular tends to vastly overestimate what foreign powers can accomplish, even if there was better planning, more boots on the ground and so on. Furthermore, such discussions too often assume that the players involved are unitary, sovereign nations, when in reality they are increasingly non-state actors. The real power players are various ethnic militias, tribal organizations and religious sects-in short, communities.
Recent developments in Iraq (as well as those in Afghanistan) suggest that one must initially work with forces loyal to local ethnic and confessional communities rather than the nation-state. Indeed, it is useful to remember, as Peter Galbraith notes, that:
Iraq has never been a voluntary union of its peoples. Winston Churchill, as Britain's colonial secretary, created Iraq from the wreckage of the Ottoman Empire in 1921. . . .Churchill later described Iraq's forced unity as one of his biggest mistakes.
But because there is a compelling interest in maintaining the frame of an Iraqi state (not the least of which is to ensure the territorial integrity of all the neighboring states of the region), the disappearance outright of an entity called "Iraq" would not allow us to recover from the initial mistake of putting Iraq together in the first place.
We need an approach based in reality-one that tailors the institutions of the state to fit the sociological reality on the ground, rather than trying to force that reality into an imported, precut outfit. Unfortunately, a major example of this highly unrealistic approach is the continuous attempt to convert the Iraqi militias (or Afghanistan's warlord armies) into unified national forces. This is highlighted by the folly of deliberately positioning Iraqi Shi‘a militias as security units in Sunni areas and vice versa-ostensibly to build up their national identification and loyalty. The fact of the matter, though, is that the first and foremost loyalty of most Iraqis is to their ethno-religious community, not to their state.
A community-based security approach sets the conditions under which most of the work for formatting a new Iraq can be carried out by the major players in that nation-the three main ethno-religious communities. This idea was championed as early as 2003 by Leslie Gelb and recently by Senator Joseph Biden (D-DE). It would reduce the incentive for sectarian armed conflicts, provide a setting for political give-and-take and greatly curtail the need for foreign troops. This approach seeks a middle way, avoiding the partition of Iraq, creating three states and instead allowing Iraqis to form several semi-autonomous regional entities, while maintaining some powers in the hands of the central government.
AS PART of letting Iraqis develop a state that suits their societal structure and commands their loyalty, Iraqis should be allowed a high degree of latitude in governing themselves in their respective provinces. Moreover, this could easily be done within the framework of the existing Iraqi constitution.1
But having outsiders work out these details and the United States and its allies impose them is not a solution. The specific features of a loose federation in Iraq can be developed best by regional representatives-perhaps drawn from each of Iraq's provinces. (Most likely, several provinces would combine to form regional governments united by ethnicity or confessional links.) And while a new Iraqi Federation might follow the U.S. federal model for some matters, it would also quite likely have a number of features far beyond what even the most vocal proponents of "states' rights" find attractive in the American context. For instance, the bulk of tax revenues would probably be collected by local and provincial entities rather than the national government, and the bulk of Iraqi security affairs would be handled by territorial defense units drawn from the local population rather than by national forces.
There might also be stronger limits on freedom of movement for Iraqis. One idea gaining currency, especially given the current security situation in Iraq, builds on existing checkpoint systems and puts American and allied forces in charge of patrolling the borders between provinces. Thus, a person who lives in one region and seeks entrance to another territory could be subject to a higher level of scrutiny (or even denied entrance), at least in the near future.
Inter-regional matters, from the sharing of oil revenues to limited population exchanges, would be worked out through negotiation among the representatives of the various communities. However, intra-regional matters, from licenses for barbershops to the legality of liquor stores, would be worked out within each province or group of provinces and would not be subject to an overarching national policy.
Granted, a loose federal system where a great deal of power has been devolved from the central government is far from an ideal institutional format; moving in this direction entails some risks. However, given current conditions, it is the best option-and possibly the only viable one. Most importantly, it would shift the United States's role in Iraq to securing regional borders rather than maintaining security throughout each region.
The benefits to be reaped from this approach are made evident by the current security situation in Kurdistan. In a February 2007 interview, Major General Benjamin Mixon, the commanding officer of U.S. forces in northern Iraq and Kurdistan, told 60 Minutes that of 20,000 troops under his command, a mere sixty or seventy are stationed in Kurdistan. Because Kurdish areas are patrolled by Kurdish troops, "there's no need" for an American presence there.
This kind of self-governance can be achieved not merely in the relatively homogenous ethnic regions such as the Kurdish Dahuk or the Shi‘a Maysan provinces, but even in mixed cities, in which neighborhoods-such as the overwhelmingly Shi‘a Sadr City or the Sunni Adhamiya neighborhood in Baghdad-are relatively homogenous and becoming increasingly so. As much as 70 percent of Baghdad is already ethnically segregated.
Much of the limited recent progress, often attributed to the surge, is actually due to the fact that the United States decided to work with local, communal forces rather than against them. This is most visible in Anbar and Amiriya, where the United States has joined with the Sunnis against Al-Qaeda. It is also the reason some parts of Baghdad have quieted down-largely because the local Shi‘a militias have taken control.
Such a separation of the vying parties into their respective corners, which experience shows greatly reduces violent conflict among them, will not prevent them from working out nationwide differences. Distribution of oil revenues and the like could be negotiated by the regional representatives in the same way that other federations hammer out such policies. There is no evidence that devolution would make such negotiations more difficult than they have been recently; on the contrary, if devolution leads to a decrease in hostilities among the groups involved, it will have the opposite effect on the national stage, making a settlement of differences more attainable.
And while we cannot predict with certainty that a loose federal state will reduce violence, there is some evidence to suggest that these measures could help minimize sectarian conflicts (as each ethno-religious community would control its own area) and curtail terrorism (because the presence and visibility of foreign troops would be much reduced). We have seen that when local forces are composed of soldiers with the same ethnic or religious background as the citizens of the areas they control, they have been able to bring considerable peace to their regions, with the exception of Basra, and even there only recently. This is not true merely for the Kurds in the north and the Shi‘a militias in parts of the south, but even in Baghdad. For instance, Sadr City was for a time an area of relative peace and prosperity-because the area was patrolled by a Shi‘a militia, the Mahdi Army. When, early in 2007, under pressure from the United States and its allies, the Mahdi Army withdrew in order to show Sunnis that Shi‘a areas are also under control of the "national" army and police, violence increased.
On the other hand, when attempts have been made, so far on a very limited scale, to work with communal forces rather than try to nationalize or defeat them, a measure of success is reported. For example, when American forces in the Babil province began to allow local forces to patrol their own community, they saw attacks on American forces halve and violence within the province decline, to give one more example.
WHEN THE proposals for community-based security in Iraq were unveiled for discussion and commentary at a meeting on Capitol Hill this past summer, a number of objections were raised. One of the most common of these is that community-based security will lead to bloodshed. Yet violence is already widespread, precisely because militias of one group are roaming the territories of others. Separating the fighting parties is likely to reduce-not increase-the carnage. The wider security situation will improve when one tribe ceases to impose on all the others the same "national" policies.Essay Types: Essay