This is a guest post by Philip Martin, PhD student in the Department of Political Science at the Massachusetts Institute of Technology.

In a recent op-ed in the New York Times, columnist David Brooks advises a U.S. approach to Iraq which uses military force to arm-twist Iraqi elites into forming an inclusive new government, since “if you get the political elites behaving decently, you can avoid the worst.” At Political Violence @ a Glance, Barbara Walter also argues in favor of a negotiated settlement based on power-sharing as the optimal solution to Iraq’s current political fragmentation, an outcome that will supposedly “become increasingly attractive to everyone as the costs and risks of war increase.”

It is true that if moderate elites had more power in Iraq this would reduce the intensity of the country’s domestic political violence; it is less clear, however, that another power-sharing coalition government brokered by foreign interveners is an effective means to this end. For the last decade or more, scholars and practitioners have advocated for inclusivity, integration and power-sharing as the principal solution to the problem of civil war termination, expecting that these arrangements can reassure combatant groups of their participation in the post-war distribution of power, and eventually establish a cooperative model of governance which builds trust and moderation. Yet empirical research on foreign-imposed regimes and the determinants of peace agreement success provides little optimism about the likely effectiveness of these institutional arrangements.

In a study of 156 peace agreements signed between 1989 and 2008, for example, I found that while agreements with more “layers” of power-sharing (i.e. political, military, territorial) were generally more durable, negotiated pacts at the elite level (i.e. executive coalitions) were the least stabilizing form of political power-sharing (as opposed to, say, establishing minority veto rights). Even when used in combination with other types of power-sharing, settlements which relied on executive-level power-sharing were more likely to collapse (by approximately 60%) within five years. Thus, while power-sharing institutions may contain benefits for stabilizing divided states after civil conflict, these findings and others suggest that this relationship is likely being driven by factors other than inclusive elite coalitions at the top.

One of the key problems with externally-created elite coalitions in deeply divided states is that they very often include hard-line elements and other spoiler parties who enter the government with no intentions of actually fulfilling the terms of the agreement. Instead, they join these “grand coalitions” in order to defend or project their own political interests, gambling that they will be able to outmaneuver opponents once international pressure for compromise recedes. While robust third-party enforcement can sometimes help to secure the post-conflict environment, external mediation to create new “all-inclusive” governments also creates artificial incentives that, in the long run, leave combatants with less durable settlements than what they might have achieved on their own.

The dilemma is even worse when political pacts are forged under limited time-horizons, either because of upcoming elections or the withdrawal of critical external support. Under these circumstances, there is limited space for the emergence of a common political project between factional leaders, since uncertainty about the future drives elites to prioritize short-term tactical advantages at the expense of longer term goals like reconstruction and democratization.  Iraq itself is a good example of this trend. Although U.S. interveners tried to fashion an inclusive power-sharing government which could offer guarantees of security and inclusion to all of Iraq’s main players, Prime Minister al-Maliki never shared that vision. Since the beginning of the U.S. drawdown in 2009, he has shown increased authoritarian tendencies, consolidating his political power by neutralizing Sunni and former Baathist rivals within the government and tightening his grip on state security forces. As one Sunni Iraqi in Karada put it: “Sunnis after 2003 are like fish inside a small pool with a shark. They can’t get out, and they can’t remain inside.”

Similarly dysfunctional power-sharing pacts can be found elsewhere in history. Rwanda’s pre-genocide power-sharing government between President Habyarimana and the Tutsi RPF; Afghanistan’s consociational deal between the Northern Alliance warlords and Pashtun exiles after the 2001 Bonn Agreement; or Côte d’Ivoire’s string of failed power-sharing governments between President Laurent Gbagbo and the northern rebel alliance Forces Nouvelles (FN). In each of these cases, factional leaders within the government agreed to a power-sharing formula in order to appease foreign mediators and rebuild their war-fighting capacity, not to reconcile grievances with their perceived enemies. Unlike more fundamental and hard-to-reverse commitments – like surrendering one’s military capability or devolving state power to regional sub-units – entering an elite coalition at the centre is a cheap promise, one that is relatively easy to abandon later.

That is why the U.S. should not place any great amount of faith in the ability of a new power-sharing pact in Baghdad to deliver a substantially different political outcome for Iraq than what is likely to occur anyway – a de facto partition of the country into self-administered and increasingly homogenous sectarian zones. Any new elite coalition at the centre that foreign interveners manage to browbeat into existence is likely to be subverted by its participants as a means of accruing resources and projecting power, not as a vehicle for reconciliation or creating a functional pluralist government at the centre.

For the foreseeable future, then, a highly decentralized state which grants each of Iraq’s ethno-sectarian groups nearly exclusive self-governance is probably the least-bad option available. External interveners, therefore, should resist pressure to “do something” to impose a new centralized regime from above, and simply wait for the self-fragmenting tendencies of groups like the Islamic State of Iraq and Syria (ISIS) to run their course.  As long as al-Maliki – or his successor – continues to believe that the U.S. will provide military and financial resources indefinitely to prop up Iraq’s central government, there will be no serious incentive to make the serious, irreversible power-sharing commitments that could form a more durable peace. The result of such a strategy of restraint may not be the creation of a functional central Iraqi state in the short term, but would be more consistent with the limited leverage that outside interveners actually have.