It's pretty obvious why the United States has cared about failed states in recent years. A territory lacking a monopoly of legitimate coercion is a vacuum that attracts, in this current age of jihadist and sectarian violence, a lot of bad actors who can hurt the interests of distant countries as they did on September 11. Neither Boko Haram nor the Islamic State would be terribly impressive political movements if they had to deal with a well-organized state; we take them seriously only because they have flourished in the weak state environments of West Africa and Syria-Iraq.
I believe that the United States has overestimated the overall dangers represented by terrorism coming from such sources. Much of the extreme concern immediately after 9/11 was driven by the fear that jihadists were going to get access to weapons of mass destruction; while this remains possible, the real consequences are much smaller attacks like the Charlie Hebdo operation, which fall well below the level of existential threats.
The United States also has a humanitarian interest in the welfare and safety of innocent people living in these states. How powerful an interest this is is questionable, however, when the costs of action are high. Nearly as many people have died in the Democratic Republic of Congo over the last decade as in the Holocaust, and yet hardly anyone in the United States has suggested it is worth intervening militarily in this conflict.
I don't think that it is important to have a precise definition of a failed state, or to delineate a boundary between failed and weak states. Weak states that can't control security within their own borders, like Nigeria or Pakistan, can potentially shelter dangerous groups just as easily as failed ones like Somalia or the DRC. We're really not interested in failed states as a category, but in the political and security externalities that flow from conflict in them. It is understandable why U.S. foreign policy since September 11 has focused on state-building in Afghanistan, Iraq, and other areas; even if this is not the highest priority in US foreign policy, it is still reasonable that America should want to create order in potentially dangerous places. The real question is whether it can.
The historical record of more developed powers transplanting institutions to less developed ones is mixed. The most successful transplants were settler colonies in lightly-inhabited territories like North America, Australia, New Zealand, South Africa, Patagonia, and the like, where the native inhabitants could be physically killed or driven into reservations. Institutions were also transplanted into more densely populated areas like British India, Hong Kong, and Singapore. Success in the former case involved staying a very long time (a couple of centuries), and in the latter two from having strong military and commercial motives for investing in the colony and being able to shape it, as it were, from scratch.
In the modern world, however, norms prevent intervening countries from killing off native populations, and democratic electorates have little tolerance for high costs, casualties, and extended periods of occupation. In Nicaragua in the 1930s, Vietnam, El Salvador, and Iraq, American patience began to run out approximately five years after the initial intervention. The United States lasted a little longer in Afghanistan, but even there we decided to get out less than a decade after we entered. So the question is, what policies can be implemented within these time and resource constraints that will help build strong state institutions?
We have found that it is much more difficult and time-consuming to build modern state institutions than it is to construct democratic ones; indeed, of the state, law, and democracy triad, the latter is by far the easiest for foreigners to put in place. While we oversaw democratic elections (or "election-like events") in Iraq and Afghanistan, we completely failed to create state institutions that were impersonal, effective, legitimate, and able to penetrate the underlying society. High levels of corruption have been characteristic of both places, and have undermined our efforts to leave behind stable polities.
Below long-term settlement and occupation, there are three realistic levels of engagement that can be usefully identified as policy options: All In, Indirect Rule, and Containment.
In a series of Rand studies, James Dobbins and his colleagues have argued that there is a linear relationship between the level of resources thrown at a nation-building effort and its success. A lot of this is drawn from NATO’s experience in the Balkans, where heavy military and political engagement, plus a peacekeeping operation that is still in place 20 years after the end of the conflict, have succeeded in producing stability. American counterinsurgency doctrine as developed during the Afghan and Iraqi wars presupposes a similarly high level of resource commitment over an extended period of time.
There are two major problems with this approach. The first is that Western taxpayers have shown themselves unwilling to fund such operations at the level and duration that is required to make them work, Bosnia notwithstanding. The surge in Afghanistan showed that the US could stabilize the regime in Kabul and even push back the Taliban for as long as it wanted; but neither Congress nor the American public was interested in maintaining that level of commitment indefinitely.
The second problem is what Jim Fearon has identified as the free rider problem. No American commander, given adequate resources, is going to prefer to rely on indigenous forces to do the brunt of the fighting when the battle can be undertaken by U.S. forces. But this tendency to want to take the lead in both fighting and development relieves the local political actors of the need to get their own act together. War incentivizes state-building; but the U.S. doesn’t want to risk losing and thus undercuts this incentive.
Indirect rule was a doctrine first articulated by Lord Frederick Lugard when he was the British governor of northern Nigeria. Lugard argued that Britain had neither the manpower nor resources to rule its new colonial possession directly, as it ruled Hong Kong and Singapore; moreover, he argued that the indigenous inhabitants were better off living under their own customs and laws rather than being ruled by foreigners. This led the British authorities to engage in a prolonged search for “native law and custom,” which among other things gave a tremendous boost to the new discipline of cultural anthropology.
A contemporary version of indirect rule applied to places like Afghanistan and Iraq would have involved recruiting local elites to form governments that made no pretense of being either modern or democratic, but which would have achieved higher levels of buy-in. In Afghanistan, this might have led to the empowering of various unelected warlords to keep the peace; in Iraq, the U.S. could have put a general from the old Iraqi army in charge once Saddam Hussein was out of the way. The US would have maintained an indirect military role advising and training indigenous military forces, but would not have done the bulk of the fighting. The chief objective here would be stability and security rather than justice.
Arguably, the U.S. practiced a version of indirect rule in El Salvador in the 1980s. Congressional Democrats limited the number of American advisors in the country to less than 60, and forbade them to play a direct combat role. As a result, the Salvadorian National Guard was forced to bear the brunt of the battle with the FMLN, which it ultimately won. Indirect rule can thus encourage state-building by limiting free-riding on the part of the local client.
The limitations of indirect rule are evident from the British colonial experience. Putting in power a coalition of local actors requires a huge amount of local knowledge. The British colonial authorities invested a lot in their search for native law and custom, but even so got it wrong in many places. Their failure to understand the true nature of local politics and culture left a problematic legacy after decolonization. Some observers have charged that the Big Man of postcolonial Africa and the phenomenon of ethnicity itself were colonial creations of British administrators seeking to make their territories more tractable.
The most minimal level of engagement (short of completely ignoring conflicts in failed states) is containment. The U.S. would seek to prevent the outflow of terrorists and act as an offshore balancer, preventing the emergence of any single large, powerful actor. It would rely primarily on the desire for survival on the part of regional players to incentivize the creation of more durable institutions. It would also as a matter of principle not develop permanent friends or enemies, but would fluidly support groups that served its security interests. Any application of force would involve primarily airpower, with perhaps some use of military assistance and special forces.
The problem with containment is, first, that it may not be sufficient to solve the underlying problem of institutional failure. The U.S. in some sense “contained” the chaos in Somalia, but more than 20 years after the fall of Siad Barre the country has still not been stabilized. Second, a policy of containment or offshore balancing requires a high degree of cynicism in dealing with regional actors. Americans like to divide the world up into clear-cut and permanent friends and enemies; offshore balancing, while practiced over the centuries by the British, is deeply un-American.