Promoting Democracy in Post-Conflict and Failed States:
Prepared for the National Policy Forum on Terrorism, Security, and America’s Purpose
September 6-7 Washington, DC
By Larry Diamond
As we move into the fourth decade since the great wave of global democratic expansion began in 1974, the task of promoting democracy faces a deepening set of challenges and contradictions. These revolve around two inter-related facts. First, as the number of democracies has increased—from about 40 in 1974 to around 120 today (slightly over 60 percent of all independent states)—the task of promoting democratic transitions and consolidation has become more difficult, because the countries with the economic, social, historical and geographic conditions most conducive to democracy have already installed (and in many cases, largely consolidated) democracy. Second, and related to this, many of the tough cases that remain are so not simply because they lack the classic facilitating conditions for democracy—more developed levels of per capita income, civil society, independent mass media, political parties, mass democratic attitudes and values, and so on—but because they lack as well the more basic conditions of a viable political order. Before a country can have a democratic state, it must first have a state—a set of political institutions that exercise authority over a territory, make and execute policies, extract and distribute revenue, produce public goods, and maintain order by wielding an effective monopoly over the means of violence. As Samuel Huntington observed in the opening sentence of his classic, Political Order in Changing Societies, “The most important political distinction among countries concerns not their form of government but their degree of government.” While this sentence (and really, the book itself) does not do justice to the importance of freedom and democracy for good governance, it does orient us to the fundamental importance of a coherent, capable state. It is an insight that has been coming back vigorously into the literatures on both democracy building and state building in recent years. The daunting reality of the contemporary world is that many of the remaining autocracies and semi-authoritarian states of the world lack this most basic foundation for building democracy. In these countries, the state has either collapsed in civil or international war, or it is weak, fragile, and at risk of collapse.
Weak and failed states pose distinctive problems for democracy promotion. In these states, the challenge is not only (or in some cases, even at all) to pressure authoritarian state leaders to surrender power but rather to figure out how to regenerate legitimate power in the first place. The imperative is not only to empower citizens and their independent organizations but to endow state institutions as well with resources, training, organization, and a sense of a common mission.
Within this broad context, there are three distinct types of cases. First are the post-conflict states that are emerging (or trying to emerge) from a period of external, or more commonly civil, war. Many of these countries have been in Africa—South Africa, Mozambique, Sierra Leone, Somalia. Some have been in Latin America (Nicaragua, El Salvador, indeed much of Central America), in Asia (e.g. Cambodia and one hopes now Sri Lanka), and in the Middle East (Lebanon, Algeria, and now Iraq). Second are the countries that are in the midst of civil war or ongoing violent conflict, where central state authority has largely collapsed, as in the Democratic Republic of the Congo. And third are the states that, while not yet gripped with large-scale internal violence, are at severe risk of it, because of weak or weakening state authority and capacity, high levels of crime and privatized violence, and increasing polarization of domestic politics (for example, Nigeria).
Each of these three types of cases requires specific kinds of strategies for democracy promotion. Obviously, the first imperative for states suffering civil war is to end it, and here international mediation, intervention, and peace implementation (as well as the more conventional forms of peace keeping) have a vital role to play. There is a large and distinct literature on this set of challenges, and I will only address it as it bears on the challenge of democracy promotion in these settings. In addition to all the usual types of efforts to build democratic civil societies, public values, political parties, and governmental institutions, states that are weak, feckless, and failing states require focused efforts to get at the sources of state failure, which frequently have to do with ethnic domination and injustice and endemic political corruption. Obviously, state institutions in this class of cases need to be strengthened in their skill and resource levels across the board, but this is generally not possible unless a new structure of incentives is institutionalized to foster commitment to the state and the country—the public interest—rather than to the advancement of individuals and their families, patronage networks, parties, and tribes. This is one of the most pervasive challenges of economic and political development assistance, and as I have also addressed this extensively elsewhere, I will only touch on it briefly as it bears on the class of post-conflict states, which will be the subject of the rest of this essay.
Democracy promotion in post-conflict states begins with the problem of order. By definition, there has been violent conflict. In some instances (such as El Salvador, Nicaragua, Rhodesia/Zimbabwe, South Africa, Cambodia, and Liberia), a peace agreement (often internationally mediated) may restore the authority of the state over its territory and implement peaceful means for sharing power or regulating the competition for power. One of the distinctive features of post-conflict state building in the past two decades has been the increasing reliance on formal democratic mechanisms, particularly elections, to determine who will rule after violent conflict. In other instances, either the pre-existing state has completely collapsed, so that there is no over-arching indigenous political authority left, or the authority of the state has shrunk back to only a portion of the territory over which it exercises international legal sovereignty. A vacuum of power is always filled, one way or another.
In the absence of an effective state, there are basically three possibilities. If there has been a civil war and a rebel force has ultimately triumphed, then the vacuum may be filled (gradually or even very rapidly) by the rebellious army and political movement as it establishes control over the state. However, this is highly unlikely to lead to democracy, as the triumph of violent insurgencies usually leads to the replacement of one form of autocracy with another (the American revolution being a striking exception). Second, there may simply be a patchwork of warlords and armies, with either no real central state (as in Somalia) or only a very weak one, as in Afghanistan. In this situation, the conflict does not really end, but may wax and wane in decentralized fashion, as in Afghanistan today. The third possibility is that an international actor or coalition of actors steps in to constitute temporary authority politically and militarily. This may be an individual country, a coalition, an individual country under the thin veneer of a coalition (essentially the case in Iraq with US administration after the fall of Saddam Hussein in 2003), or the United Nations acting through the formal architecture of a UN post-conflict mission, as in the UN Transitional Administration in East Timor (UNTAET) from 1999-2002.
Whatever the specific form of the post-conflict effort to build democracy, one thing must be stressed above all others: no order, no democracy. Democracy cannot be viable (and neither can it really be meaningful) in a context where violence or the threat of violence is pervasive and suffuses the political calculations and fears of groups and individuals. Thus, the promotion of democracy in post-conflict situations cannot succeed without the rebuilding of order in these contexts, and I therefore argue heavily in this paper that the tasks of democracy building and of peace implementation are inseparable. It is possible to implement peace without democracy, but it is not possible to build democracy without peace (and in fact, peace will be better and deeper with democracy). More generally, we can specify six distinct challenges of political reconstruction in a post-conflict setting:
1) Rebuilding the capacity of the shattered state, including its means of providing order and security (the army, police, and intelligence);
2) Controlling and demobilizing alternative sources of violence in the hands of non-state actors, such as religious and party militias, warlords, and other private armies;
3) Reducing the structural incentives to violence, through the design of political institutions that give a real stake in the system to each group that is willing to play by the rules of the democratic game;
4) Developing the political and social institutions of democracy, in the state and civil society;
5) Administering the post-conflict nation until…
6) The design and implementation of a plan for transition to a self-sustaining and democratic new political order.
These six tasks, which overlap in their temporal sequencing and often become highly compressed in time, encompass a number of contradictions, and it is in part the failure to acknowledge and somehow mitigate these tensions that accounts for the failure to build a sustainable democracy in these circumstances.
First is the tension between order and freedom. The post-conflict state needs an authoritative and capable public security establishment. But building up the police (and probably some kind of conventional armed forces) is in tension with the goal of empowering and privileging civilian political actors. The new state must have an internal monopoly on the means of violence, but this must be constructed carefully, with mechanisms and norms of civilian supremacy, so as not to create a new, anti-democratic military Frankenstein. It takes time to build the norms of deference to civilian control and respect for human rights and the rule of law, yet time is precisely what the reviving state does not have a great deal of. Moreover, the new security apparatus may face terrorists, warlords, and other violent spoilers whose brutal threats to the incipient new order can be easily seen to justify abridgements of due process and other restraints.
A second tension pits the imperatives of post-conflict democracy building against post-conflict administration and stabilization. The goal may be to establish democracy, but in a post-conflict setting it may be some time before free, fair, and meaningful elections can be organized. Thus, for some interim period, an unelected authority has to administer the country. Who? The best solution is a transitional government in which the former combatants share power by some agreed-upon formula until democratic elections can be organized (the South African model). However, it is difficult to broker such agreements in the midst of violent conflict or state collapse. A frequent model has been international intervention of both a military and political nature, with the international authority providing both a stabilization force to secure the country and a transitional authority to rule the country, or at least help referee the political situation, until a new constitution can be written elections can be held for a new permanent government. Herein emerges the dilemma. A nondemocratic (often in many respects quasi-colonial) power is asked to establish a democratic form of government.
The dilemma may be reduced when the international transitional authority “has been empowered primarily to hold an election and then withdraw” according to a defined and fairly imminent timetable (as with the eighteen-month UN Transitional Authority in Cambodia, UNTAC). It becomes more serious when the international authority is tasked with both administering the country and preparing it for sovereign democracy, with no specific end date, as was initially the case in Iraq and has been the case for quite some years now in Bosnia and Kosovo. The scale of operation and formal scope of authority also matter a great deal. In East Timor, the authority and resources of the UN gave it effective governing authority over the territory for more than two years, while the UN mission in Afghanistan operated with a much lighter footprint, involving only “a fraction of UNTAET’s staff and budget … in a country perhaps forty times the size and thirty times the population of East Timor.” The heavy footprint worked in East Timor, but those conditions (a situation of decolonization; support and acceptance from the local population; international consensus; and therefore broad domestic and international legitimacy) are likely to prove rather unique in the contemporary era. Gerald Knaus and Marcus Cox argue that the European Union’s mission in Bosnia and the UN Interim Administration Mission in Kosovo (UNMIK) have failed to build democracy in these two territories because they have ruled them as protectorates through a model of “authoritarian state-building.” While this has achieved some degree of stabilization, it has not cultivated the tools, incentives, and culture of democratic self-governance but instead has run roughshod over local resistance. By contrast, in dealing with candidate member states the EU has worked with local institutions, “giving them the capacity and the incentive to become active forces for development.”
Knaus and Cox offer a compelling critique of the Bosnia and Kosovo interventions. But these territories are certainly not Poland and Hungary, nor even Bulgaria and Romania. The problem is that the more a post-conflict situation is dominated by undemocratic leaders, parties, and movements, and by overriding ethnic or political divisions between them, the more a “light footprint” by the international community may leave only a light impact at best. A transitional administration must be strong enough to control, contain, and face down undemocratic elements, especially if they are armed and violent, and yet “light” enough to allow—and indeed cultivate—the emergence of local initiative and control, the development of democratic self-governance. This may not be an impossible combination, but in the worst post-conflict situations, it is a formidably difficult one.
Related to this is the third dilemma, involving time again. If the mission, or at least one important objective, of transitional administration is to promote democracy, then this requires the holding of free and fair elections. But if elections are to be truly free and fair (and democratically meaningful), there must be time to prepare them properly: time to construct electoral administration and disperse its offices and resources throughout the country; time to devise an electoral system that can provide the right kinds of incentives to restrain and transcend conflict; time to provide conditions of reasonable physical security for campaigns and voters; time to register and educate voters, organize election monitoring, train political parties and candidates, and enable them to build their organizations and mobilize support. Again, during this time, some non-elected authority has to rule. If that authority is international, the longer it rules, the more it risks a legitimacy crisis with the public it is trying to prepare for democracy, while falling into the model of “authoritarian state-building.” If that authority is domestic, a protracted period of interim rule may enable the unelected political forces to entrench themselves in power, generating a severely “unlevel” playing field for the elections when they do come.
Ill-timed and ill-prepared elections do not produce democracy, or even political stability, after conflict. Instead, they may only enhance the power of actors who mobilize coercion, fear, and prejudice, thereby reviving autocracy and even precipitating large-scale violent strife. In Angola in 1992, in Bosnia in 1996, and in Liberia in 1997, rushed elections set back the prospects for democracy and, in Angola and Liberia, paved the way for renewed civil war. There are compelling reasons, based on logic and recent historical experience, for deferring national elections until militias have been demobilized, new moderate parties trained and assisted, electoral infrastructure created, and democratic media and ideas generated.
In a context of shattered political order, truly free and fair elections take a long time to prepare, for they require not only a neutral and skillful administrative infrastructure but also an informed citizenry, organized parties, and a political climate largely free of coercion and violence. In theory, that would be a five- or even ten-year project in many post-conflict settings, especially where the state has collapsed or there is no previous history of democratic elections. It would have been better in the abstract for post-war Iraq if national elections could have been deferred for at least five to seven years. But the reality was that there was no way of constituting legitimate authority for very long in the interim, particularly with the country’s most important spiritual leader demanding elections for a national parliament as soon as possible. This means reaching for a difficult balance in timing (see below).
The fourth contradiction emerges out of two competing visions of post-conflict stabilization, one deeper, longer-term and more costly, the other easier to secure but far more vulnerable to failure. There is a temptation in a country that has been torn by war to reach for a false sense of peace because it is quicker and easier to obtain, to let different armed groups keep their arms and armies in exchange for thin promises of fealty to the new democratic order. Implementing a more thoroughgoing stabilization—in which alternative sources of violence outside the state are systematically demobilized—is time-consuming, financially expensive, and potentially costly in lives as well. Indeed, it is one of the ironies of a post-conflict situation that the new authority may have to wage new conflict in order to create the conditions for a more organic and sustainable peace. Such a genuine and democratic peace often requires a comprehensive “DDR” plan for the disarmament, demobilization, and reintegration (into society, and selectively into the new police and army) of various non-state armed forces. DDR plans require a lot of money and sufficient military power to forcibly disarm those groups that will not voluntarily sign up, and to ensure the compliance of those groups that have made commitments to demobilize and disarm. If stability in the transitional period is secured with international troops, they usually are not large enough in number and robust enough in their rules of engagement to take on this task.
Like other dimensions of post-conflict democracy promotion, there is no one standard model or formula for the control of violence. Social, cultural (and political) realities may require a concession that allows the citizens to keep small arms, but Joanna Spear argues that there is a common imperative to demobilize (disband) large-scale military or paramilitary formations outside the state, and this requires considerable political will and skill, knowledge of the specific environment, and often financial and military resources. Another irony is that if international military force must be mobilized to demobilize private militias and violent challengers to the new democratic order, these foreign troops may also become part of the problem, in that their presence can provoke resistance, particularly if they kill local combatants, and more especially if they become undisciplined and themselves violate individual rights and the laws of war. This has become a big part of the problem in Iraq, where American troops have been both a bulwark of security and a lightning rod for nationalist resistance and insecurity, but it has also been evident in some African cases as well.
To summarize, in a way: When we mention the term democracy promotion or democracy building, we tend to think of a fairly conventional set of tasks—helping to develop political parties, civil society organizations, representative and legal institutions, and so on. All of these are important. Indeed, all of the things that need to be done to promote and develop democracy in a historically authoritarian setting need to be done in a post-authoritarian, post-conflict setting. But post-conflict settings are distinctive in terms of the roles of violence, order, and stateness. If these challenges are not met, all the others will fail. This was a problem that the American occupation of Iraq never adequately grasped. Consequently, the ambitious conceptual plans for political, civic, and economic reconstruction could never really be implemented because of the widespread violence. To a lesser degree, Afghanistan faces the same problem today.
Some (Tentative) General Lessons and Guidelines
Post-conflict situations vary considerably in their dynamics and distributions of power, and in the hierarchy of challenges they face. Probably the single greatest lesson to be learned from previous efforts at stabilization and democratization of conflict-ridden states is that there is no one lesson or model. Just as generals always fight the last war, so do nation-builders always apply the model of the last post-conflict mission, or of some earlier historical model that may be quite limited in its fit. Thus did the late UN administrator, Sergio Vieira de Mello, bring to the new UN mission in East Timor the same basic model he had used in Kosovo—only to find that it did not fit. The American occupation of Iraq, under the sweeping authority of the Coalition Provisional Authority, seems to have been inspired to some extent by MacArthur’s postwar administration of Japan, despite the profound differences in the political, sociological, geopolitical, and historical conditions of those two occupations. Understanding the context is crucial.
Still, there are some lessons that from recent post-conflict democracy-building and stabilization experiences that appear to be generally relevant. I begin by formalizing the first imperative just mentioned.
1. Understand the local context in its historical, cultural, political, and sociological dimensions. While this is generally important for assisting democratic development in any context, it is especially vital in the wake of violent conflict or state failure, because state collapse generates conditions that are so unfavorable to the development of democracy, and that often require not just democratic assistance but a much more massive and wide-ranging set of international commitments (see below). Thus, in post-conflict settings, the scope of international intervention is likely to be far greater (if there is to be any chance for democratic success), and the margin for error is at the same time much less. Inadequate understanding of the local context—including such vital issues as political leaders and alliances, historical trends and grievances, religious, ethnic and sub-ethnic divisions, the sources of legal and illicit revenue, and the structure and loyalties of private militias—can be crippling. The problem is made worse by the fact that many of the countries whose states are emerging out of conflict or collapse are poorly understood in the West or by actors in the international community. They may be small and historically lacking in strategic importance, like Sierra Leone or Liberia, and so not the subject of much good intelligence or even academic insight. War, and before it a long period of brutal misrule or social disintegration, may have made the country difficult to study, or may have radically altered many of the structural parameters (political leadership, ethnic divisions and alliances) which had previously been understood by academic, diplomatic, and intelligence experts.
Thus, a serious effort to promote democracy in a post-conflict state must begin early on with a fairly intensive and comprehensive mobilization and integration of existing country knowledge from all sources, private and public, governmental and non-governmental, academic and operational. Next, any international mission should be advised on the ground, on an ongoing basis, by some number of leading experts on the country, not just political scientists but ideally historians and anthropologists as well (economists would also be helpful, but they tend not to focus on specific countries, and especially not on poor, marginal ones in danger of state collapse). These experts should be drawn from across the available resources in the international community, not just from one country. The failure of the United States to mobilize and more fully utilize expert knowledge of Iraq from the beginning was an important contributing factor to the bumbling, ineffectual character of its occupation.
2. Mobilize and commit adequate military and financial resources. This is probably the most difficult lesson to apply, because all resources are scarce, and it is very difficult to get the primary national and multilateral actors in the international community to commit the military force necessary to truly stabilize a country where the state, and with it civil order, has broken down over an extended period of time. For one thing, it is financially costly. For another, it is risky, in that countries contributing forces may suffer casualties, and their leaders may then pay a high political price. Finally, deployable military force is a far more finite resource than money alone. The recent Report of the UN Secretary-General’s High-level Panel on Threats, Challenges and Change noted that with 60,000 UN peacekeepers deployed in 16 missions around the world by the end of 2004, and more likely to be committed soon to other war-torn African states, the world is running out of available forces for peacekeeping and peace enforcement. (The panel did not mention the military engagement of the American-led coalition in Iraq, since it is not a UN mission, but even 150,000 troops have not been able to stabilize that country over more than two years, and it is clear that the United States, with the best military in the world, is already feeling the strain on its own sustainable military capacities). As the UN panel’s report makes clear, significantly more international peacekeeping/enforcement forces must be made available, along with “sufficient transport and logistic capabilities to move and supply those who are available…. The developed states have particular responsibilities here, and should do more to transform their existing force capacities into suitable contingents for peace operations.” Currently, the armed forces of many of these countries have outmoded, cold war-era structures, “with less than 10 percent of soldiers in uniform available for active deployment at any given time.”
Sometimes, it takes a lot of troops to create the enabling environment for democracy building and national reconstruction, because the situation confronting international actors is not truly one of “post-conflict.” War in its conventional form may have ended, but order has been shattered, violence continues, and armed groups stand ready to use violence and intimidation to enhance their political position or to undermine the implementation of any peace agreement that does not meet all of their key demands (the latter marking them as “spoilers.”) One of the two greatest obstacles to the democratization of post-war Iraq has been the lack of adequate force for stabilization of the country after the end of formal hostilities (with the collapse of Saddam Hussein’s Baathist regime). As I and others have argued repeatedly, this was not due to lack of advance warning of what could be expected during the “post-war” period, but rather to a stubborn, blinding refusal on the part of Secretary of Defense Donald Rumsfeld and other key Pentagon and Administration officials, including President Bush himself, to heed the experts on post-conflict stabilization and on Iraq (including the senior command of the United States Army, which sought an invasion and stabilization force of several hundred thousand troops, or at least twice as large as what was utilized).
The Army’s initial request for troops in Iraq was much more in line with the ratio of foreign troops to domestic population in the international interventions in Bosnia and Kosovo, which if replicated in Iraq would have meant an initial international force of 460,000 to 500,000 troops. Pentagon planners probably worried about the capacity of the United States to mobilize such a large force, and about the resulting casualties. But the RAND study, led by James Dobbins—who had served in the previous decade as U.S. special envoy for the post-conflict missions in Bosnia, Kosovo, Haiti, Somalia, and Afghanistan—concluded: “There appears to be an inverse correlation between the size of the stabilization force and the level of risk. The higher the proportion of stabilizing troops, the lower the number of casualties suffered and inflicted.”
Being able to mobilize adequate resources for post-conflict stabilization and (democratic) reconstruction requires three further imperatives. First, the actors who would intervene must assess the difficulty of the mission and the prospects for success. Stephen Stedman and his colleagues in the most systematic and comprehensive study to date of peace implementation efforts find three factors in particular that are “most commonly associated with a difficult environment.” These are first, the likelihood of spoilers, especially total spoilers; second, neighboring states hostile to the peace agreement (or the new, democratic post-conflict political order); and third, “spoils—valuable, easily tradable commodities.” The difficulty of peace implementation also increases with the number of warring parties, the absence of a (non-coerced) peace agreement before intervention, a collapsed state, the number of combatants in the conflict, and demands for secession.
Iraq in this regard was completely off the charts in terms of difficulty—virtually all of the eight unfavorable conditions specified by Downs and Stedman were strongly present. Analysts warned that a violent insurgency (consisting of dedicated spoilers) would emerge in the postwar situation, especially if there was a prolonged occupation. It was clear that neighboring states, particularly Syria and Iran, would be hostile to the construction of democracy and would try to sabotage it—as indeed they have tried. The insurgency has consisted of a number of different elements with different interests, and there are various other armed militias in the mix as well. Oil looms large, the state has collapsed, there is no peace agreement, there are a fair number of armed combatants (though not in terms of formal armies), and there is powerful Kurdish sentiment for secession (along with periodic threats from Kurdish elites).
The more difficult the (post)conflict situation, the more peace implementation requires for success an accurate estimate of the difficulties involved and the resources—and sacrifices—that will be required, as well as a powerful state willing to bear at least a considerable share of those risks and costs. Facing down spoilers and stabilizing a war-torn country where peace must really be imposed, not simply kept, requires the commitment of a major international or at least regional power, which views stabilization in its own vital strategic interests. This means not just an international organization (like the UN) or loose international coalition, but at least one powerful state. After surveying the bloodied landscape of peace implementation, Stedman found: “All too often in the 1990s international and regional organizations were sent to implement peace agreements in extremely challenging environments where no major state possessed a security interest. When implementers were challenged, the missions failed, usually with catastrophic consequences.”
The lessons learned from failed peace implementation efforts are sobering. If the international actors that are intervening to implement peace—and democracy can only be meaningful and viable in a context of peace—do not judge the difficulty wisely, and are not willing to commit adequate resources, they will likely fail. And since failure entails a tragic loss of lives and resources, and can discourage future interventions (even ones that are more likely to succeed), it is better not to intervene than to do so with a level of resources and commitment that makes failure quite likely. This generates a third lesson, about the international circumstances that will more likely call forth the necessary resources.
3. Establish international legitimacy and active support for the post-conflict intervention. Such an intervention will more likely raise and sustain the necessary resources and commitment if there is a shared sense of importance and commitment in the international community, ideally formalized by United Nations Security Council mandate. As the difficulty of the challenge rises, the need for a powerful state to take a vital interest in the mission increases, but that alone will probably not be enough to generate success unless there is significant international participation. This is not only because of the need to distribute and share the burdens, but also because of the imperatives of legitimacy internally in the country, which require that the intervention not be seen as (and truly, not be) the imperial action of another powerful state.
4. Generate legitimacy and trust within the post-conflict country. No international reconstruction effort can succeed without some degree of acceptance and cooperation—and eventually support and positive engagement—from the people of the failed state. If the local population has no trust in the initial international administration and its intentions, the intervention can become the target of popular wrath, and will then need to spend most of its military (and administrative) energies defending itself rather than rebuilding the country and its political and social order.
In the final page of an impressively wise and learned book on post-conflict state building, Simon Chesterman writes: “Modern trusteeships demand, above all, trust on the part of local actors. Earning and keeping that trust requires a level of understanding, sensitivity, and respect for local traditions and political aspirations that has often been lacking in international administration.” Unfortunately, the occupation of Iraq lacked these qualities, and the Iraqi people knew it.
Chesterman advises that when the United Nations and other international actors come “to exercise state-like functions, they must not lose sight of their limited mandate to hold that sovereign power in trust for the population that will ultimately claim it.” This requires a balancing of international trusteeship or imperial functions with a distinctly non-imperial attitude and a clear and early specification of an acceptable timetable for the restoration of full sovereignty. The humiliating features of an extended, all-out occupation should be avoided as much as possible. In fact, whenever possible, the better course will be to avoid international occupation altogether and organize a broad-based national conference to choose an interim government.
Some theorists and practitioners have been searching for a formula for international intervention to democratize failed states that stops short of full-scale imperial rule (whether by one nation or many). One possible approach is through some form of “shared sovereignty.” However, these formulas are viable precisely because they build not only on the de jure sovereignty of a state, but also on that state’s retention of de facto sovereignty over most conventional aspects of policy. Such formal abridgements of sovereignty are likely to be more palatable if they are negotiated with international institutions or multilateral actors rather than a single powerful state.
Shared sovereignty is for the longer run, when failed states have begun to revive. In the nearer term, only international military intervention in some form can fill the vacuum left behind when a state has collapsed and a country is in or at the edge of or just emerging from chaos and civil war. Force must be used, or at least effectively deployed and exhibited, to restore order. Military occupation does not legitimate itself, however, but needs to be paired with a clear indication, from the very beginning, “as to how a temporary military occupation is to begin the process of transferring political control to local hands.” Such a framework should limit the political occupation not only in time, but in scope as well, allowing for the occupier, or the peace implementation force, to be held accountable. Such mechanisms of accountability can “encourage the emergence of an indigenous human rights and rule of law culture as well as improve the day-to-day governance of the territory,” while also stemming the accumulation of local resentment and frustration. (Two huge mistakes of the American occupation of Iraq were establishing an indefinite occupation with no clear timetable initially for the return of sovereignty, and not allowing any means by which the occupying authorities could be questioned, scrutinized, and themselves held accountable).
5. Hold local elections first. International interventions that seek to construct democracy after conflict must balance the tension between domination for the sake of implanting democracy and withdrawal in the name of democracy: The two competing temptations are 1) to transform the country’s institutions and values through an extended and penetrating occupation (à la British colonial rule); and 2) to hold elections and get out as soon as possible. A key question is always how long international rule can be viable. In the case of Iraq, the answer—readily apparent from history and from the profound and widespread suspicion among Iraqis of U.S. motives—was “not long.” The failure to establish early on a date for national elections to choose a constitutional assembly became a major bone of contention between the U.S.-led occupation and the most revered religious and moral leader in Iraq, Ayatollah Sistani.
The pressure for rapid national elections might have been contained better if the United States had not constructed a full-blown occupation, but rather transferred power back to Iraqis quickly through a broad-based national conference with UN assistance, and if the international authorities in Iraq had allowed local elections to take place fairly soon. Even when the Coalition Provisional Authority (CPA) did organize at least indirect elections for provincial and local councils, it then undermined their authority by failing to give them meaningful resources and authority. This violated Chesterman’s general guideline that executive authority should be devolved to local actors as soon as practical, and that “once power is transferred to local hands, whether at the municipal or national level, local actors should be able to exercise that power meaningfully, constrained only by the rule of law.”
In general, there is a strong logic to holding local elections before national ones, and as soon as practicable. Dobbins and his RAND coauthors find that holding local elections first “provides an opportunity for new local leaders to emerge and gain experience and for political parties to build a support base.” That could well have happened in Iraq if local elections had been allowed to proceed during 2003, and if some meaningful scope of authority and resources had been devolved to the newly elected bodies. Then the United States would have faced a broader, more diverse, and more legitimate array of Iraqi interlocutors, and the elected local bodies could have provided one basis for selecting an interim government.
6. Promote knowledge of institutional choices for democracy, and of democratic principles and norms. Post-conflict societies are generally weak in knowledge of the institutional options for structuring democracy to manage ethnic and other group conflict, to protect individual and group rights, and to generate incentives for moderation in political behavior. Any effort to promote democracy in this setting has to involve the dissemination of this knowledge, with some detail and sophistication for political elites, but in simpler terms for the mass population as well. Assisting the formation and development of research institutes and NGOs that promote understanding of institutional designs to manage ethnic conflict and respectful debate on constitutional options for political reconstruction should be an early priority for democratic assistance. So should efforts by various types of NGOs and state institutions to educate the public about democratic norms, principles, and values. A mass civic education campaign must make people aware of their rights, train them in the arts of active citizenship, and lead them to hear, tolerate, and respect opposing views and interests.
7. Disperse economic reconstruction funds and democratic assistance as widely as possible. Both for the effectiveness and speed of economic revival, and for the building of local trust and acceptance, there is a compelling need to decentralize relief and reconstruction efforts, as well as democratic civic assistance. The more the international administration and private donors work with and through local partners, the more likely that economic reconstruction and democracy building efforts will be directed toward the most urgent needs, and the better the prospect for the accumulation of political trust and cooperation with the overall transition project. In Iraq there was a particularly compelling need for the creation of jobs, a need that could have been met more rapidly if the repair and reconstruction contracts had been channeled more extensively through a wide range of local Iraqi contractors, instead of through the big U.S. corporations. Decentralization and rapid dispersal requires dispersing some operational and spending authority to lower-level international officials who are resident in different cities and provinces. It may make it more difficult to obtain the kinds of receipts that permit auditing. It may be more conducive to corruption—though corruption was hardly contained in the centralized operation in Iraq. Hard choices must be made. But in the hard and urgent circumstances of a post-conflict situation, I think there is a case to be made for decentralized dispersal of small contracts and grants, along with early efforts, gathering as much information as possible, to evaluate performance.
8. Promote local participation, and proceed with humility and respect for the opinions of the people in whose interest the intervention is supposedly staged. There is, or certainly can be, a large dose of arrogance in any effort at international assistance, including democracy promotion. The danger of arrogance, or fatal conceit, grows with the weakness, poverty, and urgency of need of the recipient state—up to the point where that state has failed altogether and is more or less helpless. Such ambitious international intervention cannot succeed, and the institutions it establishes cannot be viable, unless there is some sense of participation and ownership on the part of the people in the state being reconstructed. This is why holding local elections as early as possible is so important. It is why it is so vital to engage local partners, as extensively as possible, in post-conflict relief and economic reconstruction. And it is why the process of constitution-making must be democratic and broadly participatory, not merely through the election of a constituent assembly or a constitutional referendum (or ideally, both), but through the involvement of the widest possible range of stakeholders in the substantive discussions and procedural planning, and through the organization of an extensive national dialogue on constitutional issues and principles. As Jamal Benomar observes, “Constitutions produced without transparency and adequate public participation will lack legitimacy.” And illegitimate constitutions augur poorly for future stability.
9. Institutionalize the capacity for effective intervention and democracy promotion in post-conflict settings. The UN Secretary-General’s High-level Panel has offered a number of promising and ambitious suggestions for enhancing and institutionalizing capacity in the UN and the international system. Prominent among these is the establishment of a Peacebuilding Commission (with a significant permanent support staff) to identify countries at risk of state failure and intervene early to prevent it; to assist the transition from conflict to peace (and one hopes to democracy, though the Panel does not mention this); “and in particular to marshal and sustain the efforts of the international community in post-conflict peacebuilding over whatever period may be necessary.” In addition, the Panel recommends that member states “strongly support” the efforts of the UN’s Department of Peacekeeping Operations to facilitate more rapid deployment by enhancing strategic stockpiles and standby arrangements for peacekeeping deployment, and it proposes to strengthen, and in essence shake up and professionalize, the UN Secretariat.
On the American side, I believe we need to strengthen our own efforts by creating a cabinet-level Department of International Development and Reconstruction, building on the core of USAID and incorporating as well the State Department’s Office of the Coordinator for Reconstruction and Stabilization. We need a permanent, institutionalized standing capacity to land on the ground quickly after conflict and (in concert with local actors and a wide range of other international donors) help the country reconstruct itself politically, economically, and socially. This requires specific training and skills, as well as lesson-learning and institutional memory, which are not well situated in the Departments of State or Defense. A separate cabinet department would enable us to have a standing administrative and reconstruction capacity, much readier to deploy, with a mix of experiences and language capabilities and close working relations with other government departments and agencies.
I sincerely believe that every country can become a democracy—eventually. No country is ruled out because of its preceding history, culture, or social structure. But not any country can become a democracy at any particular moment, and certainly not quickly. Failed states pose among the most difficult challenges for democratization. Sometimes democratization—assisted heavily from the outside—will be an indispensable means for the restoration of order, as in South Africa and Nicaragua. Sometimes, as in Cambodia, the international community will claim to be promoting democracy when it does not have the stomach or resources for the fight. Even then, the international intervention may leave behind fragments of hope for political pluralism, and at least a less thoroughgoing authoritarianism than what otherwise might have emerged. Authoritarian states do not become democracies just because they hold elections in which opposition parties compete and win some seats. But in granting some space for opposition and dissent, they are often preferable to the harder authoritarian alternative, and leave open the possibility of eventual democratization.
It would be better if, in countries like Cambodia, the international community would summon the resources and the will really to promote and insist upon democracy. But the hard truth is that we lack in the international community today the finances, the troops, the political will—and probably also the knowledge—to promote democracy successfully in the most forbidding cases. I fear Iraq will bear out this sad truth once again. In these circumstances, it is possible that we are better off having tried, even if half-heartedly, to build democracy, while winding up with a partially democratic system—a country at least struggling in the “gray zone”—than we would have been if we had just resigned ourselves to dictatorship from the start. But the people who suffer under new forms of oppression—however much they fall short of genocide, absolute dictatorship, or civil war—still wish for something better politically. At a minimum, we owe it to them to remain engaged, morally, rhetorically, diplomatically, and with concrete programs for democracy assistance, once the failed state has begun to take shape on less than democratic grounds.
It is not inevitable that we will fail to promote democracy, even in the hard cases. When we do, we should not regard that as the end of the story. If a new authoritarian state emerges, the struggle to promote democracy resumes, on more familiar if incremental terms. The tragedy, however, is that once a new authoritarian regime consolidates its grip, as Hun Sen and his Cambodian People’s Party have done in Cambodia, it may be very difficult to dislodge, particularly if it has authoritarian neighbors and defenders. For all their challenges and vulnerabilities, post-conflict countries do provide an arena of considerable fluidity for building a more democratic state and society.
 Samuel P. Huntington, Political Order in Changing Societies (New Haven: Yale University Press, 1968): 1.
 See for example Juan J. Linz and Alfred Stepan, Problems of Democratic Transition and Consolidation (Baltimore: Johns Hopkins University Press, 1996), and Francis Fukuyama, State Building: Governance and World Order in the 21st Century (Ithaca, NY: Cornell University Press, 2004).
 Perhaps the definitive work to date on this subject is Stephen John Stedman, Donald Rothchild, and Elizabeth M. Cousens, eds., Ending Civil Wars: The Implementation of Peace Agreements (Boulder, CO: Lynne Rienner, 2002).
 Foreign Aid in the National Interest: Promoting Freedom, Security, and Opportunity (Washington, DC, 2002): chapter 1 (available at www.usaid.gov/fani/); Larry Diamond, “Promoting Real Reform in Africa,” in E. Gyimah-Boadi, ed., Democratic Reform in Africa: The Quality of Progress (Boulder, CO: Lynne Rienner, 2004): 263-292; and Larry Diamond, “Building a System of Comprehensive Accountability to Control Corruption,” in and Adigun Agbaje, Larry Diamond, and Ebere Onwudiwe, Nigeria’s Struggle for Democracy and Good Governance: A Festschrift for Oyeleye Oyediran (Ibadan, Nigeria: University of Ibadan Press, 2004): 221-242. See also Larry Diamond, “Democracy, Development and Good Governance: The Inseparable Links,” First Annual Democracy and Governance Lecture, Center for Democratic Development, Accra, Ghana, March 1, 2005, at http://www.stanford.edu/~ldiamond.
 Simon Chesterman, You, the People: The United Nations, Transitional Administration, and State-Building (Oxford: Oxford University Press, 2004): 72-3.
 Chesterman, p. 97.
 Gerald Knaus and Marcus Cox, “Building Democracy after Conflict: The ‘Helsinki Moment’ in Southeastern Europe,” Journal of Democracy 16, no 1 (January 2005): 49.
 Stephen John Stedman, Donald Rothchild, and Elizabeth M. Cousens, eds., Ending Civil Wars: The Implementation of Peace Agreements (Boulder, Colo.: Lynne Rienner Publishers, 2002). See in particular in this collection Terrence Lyons, “The Role of Postsettlement Elections,” 215–236.
 Joanna Spear, “Disarmament and Demobilization,” in Stedman, 141-182.
 Chesterman, You the People, 63.
 A More Secure World: Our Shared Responsibility. Report of the High-level Panel on Threats, Challenges, and Change (New York: United Nations, 2004), p. 69, paragraph 216.
 Stephen John Stedman, “Introduction,” in Stephen John Stedman, Donald Rothchild, and Elizabeth M. Cousens, eds., Ending Civil Wars: The Implementation of Peace Agreements (Boulder, CO: Lynne Rienner, 2002), pp. 9, 11-14. Stedman reserves the term “spoiler” for leaders and factions who use violence to undermine a signed peace agreement, but I believe the term can usefully be applied as well to any post-conflict situation where there is broad domestic participation in the effort to stabilize and rebuild the country but some armed groups use violence to undermine the process for strategic gain. In either case, a judgment must be made (following Stedman’s model) as to whether the violent actor in question is a “total spoiler” who “sees power as indivisible” and seeks to conquer it all, or merely a tactical, “greedy” spoiler who utilizes violence to obtain more power and resources (pp. 12-13). The effort to stabilize Iraq has been bloodily undermined by both total and tactical spoiler forces, and one of the great political failings of the United States has been its inability to separate the latter from the former and draw them into the political process with a judicious mix of assurances and inducements.
 Larry Diamond, Squandered Victory: The American Occupation and the Bungled Effort to Bring Democracy to Iraq (New York: Times Books, 2005): chapter 10.
 James Dobbins et al., America’s Role in Nation-Building: From Germany to Iraq (Santa Monica, Calif.: RAND, 2003): 198.
 James Dobbins et al., America’s Role in Nation-Building, 165–166.
 Stedman, “Introduction,” p. 3. See also George Downs and Stephen Stedman, “Evaluation Issues in Peace Implementation,” in Stedman et al., Ending Civil Wars, 43-69. With regard to commodities, diamonds and timber are much more easily tradable than oil, but oil provides a huge incentive for a potential spoiler group to seek substantial or total control over the area producing it, especially if the oil lies in their traditional area of ethnic and political dominance. In general, the more rents that flow to the new state from tradable commodities or even huge foreign aid flows, the greater the stakes in controlling the state.
 Downs and Stedman, 55-57.
 Stedman, “Introduction,” 3.
 Simon Chesterman, You the People, 257.
 Simon Chesterman, You the People, 257.
 Stephen Krasner, “Building Democracy after Conflict: The Case for Shared Sovereignty,” Journal of Democracy 16, no 1 (January 2005): 69-83.
 Simon Chesterman, You the People, 153.
 Simon Chesterman, You the People, 153.
 Simon Chesterman, You the People, 243.
 James Dobbins et al., America’s Role in Nation-Building, 154.
 Jamal Benomar, “Constitution-Making After Conflict: Lessons for Iraq,” Journal of Democracy 15 (April 2004): 89.
 A More Secure World: Our Shared Responsibility. Report of the High-level Panel on Threats, Challenges, and Change (New York: United Nations, 2004), p. 84, parag 164.
 Ibid, pp.69, 91-92.
 As I envision it, the Department would also incorporate traditional development assistance work now done by USAID and would give the US the ability to engage and coordinate among other donors with the same cabinet-level representation that many other industrialized democracies have for their international cooperation and development work.
 See the cluster of essays, “Elections Without Democracy,” in the Journal of Democracy13 (April 2002): 21-80.
 See the article by Duncan McCargo in the October 2005 Journal of Democracy, forthcoming.