All good things must end. Bad things, on the other hand, seem to get a pass from nature. In the universe observed thus far, life on Earth is anomalous. Birth, growth, order, and building are difficult uphill struggles while death, decay, disorder, and destruction—the engines of entropy—roll on unless stopped. War is humanity’s entropy accelerator. The previous century witnessed war on an unprecedented scale, waged by industrialized states against one another, propelled by secular ideologies of the One Best Way or the one best variety of human. Today, most wars are fought on relatively smaller scales, within nominal state borders, by combinations of quasi-state forces, nonstate forces, and externally orchestrated, state-based interventions. Yet the implications of these smaller wars are global, as the pools of entropy they create are havens or way stations for disorder in the form of two-bit pirates and four-bit drug lords as well as this century’s dominant extremist ideologies, which claim not the mantle of history or superior genes but rather divine warrant to mete out infinite justice as they see fit.

With such forces operating in the background and with the broad array of personal, cultural, and institutional interests at play in any place trying to rise out of conflict, it is obvious that efforts to build peace will be difficult, lengthy, and contested. It should be equally obvious that failure to try to build peace, and the civil order that it implies, only guarantees that entropic pools will grow and connect—if not geographically, then commercially (where “commerce” includes all forms of interchange at a distance). The questions for international engagement, then, involve capacity (the ability to undertake a task), resources (the money, people, and time available to underwrite capacity), understanding (of what a specific environment needs to transition to peace and how applied capacity and resources are likely to affect that transition), goals (what applied capacity and resources are supposed to achieve, based on understanding), and exit strategies (plans for executing reduction or disengagement that may be triggered by reaching transition thresholds for key mission goals or by failing to reach those thresholds within a specified time, specified expenditure ceiling, or the parameters of political tolerance of the host state or the engaging country, countries, or organizations).

This article focuses on exit strategies for peace support operations (PSOs) but addresses other elements of the engagement chain and other operations to illustrate how exit strategies are influenced and altered by those other elements. It is as much about how PSOs do end as about how they should end.

Defining Relevant Scope

The operations of greatest interest here are those that have some element of state-building in their mandates or that have contributed to state-building indirectly—for instance, by laying the groundwork for future state-building operations. Twenty-one states and territories have been host to twice as many relevant PSOs (a number of states have hosted operations sequentially and/or several simultaneous missions led by different entities), as indicated in table 1. Yet only 12 “completed” missions have seen security and governance transition to local hands.

Table 2 categorizes operations from a maximum to a minimum level of state-building authority or involvement. For each type of operation, the table sketches the essential characteristics and lists the main factors driving its exit strategy, divided into factors within and beyond that operational type’s direct control. The final column notes unique exit issues.

For the first three types of operations, I have borrowed and relabeled concepts from Jarat Chopra et al.¹ Their work on the requirements of forceful international peace implementation predated and prefigured the United Nations (UN) missions in Kosovo and Timor-Leste, which are instances of international administration or interim law enforcement to fill a vacuum in local governance.

Directive state-building is the best way to describe the painful evolution of the international presence in Bosnia and Herzegovina under the 1995 Dayton Accords or the “twinning” of a new generation of Liberian officials with international mentor/monitors in the World Bank Government Economic Management Assistance Program, where the internationals had cosigning authority on government expenditures.²

Collaborative state-building might best be exemplified by present UN operations in Haiti, as all mandated activities are undertaken in support of the elected government. Since the earthquake of January 2010, the UN Stabilization Mission in Haiti has doubled its armed police presence and helped to keep order in cooperation with the Haitian National Police.³

Securing humanitarian relief has palliative rather than political objectives, although such action always affects political outcomes by feeding the besieged or the displaced or by putting off more decisive international action. UN operations in pre-Dayton Bosnia had such a focus on humanitarian relief, as did the U.S.-led relief force in Somalia in 1992–1993. Both relief efforts transitioned to directive state-building operations.

Creating political space means providing international support for a stable host-state security environment, as the UN Mission in Côte d’Ivoire has strived to do since 2004.

Providing fair witness, often labeled “traditional” peacekeeping or peace observation, does not directly engage the politics of peacemaking or the resolution of national animus. As the price of that detachment, exit depends on actions and agreements in other venues. Fair witness operations serve the purposes of state-building by helping to reduce the probability of state damage or destruction associated with resumed conflict.

Exit Strategy for a Complex PSO

The operating environments of complex PSOs are like parties with ugly drunks in a variety of guises: disgruntled factions who do not like their cut of the peace deal, former combatants who cannot find work, arms and commodities smugglers and traffickers in human beings—all of whom thrive in a climate of anarchy; avaricious warlords; and corrupt government officials. Exit would be easy if peacekeepers could simply put down their tools and walk away at a moment’s notice, but there are always substantial constraints, from the reputations of those who constitute, authorize, or lead the operation, to concerns about host-state political and economic stability and the knock-on effects of exit for the surrounding region.

Exit strategies always entail elements of prediction. The wisdom of any given exit-related choice may only be demonstrable (or refutable) several years after the choice, as the exited territory either remains stable or dissolves into disorder again. Any transition or exit plan thus embodies projections about what is likely to happen after transition/exit. Milestones met relate to past and present action and do not necessarily indicate future performance, especially independent host-state performance. A positive trend may have been due to the presence of an operation or the interaction of that presence with local parties. Trends and projections from other operations may or may not be valid locally owing to just such strong interactions between international engagement and the specifics of the engagement environment. International election security and logistics may have been effective in one operation, for example, but the effect may have been magnified by terrain and infrastructure favorable to the operation (as with a smaller area of operations, lower mountains, more open terrain, less extreme temperatures, no seasonal flooding, good roads, or plentiful airstrips).

Exit strategies are also political constructs. Gideon Rose observed that the term exit strategy was not applied to foreign or military policy matters before the 1993 U.S. engagement in Somalia.⁴ Jeffrey Record argued similarly that, pre-9/11, the pressure for clear exit strategies was greatest “among those who believe that the military should not be exposed to the risks of peace enforcement and other small-scale contingencies,” making exit strategy a kind of barrier to entry for operations that politicians or military leaders preferred not to undertake.⁵

America’s preferences for its own forces notwithstanding, as a member of the Security Council, the United States has voted frequently to establish new UN operations for which the lack of an exit strategy has not seemed to pose any barrier to council action. The United Nations and its member states have been invited repeatedly to send troops, police, and civilian personnel into semistable situations still aspiring to be postconflict without clear guidance on how to see their mandated tasks through to a successful outcome.

Through the mid-1990s, the dominant benchmark indicating fulfillment of a complex UN operation’s mandate was the conduct, supervision, or certification of host-state national elections for constituent assemblies (Namibia) or for national parliaments and executives (Cambodia, Mozambique, Angola, and Bosnia). An election is a logistically and politically complex endeavor but, properly conducted, has an outcome that is objectively measureable, is technically achievable within an operation’s mandated means, occurs at a predictable time, and involves a time-limited activity. It is the easiest sort of benchmark to meet. Elections are visible symbols of apparent transition from wartime warlord politics to the transparency and public order that peacebuilders seek. But for elections to be effective, all local parties must support them (as in Namibia), and those who waver must either be brought back into line (as was the Mozambican National Resistance) or decide not to interfere (as did the Khmer Rouge in Cambodia). Where irritable losers return to violence (as did Jonas Savimbi in Angola and Hun Sen in Cambodia), the consequences can weigh heavily on the population, in terms of not only lives but also lost credibility for democracy. Premature elections can also cement nationalist or extremist leaders’ hold on power: the September 1996 national elections in Bosnia confirmed nationalist parties’ power and helped ensure the continued need for the presence of the North Atlantic Treaty Organization (NATO).⁶

These negative experiences demonstrated that quick elections are not enough to generate lasting success for a complex PSO except in the easiest of political circumstances. Multiple analyses suggest, however, that complex PSOs tend to be sent into more difficult circumstances, where the first election is the starting block for democratic state-building, not the finish line.⁷ Debate continues over how much democratic state-building is enough to reinforce post-peacekeeping political stability and even whether democracy itself is a feasible or desirable near-term peacebuilding goal. Jack Goldstone and company argue for fully institutionalized democracy as the most stable form of governance consistent with human rights,⁸ while Max Manwaring and Kimbra Fishel argue that locally defined “legitimate” and “competent” governance is the best that outsiders can reasonably expect to achieve.⁹ Roland Paris advocates a similar “institutionalization before liberalization,” also known as IBL.¹⁰ In order to work, however, IBL requires outside manipulation of internal politics: encouragement of “moderate” political parties, “lustration” (which is defined as “ceremonial purification” but which can be unceremonious in practice, as with de-Ba’athification in Iraq), and design of electoral systems that reward moderation. In short, IBL more or less requires an ironically authoritarian international trusteeship that continues “for as long as it takes” to produce free and open politics and markets.¹¹ Experiences in the Balkans confirm that reaching such goals takes time. The ethnic Albanian majority of Kosovo grew tired of waiting for the process to finish and just declared independence in 2008. But even if local parties everywhere exhibited great patience and cooperation, the international resources committed to building fully institutionalized democracy would need to grow far beyond present levels of investment (about $10 billion–$12 billion annually, outside of Afghanistan and Iraq) if more than a handful of places were to benefit. The debate also begs the question of when and to what degree international security providers, in particular, can disengage from the peacebuilding process.

Prioritizing a Path to the Exit

For some years, and especially since the establishment of interim international governance in Kosovo and East Timor, analysts have been grappling with the issue of priorities for complex PSOs. Three examples include Graham Day’s policekeeping model,¹² the Beginner’s Guide to Nation-Building by James Dobbins et al.,¹³ and the seven-step program of Michael Doyle and Nicholas Sambanis.¹⁴

Day assigned priority to stopping the war (a military task); then to providing public security and governance, including contract enforcement (policing and civilian tasks that resume as wartime violence ebbs); next, jump-starting the local economy (a financial donor/development agency task that requires public order and law enforcement to attract investment and allow business to generate jobs at a reasonable level of risk); and finally, reconciliation (a civil society task best undertaken when more people have work and are meeting basic needs other than through international relief). Each successive phase begins at some point after the preceding phase begins but before it is complete, in a pattern Day describes as a cascade. Appropriate starting points for each phase will vary with the mission, and all physical and institutional reconstruction efforts should be designed for local sustainability in operations, maintenance, and cost, from the capacities of hospitals to the usability of mission office space sans air conditioning, if it is intended for transfer to local use upon drawdown.

Dobbins et al. proposed the following priorities in the indicated order: public security and humanitarian assistance (including the return of refugees and internally displaced persons), governance (restoring public services and administration), economic stabilization (including a stable currency and a legal/regulatory framework for commerce), democratization (including an essential free press), and long-term development aid. Democracy is accorded a lower priority than standing up a government that can provide law enforcement, education, public health, power, and telecommunications (often regulated at the national level), plus water and sanitation (usually the job of local levels of government).

Doyle and Sambanis see the greatest prospects for success (and exit) in UN missions with “transformational” mandates and authority to use “discrete” enforcement measures against noncompliance in a steady international effort that may last 10 or more years.¹⁵ Step one is internal security, with a new “sovereign Leviathan,” either domestic or international, to enforce order. Step two is regional security, to get the neighbors on board the peace process. Step three is “quick wins to generate popular support and buy time”: food distribution, health clinics, reliable power, and cleaning up rubble. Step four implements “the rule of law and constitutional consent” as the “foundations of all that follow.” Step five involves attention to property rights, which “the poor need even more than the rich.” Step six is democracy or other “wider participation” that is “essential for the long run.” Step seven is moral and psychological reconciliation.

Past their step two, it is not clear that a complex peace operation with full security functions need be present to achieve their remaining peacebuilding goals. Whether it can draw down or depart completely depends in large part on the effectiveness of the operation and donor cooperation in training and equipping acceptable new armed forces and police services and, presumably, criminal justice systems (prosecutors, judges, defense lawyers, recordkeeping, and corrections facilities). These have all been slow to reform in recent operations; if they are the responsibility of the operation, they may be the pacing factor for its exit. A nagging question for both the Doyle and Sambanis and the Dobbins/RAND models is that of oversight: if democracy appears late in the process, who or what legitimates and monitors the activities of the governing institutions built with outside support? What gives them the incentive to provide efficient public services or to buy into rule of law and constitutionality? Absent a role for voters, such oversight must come from other states or international institutions, which means that the effort to promote efficient, effective government is externally imposed rather than home grown, with implications for whether it will embed, politically or culturally. This is only one of many dilemmas inherent in peacebuilding.¹⁶

Exit as Process (Transition) Rather Than Event

Rose stressed that exit should be conceptualized as a process, not an event.¹⁷ As noted, for UN operations, exit was an event until the latter 1990s. Phased withdrawal from Somalia simply shrank the mission’s perimeter over the course of 15 months, ending with its final rapid evacuation from Mogadishu in March 1995 by a covering force of U.S. Marines. Some months later, UN forces in Bosnia were replaced by (and in many cases rehatted as) NATO forces. Somalia and Bosnia remind us, as if it were necessary, that transitions may result from mission failure as well as success. The textbox summarizes the array of potential outcomes.

It is a commonplace that operations—and transitions—should be milestone- rather than schedule-driven. But certain operations are most often designed to be schedule-driven. These include rapid-reaction interventions by coalitions of willing states (in Haiti, 1994 and 2004, and in East Timor, 1999) and military interventions to date by the European Union in Africa (in the Democratic Republic of the Congo for 3 months in 2003 and 6 months in 2006, and in Chad for 12 months in 2008–2009). All of these short-term interventions handed over or handed back responsibility to UN peace operations. Such schedule-driven transitions are successful for the originating organizations if they transition on time, but are perhaps less successful from the perspective of those local or international interests they supported temporarily or those who need to act rapidly to fill the security vacuum that might otherwise result from their withdrawal.

Lessons from the ragged U.S.-to-UN handover in Somalia were applied successfully to the 1995 handover in Haiti. A substantial U.S. contribution to the first 15 months of the UN operation was followed, however, by successive reductions in the size of the UN force, a narrow emphasis on training the Haitian National Police, and then dropping even that goal as the police themselves were reabsorbed into the country’s still-corrupt political and economic system. The UN’s Haiti operations in the 1990s ended with a sense of futility.

The United Nations began serious planning for drawdown of their complex operations in perhaps 2001 with the UN Transitional Administration in East Timor (UNTAET) and the UN Mission in Sierra Leone. UNTAET served for a little over 2 years as de facto sovereign, relinquishing political control to a newly elected government in May 2002. A smaller follow-on mission, the UN Mission of Support in East Timor, remained for 3 more years, handing off in turn to a largely political advisory mission, the UN Office for Timor-Leste. The latter could not manage the tensions within and between the Timorese security forces and political leadership, leading to disorder that in turn caused the government to request outside help. An Australian-led multinational force responded quickly and remains in Timor-Leste at this writing—a major exception to the rule on rapid coalition withdrawals. The force did not revert to UN command, however, when a new policing mission, the UN Integrated Mission in Timor-Leste (UNMIT), began its work in August 2006. UNMIT’s 1,600 police were once again given law enforcement powers.¹⁸ Timor-Leste is thus the current best example of a premature transition.

Sierra Leone held national elections in 2002 at about the same time as Timor-Leste. Judged free and fair, they marked the start of the gradual drawdown of UN forces from a peak of 17,500 in 2002 to none at the end of 2005. The follow-on UN Integrated Office for Sierra Leone continued to work with the national police, a British military training team also remained in place, and the country enjoyed a peaceful change of party in power following national elections in 2007. Here, then, the United Nations seems to have judged the circumstances correctly, planned its exit carefully, and executed it over a period of years. Its ability to execute that strategy benefited from the early collapse of the government’s main adversary, the Revolutionary United Front, whose leader died in custody and whose forces were roughed up by the Guinean army and by British military intervention. The Special Court for Sierra Leone (SCSL) indicted not only local militia leaders but also the sitting president of Liberia, Charles Taylor, who at this writing remains on trial in the SCSL’s chamber in The Hague, indicted for criminal support of the Revolutionary United Front.

The Democratic Republic of the Congo (DRC) lacks everything that Sierra Leone enjoys in terms of the passing of its warlords, the free election of a new generation of leaders, regional support, an ethnically nonhostile social structure, and small size. DRC warlords were elected to high office in national elections, but most retained their provincial links. Many neighboring states are unstable and those that are not are actively exploiting DRC ethnic divisions and economic resources.¹⁹ The country is so large, with such a weak communications infrastructure, that it is difficult for the government, based in the far southwestern corner, to maintain much influence over its provinces bordering Burundi, Rwanda, and Uganda. The UN operation in the DRC has nearly 20,000 troops who are mostly deployed in the unstable east, where they are outnumbered by sundry militias and the ragtag Congolese national army. Still, the UN Mission in the Democratic Republic of the Congo (MONUC) has helped tie the country together—physically with Air MONUC and electronically with Radio Okapi, which blankets the country with nationally and regionally based programs, uses Congolese reporters and announcers, and offers a mix of cultural programming, music, and objective news. An exit strategy for MONUC would be dauntingly complex, but the keys to its exit may be some combination of ridding DRC of the last Rwandan fighters (thus relieving Rwanda of an excuse to meddle in Congolese affairs) and gaining control of its own mineral resources. Congo’s leadership, however, remains in the hands of the same men who tore it apart and indirectly caused the displacement and death of more than 4 million of their fellow citizens.²⁰ The government of President Joseph Kabila demanded that MONUC withdraw in time for the country’s 50th anniversary of independence in June 2010, but continued instability in the east led the Security Council to instead reconfigure a slightly reduced UN presence as a stabilization force focused primarily on protection of civilians.²¹ The likelihood of a dignified exit, or any exit, remains low for the mission, rebranded as the UN Mission for Stabilization of Democratic Republic of the Congo (MONUSCO).

Exit Strategy: Know It When You See It?

This sample of the problems and prospects for exit strategies in the world of complex peace operations is necessarily nonscientific because the success of these operations is a matter of circumstance and art: the art of diplomacy, the art of (selective) war, the art of reconciliation, and the art of the law and politics. No definitive formula for a success-based exit is possible, although grounds for exit can be seen in the accomplishment of key elements of a mission’s mandate. Much, therefore, rides on the content of that mandate, the knowledge and wisdom that went into its preparation, and whether there are opportunities to refine it over time to better fit circumstances on the ground and to adapt to them as they change. This is one area in which UN mission mandates may hold an advantage over those of some other organizations: just a dozen states must be convinced of the need to change the way a mission does business, assuming that a veto-wielding permanent member does not object to the alteration.

Flexibility in the Security Council notwithstanding, it is perfectly possible for an operation to devise an exit strategy consistent with fulfillment of its mandate and then be utterly unable to implement it owing to circumstances on the ground, the recalcitrance of local leaders, or insufficient resources—the external factors noted in table 2. Recall, for example, that while MONUSCO has many troops by UN standards, it has built up to that number in trying to secure a land the size of Western Europe with less than half the number of troops that NATO initially deployed to secure Kosovo, a parcel that could be a large ranch in one of Congo’s central or eastern provinces. Rather than an exit strategy, an operation such as MONUSCO is more likely to have an influence strategy, a survival strategy, a retrenchment strategy, and the occasional attack strategy. It cannot succeed in any strategic sense, however, without major political cover of the sort that Doyle and Sambanis and many other analysts stress as necessary to promote regional good behavior. Without such continuous external support, a country like Congo and an operation like MONUSCO will be stuck in a perpetual “Groundhog Day” cycle of endlessly reliving the same militia gambits, dealing with the same greedy officials, and facing the same people formerly so full of hope that tomorrow might be a little better than today, but lately more likely to stone peacekeepers who have proven unable to keep them safe. PRISM

Notes

¹ Jarat Chopra et al., “Political Tasks of Peace Maintenance,” Small Wars and Insurgencies 10, no. 2 (Autumn 1999), 21–39.

² Alix J. Boucher et al., Mapping and Fighting Corruption in War-Torn States (Washington, DC: Henry L. Stimson Center, 2007).

³ United Nations (UN), Twenty-fourth Report of the Secretary-General on the United Nations Operation in Côte d’Ivoire, S/2010/245, May 20, 2010.

⁴ Gideon Rose, “The Exit Strategy Delusion,” Foreign Affairs 77, no. 1 (January–February 1998), 56–67.

⁵ Jeffrey Record, “Exit Strategy Delusions,” Parameters 31, no. 4 (Winter 2001), 21–27.

⁶ Elizabeth Cousens and David Harland, “Post-Dayton Bosnia and Herzegovina,” Twenty-First-Century Peace Operations, ed. William J. Durch (Washington, DC: U.S. Institute of Peace, 2006), 94–96.

⁷ William J. Durch and Tobias C. Berkman, Who Should Keep the Peace? Providing Security for Twenty-First-Century Peace Operations (Washington, DC: Henry L. Stimson Center, 2006), 14–15.

⁸ Jack A. Goldstone et al., “A Global Forecasting Model of Political Instability,” paper presented at the annual meeting of the American Political Science Association, Washington, DC, September 3, 2005, available at <www.globalpolicy.gmu.edu/pitf/pitfp5.htm>; “A Global Forecasting Model of Political Instability,” American Journal of Political Science 54, no. 1 (January 2010), 190.

⁹ Max Manwaring and Kimbra Fishel, “Lessons That Should Have Been Learned,” in The Savage Wars of Peace, ed. J. Fishel (Boulder, CO: Westview Press, 1998), 198–209.

¹⁰ Roland Paris, At War’s End: Building Peace after Civil Conflict (Cambridge, UK: Cambridge University Press, 2004), 190–200.

¹¹ Ibid., 206.

¹² Graham Day, “Policekeeping,” mimeo (Washington, DC: U.S. Institute of Peace, 2001).

¹³ James Dobbins et al., The Beginner’s Guide to Nation-Building (Arlington, VA: RAND, 2007).

¹⁴ Michael W. Doyle and Nicholas Sambanis, Making War and Building Peace: United Nations Peace Operations (Princeton: Princeton University Press, 2006).

¹⁵ Ibid.

¹⁶ Roland Paris and Timothy D. Sisk, eds., The Dilemmas of Statebuilding (London: Routledge, 2009).

¹⁷ Rose.

¹⁸ UN, “United Nations Mission of Support in East Timor,” 2005, available at <www.un.org/Depts/dpko/missions/unmiset>.

¹⁹ Global Witness, “Democratic Republic of the Congo,” available at <www.globalwitness.org/pages/en/drc_conflict.html>.

²⁰ Benjamin Coghlan et al., Mortality in the Democratic Republic of Congo: An Ongoing Crisis (Washington, DC: International Rescue Committee, 2007).

²¹ UN, Security Council Resolution 1925, S/RES/1925, May 28, 2010.