Fearon, James D., and David D. Laitin. 2004. "Neotrusteeship and The Problem of Weak States." International Security 28 (4) - 5-43

Neotrusteeship and the Problem of Weak States
Neotrusteeship and the James D. Fearon and

David D. Laitin
Problem of Weak States
G
eorge W. Bush and
his administration came into ofªce with a self-consciously realist orientation in
foreign policy. The president and his advisers derided the Clinton administra-
tion’s multilateralism as mere form without national security substance. They
viewed Russia and China as the main potential threats or sources of danger,
and regarded Bill Clinton as a naïve idealist for neglecting these great powers
in favor of “foreign policy as social work”—humanitarian ventures in areas pe-
ripheral to U.S. national security concerns.1 Consistent with a realist suspicion
of multilateralism and conªdence in self-help, the administration’s principal
foreign policy project in its ªrst months was the unilateral pursuit of ballistic
missile defense.
The Bush team was particularly critical of U.S. participation in quixotic
efforts at nation building for failed states. As a candidate, Vice President Dick
Cheney created a signiªcant ºap in August 2000 when he suggested that the
Bush administration would end U.S. participation in NATO’s Bosnia mission.2
Condoleezza Rice, who would become Bush’s national security adviser,
expressed dismayed amazement that U.S. troops were being used to take chil-
dren to kindergarten in Bosnia.3 The message was clear: The Bush administra-
tion would not engage in state-building efforts.4
James D. Fearon and David D. Laitin are Professors of Political Science at Stanford University.
We are extremely grateful to the Carnegie Corporation for funding this research, and to Stephen
Del Rosso of the Carnegie Corporation for his help and encouragement at all stages of this project.
We acknowledge the assitance of Salman Ahmed, who encouraged our project at its early stages
and also introduced us to the complexities of the UN. We have received helpful comments from
Michael Doyle, William Durch, Lynn Eden, Michèle Grifªn, Robert Keohane, Stephen Stedman,
J. Matthew Vaccaro, and three anonymous reviewers on earlier versions of this article, for which
we are also grateful.
1. See especially Condoleezza Rice, “Promoting the National Interest,” Foreign Affairs, Vol. 79,
No. 1 (January/February 2000), pp. 45–62. The epithet “foreign policy as social work” is from the
article by that title by Michael Mandelbaum, Foreign Affairs, Vol. 75, No. 1 (January/February
1996), pp. 16–32, the drift of which is echoed in Rice’s essay.
2. For Vice President Al Gore’s acrid response, see Michael Cooper, “The 2000 Campaign: The Re-
publican Running Mate; Cheney Urges Rethinking Use of U.S. Ground Forces in Bosnia and
Kosovo,” New York Times, September 1, 2000, p. A22.
3. David E. Sanger and Katharine Q. Seelye, “Gore Defends Civilian Uses of Military,” New York
Times, October 4, 2000, p. A4.
4. Since nation building implies the construction of a common nationality, “state building” is a
more accurate term for what both the Clinton and Bush administrations have attempted—namely,
the reconstruction of polities and economies. We use “state building” henceforth.
International Security, Vol. 28, No. 4 (Spring 2004), pp. 5–43

© 2004 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.
5
International Security 28:4 6
Ironically, the Bush administration has since undertaken state-building pro-

jects that are vastly larger and more difªcult than anything the Clinton admin-
istration ever attempted. The U.S. military is now building kindergartens in
Afghanistan, in addition to paving roads and assisting with many other major
infrastructure projects in both Afghanistan and Iraq.5 GIs report on instructing
Iraqis in how to run a town meeting with an agenda and turn taking—”It’s
basic P.T.A. stuff,” one commented.6 These are local-level complements to the
complex, higher-level efforts to build workable national political structures in
both countries. And all this is happening without any signiªcant reduction in
U.S. involvement in ongoing peacekeeping operations in Kosovo or Bosnia.
Indeed, the Bush administration even took on new peacekeeping responsibili-
ties in Liberia, albeit very small ones thus far.7
It can be argued that despite the apparent about-face, the Bush administra-
tion has actually kept true to its realist principles. It is attempting to rebuild
“rogue” states that the United States attacked and destroyed as perceived
threats to national security, rather than states that failed largely on their own.
Arguing that chaos in Liberia does not threaten U.S. security, Pentagon ofªcials
successfully resisted the strong “CNN effect,” as well as international and pos-
sibly State Department pressure for a more active U.S. role. In broader terms,
the administration claims that the terrorist attacks of September 11, 2001,
changed the game, clarifying a new security threat.
We argue to the contrary that the Bush administration’s brand of realism has
collided with post–Cold War realities that shaped the Clinton administration’s
foreign policy as well. Even before September 11, the world was changing in
such a way that the main security threats and problems now emerge not from
great power security competition—Russia and China, for example—but from
the consequences of political disorder, misrule, and humiliation in the third
world. These threats and problems have the character of “public bads” for the
major powers. That is, collapsed states and rogue regimes seeking nuclear
weapons impose diffuse costs on the major powers and other states. The total
5. See Carlotta Gall, “In Afghanistan, Violence Stalls Renewal Effort,” New York Times, April 26,
2003, p. A1, who reports on the effects of the U.S. military’s fall 2002 decision to use military civil
affairs teams, backed by Special Forces for protection, to undertake reconstruction projects outside
of Kabul.
6. Col. Ralph Baker, cited in Thomas Friedman, “Starting from Scratch,” New York Times, August
27, 2003, p. A21.
7. On the diplomatic front, the Bush administration has supported the creation of a very large UN
peacekeeping operation to Liberia, major expansion in the UN’s mission to the Congo (Kinshasa),
continuation of a state-building mission in East Timor, and creation of a small mission to the Ivory
Coast.
Neotrusteeship and the Problem of Weak States 7
costs are often large enough that it would pay to address them, but not so large
that doing so is necessarily worthwhile for any one state. Given the nature of
the problem, the incentives for burden sharing through multilateral arrange-
ments are strong. Furthermore, whether the problem is a failed state or a rogue
regime that has been attacked and destroyed, state-building efforts led by ma-
jor power interveners and international organizations are practically inevitable
(to the Bush administration’s chagrin).
As a result, the United States is now drawn toward a form of international
governance that may be described as neotrusteeship, or more provocatively,
postmodern imperialism. The terms refer to the complicated mixes of interna-
tional and domestic governance structures that are evolving in Bosnia, Kosovo,
East Timor, Sierra Leone, Afghanistan and, possibly in the long run, Iraq. Simi-
lar to classical imperialism, these efforts involve a remarkable degree of control
over domestic political authority and basic economic functions by foreign
countries. In contrast to classical imperialism, in these new forms of rule sub-
jects are governed by a complex hodgepodge of foreign powers, international
and nongovernmental organizations (NGOs), and domestic institutions, rather
than by a single imperial or trust power asserting monopoly rights within its
domain. In contrast to classical imperialism but in line with concepts of trust-
eeship, the parties to these complex interventions typically seek an interna-
tional legal mandate for their rule. Finally, whereas classical imperialists
conceived of their empires as indeªnite in time, the agents of neotrustee-
ship want to exit as quickly as possible, after intervening to reconstruct or
reconªgure states so as to reduce threats arising from either state collapse or
rogue regimes empowered by weapons of mass destruction (WMD).
Can this evolving form of neotrusteeship effectively coordinate international
action to address the problems posed by collapsed states? Much more than is
generally appreciated, the approach worked in the 1990s, with the United Na-
tions playing a central role. Yet the international system remains badly orga-
nized and badly served for dealing with the implications of state collapse,
whether indigenous or induced by invasion.
In the body of the article, we analyze the largely ad hoc arrangements for
peacekeeping and “transitional administration” as they evolved in the 1990s.
First, we discuss in greater detail what we mean by neotrusteeship and why it
is happening.8 Second, we consider a remarkable piece of self-criticism by the
8. Since Gerald B. Helman’s and Steven R. Ratner’s prescient call for some form of UN “conserva-
torship” in 1992, a small but growing number of analysts has suggested new international institu-
tional arrangements for dealing with collapsed states. See Helman and Ratner, “Saving Failed
UN, the Brahimi Report of August 2000,9 which diagnosed the troubles with
UN peacekeeping operations (PKOs) as practiced in the 1990s, when these mis-
sions spread around the globe. Third, we argue that neither the UN’s nor the
United States’ political and military doctrine for postconºict intervention
grasps the central problem posed by weak states—namely “mission creep.”
Mission creep toward state building is practically inevitable when the state’s
administrative and police apparatuses barely function, and other environ-
mental conditions favor guerrilla techniques.
In the fourth section, we identify four principal problems with the current
system for organizing and implementing international collective action in col-
lapsed states: recruitment, coordination, accountability, and exit. We offer
some proposals for how these problems might be addressed, though we do not
pretend that our suggestions are wholly satisfactory or without problems
themselves.
On the question of recruitment—who will pay for and provide these interna-
tional public goods?—we maintain that the nature of the problem favors multi-
lateral interventions led by a major power with regional or other interests at
stake, supported by troops from developing countries and ªnanced in part by
loans to the collapsed state to be repaid after reconstruction. On the problem of
coordination—how to rationalize and harmonize the activities of the diverse
state and nonstate actors involved in these projects—we argue that in anarchi-
cal settings the UN is an inappropriate lead agent and that greater efªcacy and
coordination is likely to result from missions led by a major power with a dom-
inant military force. On the problem of accountability—how to make trustees
responsible for their actions without exacerbating the recruitment problem—
we recommend a newly constructed arm of the UN to address some of the
issues once handled by the now-defunct UN Trusteeship Council.10
Finally, we argue that in countries whose state institutions have been de-
States,” Foreign Policy, No. 89 (Winter 1992), pp. 3–18; Richard Caplan, A New Trusteeship? The Inter-
national Administration of War-torn Territories, Adelphi Paper No. 341 (Oxford: Oxford University
Press, 2002); Robert O. Keohane, “Political Authority after Intervention: Gradations in Sover-
eignty,” in J.L. Holzgrefe and Keohane, eds., Humanitarian Intervention: Ethical, Legal, and Political
Dilemmas (Cambridge: Cambridge University Press, 2003); Sebastian Mallaby, “The Reluctant Im-
perialist: Terrorism, Failed States, and the Case for American Empire,” Foreign Affairs, Vol. 81,
No. 2 (March/April 2002), pp. 2–7; and Stephen D. Krasner, “Troubled Societies, Outlaw States,
and Gradations of Sovereignty,” Stanford University, July 20, 2002.
9. The report is conventionally so called to acknowledge the chair of the committee that produced
it, Lakhdar Brahimi. See United Nations, Report of the Panel on Peace Operations, A/55/305-S/2000/
809, August 21, 2000.
10. Keohane, “Political Authority after Intervention,” identiªes and analyzes this same problem.
stroyed by years of civil war, the problem of exit is nearly intractable. In such
cases, a long-term international contribution to local security will likely be nec-
essary. To provide the right incentives for the development of local capacity
and to limit the costs of potential interveners, we advocate international sup-
port for developing local tax-collecting capability, coupled with tax sharing be-
tween domestic and international agents as an incentive to move the country
from international welfare toward self-governance. For some cases, we suggest
the notion of transfer not to full sovereignty but rather to an entity embedded
in and monitored by international institutions.11
What Is Neotrusteeship and Why Is It Happening?
Despite a common media perception, the prevalence of civil war since the
early 1990s is not a post–Cold War phenomenon. Instead, it is the result of the
steady accumulation of protracted civil conºicts since the early 1950s. Figure 1
shows the number of ongoing civil wars by year for the period 1945 to 2002.12
Although there was indeed a signiªcant jump in the early 1990s associated in
part with the end of the Cold War, the current level of twenty-four ongoing
civil wars had already been reached and exceeded by the mid-1980s.
In sharp contrast, the proliferation of international interventions that include
11. In researching this article, we conducted interviews with ofªcials in the United Nations, sev-
eral U.S. government agencies involved in PKOs, and the Russian government and military. We
conducted nine formal interviews from April 2 to April 5, 2001 in New York, principally among
ofªcials at the UN Department of Peacekeeping Operations, the UN Department of Political Af-
fairs, and several diplomatic missions to the UN. We thank John Ruggie and Graciela Bazet-
Broitman for facilitating these interviews. We conducted nine interviews in Washington, D.C., on
March 18–21, 2002, interviewing principally present and former ofªcials in the Department of
State, the Department of Defense, and the Central Intelligence Agency. We thank Joseph McGhee,
Peter Cowhey, and the Institute for Global Cooperation and Conºict at the University of California
for facilitating these interviews. David Laitin conducted fourteen formal interviews in Moscow
from June 24 to July 2, 2002, principally among ofªcials at the Ministry of Defense, the Ministry
for CIS [Commonwealth of Independent States] Affairs, the Collective Security Council of the CIS,
the Ministry of Nationalities, and the State Duma. We thank Andrei Melville, Oleg Shulga, and
Boris Bykov for facilitating these interviews. All interviews were conªdential. Footnoted refer-
ences to these interviews represent place of interview (DC ⫽ Washington, D.C.; M ⫽ Moscow; and
UN ⫽ New York), number of interviewee (in our research ªles), and date.
12. Civil wars are deªned here as conºicts among organized groups within a state for state or re-
gional power that kill at least 1,000 individuals over their course, with at least 100 dead on each
side, and an average of at least 100 killed per year. For our full deªnition and coding criteria, see
James D. Fearon and David D. Laitin, “Ethnicity, Insurgency, and Civil War,” American Political Sci-
ence Review, Vol. 97, No. 1 (February 2003), pp. 75–90. Because the number of countries increased
markedly over the 1945 to 2002 period, we also show in Figure 1 the percentage of countries expe-
riencing civil war by year. The sample refers to countries that had a population of at least 500,000
in 1990.
Figure 1. Civil War and UN PKOs since 1945.
a UN peacekeeping operation—a good indicator of the spread of neotrustee-

ship—is a post–Cold War phenomenon. Figure 1 also shows the number of UN
PKOs by year. It reveals a dramatic change beginning in 1988, an increase from
ªve per year in the 1980s to sustained levels in the high teens since the early
1990s.13 The number of countries with an ongoing PKO has been only slightly
lower. Since 1993, it has varied between ten and ªfteen, or 6.5 to 10 percent of
the world’s countries that have a population of at least half a million. In total,
the Security Council has mandated some forty-ªve peacekeeping missions
since 1988, as compared with just thirteen from 1948 to 1987.14
13. We collected the data from the UN Department of Peacekeeping Operations website, http://
www.un.org/Depts/dpko/dpko/home.shtml. For valuable case studies of PKOs in this period,
see William J. Durch, ed., The Evolution of UN Peacekeeping (New York: St. Martin’s, 1993); William J.
Durch, ed., UN Peacekeeping, American Power, and the Uncivil Wars of the 1990s (New York: St. Mar-
tin’s, 1996); Michael Doyle, Ian Johnstone, and Robert C. Orr, eds., Keeping the Peace: Multidimen-
sional UN Operations in Cambodia and El Salvador (Cambridge: Cambridge University Press, 1997);
and Stephen John Stedman, Donald Rothchild, and Elizabeth M. Cousens, eds., Ending Civil Wars:
The Implementation of Peace Agreements (Boulder, Colo.: Lynne Rienner, 2002). Two analytical contri-
butions are George Downs and Stephen John Stedman, “Evaluation Issues in Peace Implementa-
tion,” in Stedman, Rothchild, and Cowens, Ending Civil Wars, on the determinants of PKO success;
and Stephen John Stedman, “Spoiler Problems in Peace Processes,” International Security, Vol. 22,
No. 2 (Fall 1997), pp. 5–53.
14. These totals are in a way misleading as compared with the annual numbers reported in
Figure 1, because they include small precursor and follow-on missions that might better be
The nature of the PKOs also changed with the end of the Cold War. Of the
thirteen pre-1988 missions, eight were classic chapter 6 operations in which
UN peacekeeping forces monitored a border or cease-ªre line after an inter-
state war, and one was a close equivalent within a state (UNFICYP in Cyprus,
especially since 1974). There have been only four new operations of this sort
since 1988.15 The rest have virtually all addressed problems of peacekeeping
and state reconstruction in countries torn by civil war. These missions have
varied dramatically in size, complexity, and core tasks, ranging from provision
of a few dozen military observers of a cease-ªre agreement to operations with
tens of thousands of peacekeeping forces, the organization and supervision of
elections, or transitional administration in which the UN takes over some or all
day-to-day government of the country or region in question.16
The spread of such complex UN PKOs is an indicator of neotrusteeship, but
it is not the same thing. In many cases, the UN is not the only important inter-
national intervener. States, other international and regional organizations,
NGOs, and corporations are often heavily involved in collapsed-state gover-
nance and reconstruction. NATO troops are supplying security and state-
building services in Bosnia, Kosovo, and Afghanistan. The 1995 Dayton
accords created an entity known as the Ofªce of the High Representative that
approximates the sovereign power in Bosnia, although it does not have power
of command over NATO troops there. The special representative of the secre-
tary-general (SRSG) in Kosovo has formal authority over Kosovo’s transitional
administration, although the European Union (EU), the Organization for Eco-
nomic Security and Cooperation (OSCE), and NATO also play major, formal-
ized roles in political institution building, economic reconstruction, and
provision of security. Afghanistan is presently governed by a regime that was
created and blessed by international negotiations in Bonn 2001, but owes its
continued existence and form to the presence of U.S. troops. NATO peacekeep-
ing forces under a UN mandate provide security in Kabul, while an extensive
UN mission trains police and engages in many other facets of state building
grouped with the major mission (e.g., UNOMSIL was a precursor to UNAMSIL in Sierra Leone).
Still, even if we do not count small precursor and follow-on missions, the comparison is thirty-two
PKOs after 1987 to thirteen before.
15. These are UNIIMOG after the 1980–88 Iran-Iraq war, UNASOG in 1994 policing the disputed
Aouzou Strip between Chad and Libya, UNIKOM on the Iraq-Kuwait border after the 1991 Gulf
War, and UNMEE after the 1999 Ethiopia-Eritrea war.
16. UN transitional administration cases include UNTAG (Namibia, 1988), UNTAC (Cambodia,
1992), UNTAES (Eastern Slavonia in Croatia, 1996), UNMIK (Kosovo in Serbia and Montenegro,
1999), and UNTAET (East Timor, 1999). For a comparison of the pre- and post-1989 PKO tasks that
shows the remarkable expansion after the Cold War, see William J. Durch, UN Peace Operations and
the “Brahimi Report” (Washington, D.C.: Henry L. Stimson Center, October 2001).
and administration. Freetown, the capital of Sierra Leone, is swamped with

UN peacekeepers and other foreign aid personnel, although the backbone of
the regime is composed of the British troops who intervened successfully in
1999 on the side of the government against the Revolutionary United Front.
Such cases, and most of all recent U.S. actions in Afghanistan and Iraq, have
fueled considerable discussion about “American imperialism” or “neo-
imperialism” in the media. For example, several journalists and historians see
the United States playing an imperial role in extending U.S. power while polic-
ing and governing chaotic regions, parallel to the nineteenth-century British
Empire.17
There is indeed a valuable analogy between contemporary developments
and nineteenth- and early twentieth-century classical imperialism. As with
classical imperialism, we increasingly see the strongest states taking over, in
part or whole, the governance of territories where Western-style politics, eco-
nomics, and administration are underdeveloped.18 Further, excepting Iraq,
their actions generally have had international legal authority behind them, in
parallel with the international legalization of the former German and Italian
colonies (as League of Nation mandates) after World War I.
There are at least two striking differences, however. First, today’s rule by
and with foreigners is largely multilateral, whereas in classical imperialism the
great powers jealously monopolized control of their imperial domains. Second,
whereas nineteenth-century imperial ventures were conceived as indeªnite in
duration, postmodern imperialists want to rebuild self-supporting but politi-
cally and economically acceptable state structures and then leave as quickly as
feasible.19
17. In fact, they applaud this role and urge the United States to act as traditional imperialists
would. See, for examples, Mallaby, “The Reluctant Imperialist”; Michael Ignatieff, “Nation-
Building Lite,” New York Times Magazine, July 28, 2002; Max Boot, The Savage Wars of Peace: Small
Wars and the Rise of American Power (New York: Basic Books, 2002); Irving Kristol, “The Emerging
American Imperium,” Wall Street Journal, August 18, 1997, p. A14; and Niall Ferguson “No Way
to Run an Empire” New York Times Magazine, April 27, 2003 (who criticizes U.S. planners for their
desire for a quick exit).
18. Roland Paris, “International Peacebuilding and the ‘Mission Civilisatrice,’” Review of Interna-
tional Studies, Vol. 28 (2002), pp. 637–656, stresses the continuity of current “peacebuilding” efforts
with colonial projects.
19. Imperialism is a contested and fuzzy concept. We take it to entail the extension of formalized
rule by one polity (or polities) over another, where there is no anticipation that the ruled will ac-
quire full citizenship rights in the ruling polity. If, by contrast, “imperialism” requires that the im-
perial power wants to remain sovereign indeªnitely, then what we are calling postmodern
imperialism is not imperialism at all. We stress that we are not advocating or endorsing imperial-
ism with the connotation of exploitation and permanent rule by foreigners; postmodern imperial-
ism may have exploitative aspects, but these are to be condemned.
These differences stem from changed motivations for international interven-

tion. Nineteenth-century empire building was driven by notions of national
glory, hopes for proªt, civilizing missions, and perhaps most fundamentally
the fear that if other great powers acquired too much imperial territory, they
would gain enough strength to shift the balance of power.20 Today, by contrast,
nuclear weapons make territorial conquest among the great powers nearly un-
thinkable, thus removing mutual fears of military invasion. At the same time, a
successful world trading system has lowered incentives for exclusive imperial
control to assure access to vital raw materials.21 The major powers perceive lit-
tle serious military danger from one another, and so have lost the incentive for
imperialism as a means of defense or attack against other major powers. Long-
term, formal imperial annexation thus has much less appeal than it once did.
Increasingly, however, the major powers must worry about bad “externali-
ties” that result from the combination of the scientiªc revolution and political
disorder, economic collapse, and anger in the third world. These externalities
include risks of catastrophic terrorism using WMD, refugee ºows, health
threats, enhanced drug smuggling networks, and disruption of oil supplies.
Major powers can also suffer from destabilizing consequences of protracted
civil wars for whole regions, as neighboring states are weakened or regional
incentives for weapons acquisition and proliferation increase. Finally, the ma-
jor powers have faced signiªcant and justiªed pressures for intervention on
humanitarian grounds as well.
These diffuse threats create a classic collective action problem for the major
powers. Given the dangers posed by collapsed states and rogue regimes in a
world with WMD, open economies, and easy international travel, all would
beneªt from political order and responsible (if possible, democratic) govern-
ments in the periphery. But the costs to provide effective support for political
order and democracy after a state collapses often exceed the expected beneªts
for any one power. The logic of this situation creates an incentive for burden
sharing. Affected states have an incentive to share the costs to mitigate the
20. See, for example, Benjamin Cohen, The Question of Imperialism (New York: Basic Books, 1973).
21. Analysts differ on the causal priority to be assigned to the nuclear revolution, free trade, de-
mocracy, and changing social norms. See, for example, Richard N. Rosecrance, The Rise of the
Trading State: Commerce and Conquest in the Modern World (New York: Basic Books, 1987); John
Mueller, Retreat from Doomsday: The Obsolescence of Major War (New York: Basic Books, 1989); and
Bruce Russett and John Oneal, Triangulating Peace: Democracy, Interdependence, and International Or-
ganizations (New York: W.W. Norton, 2001). For a contrary argument that nothing has really
changed (the great powers are and should remain focused on great power military competition),
see John J. Mearsheimer, The Tragedy of Great Power Politics (New York: W.W. Norton, 2001).
public bads of state collapse and rogue regimes at an acceptable price for each
one.22
In sum, in the time of classical imperialism the great powers threatened each
other while facing no serious, autonomous threats from Africa, the Middle
East, or most of Asia. Today, by contrast, the strongest and richest states face
no serious military threats from one another, but various security threats from
a third world that, as Figure 1 illustrates, is suffering from a great deal of politi-
cal violence and chaos. Because the costs of addressing the implications of state
collapse or the dangers posed by rogue regimes are concentrated while the
beneªts are often diffuse, the major powers now confront a collective action
problem whose internal logic should favor a multilateral response.
The UN PKOs and the Brahimi Report
Collective action is no simple matter, however, especially absent a suprana-

tional authority and given conºicting state preferences and great uncertainties
about costs and beneªts. Even if each NATO member agrees in principle that
the International Security Assistance Force should be expanded to take on
peacekeeping duties beyond Kabul, each may prefer that other NATO coun-
tries provide the troops. Members of the UN Security Council might agree that
there would be collective beneªts to a large and adequately manned peace-
keeping mission for Bosnia, Rwanda, or the Democratic Republic of the Congo,
but hesitate to send troops or adequate resources themselves. The willingness
of a Security Council power to pay costs for a mission that mainly beneªts an-
other state (e.g., the 1994 mission to Haiti that the United States strongly de-
sired) may depend on a reciprocal willingness to support a mission that the
ªrst power strongly desires (e.g., the 1994 agreement on UNOMIG, the UN
mission to Georgia concerning the Abkhaz conºict).23 And, of course, the
major powers may strongly disagree about the scale of the threat posed by a
22. The structure of the collective action problem can be represented as follows: Let b ⬎ 0 be the
expected long-run cost of a state collapse to each major power, and let c ⬎ b be the total expected
cost of remedying it. If there are n states with these preferences and if c ⬍ nb, they would all be
better off to strike a deal in which they jointly contribute a total of c to the reconstruction effort,
with each one contributing no more than b. Of course, there is a possibly intense bargaining prob-
lem over who will contribute how much. See James D. Fearon, “Bargaining, Enforcement, and In-
ternational Cooperation,” International Organization, Vol. 52, No. 2 (Spring 1998), pp. 269–306.
23. David Malone, Decision-Making in the UN Security Council: The Case of Haiti, 1990–97 (Oxford:
Oxford University Press, 1998), suggests that this trade facilitated a Security Council agreement on
both missions.
rogue regime or a collapsed state, and the appropriate response—witness Iraq

in 2003.
Despite such obstacles, the UN has authorized and mobilized an impressive
number of collective action missions in the last ªfteen years on the interna-
tional problem of civil-war-torn states.24 The natural locus for negotiating and
organizing this collective action has been the UN Security Council. The UN
Charter gives the Security Council a legal responsibility for the maintenance of
international peace and security, and it serves as the closest approximation to a
club of major powers jointly affected by the bad externalities of state collapse.
Although peacekeeping is not speciªcally mentioned in the charter, it has
evolved into one of the UN’s preeminent tasks, as illustrated by Figure 1.
The ºurry of UN PKOs between 1989 and 2000 included a number of
underreported success stories but also some highly visible human disasters. At
the start of the period, with the thaw of the Cold War facilitating agreements in
the Security Council, newly ambitious missions were authorized for address-
ing internal conºicts in Namibia (UNTAG, 1989), Nicaragua (ONUCA, 1989),
El Salvador (ONUSAL, 1991), Angola (UNAVEM II, 1991), and Cambodia
(UNAMIC, then UNTAC, 1991). Except for UNAVEM II, these were largely
viewed as successful, valuable missions.25 Debacles and tragedies soon fol-
lowed, however. The policies of UNPROFOR, the UN peacekeeping force sent
to the former Yugoslavia in 1993, may have furthered the conºict and ulti-
mately enabled the Srebrenica massacres in the summer of 1995. The expanded
mandate given to UNOSOM II for Somalia in 1993 was quickly shown to be
unsustainable and went unfulªlled with horriªc consequences.26 Worst of all,
24. Collective action has been less successful in confrontations with rogue regimes. These present
a more difªcult problem for the UN system, based as it is on the mutual recognition of members’
sovereignty. The NATO attack on Serbia in 1999 is one example, however, of an intervention
against a rogue regime that subsequently received a UN-sanctioned PKO in Kosovo. The UN-
sanctioned, U.S.-led multilateral force that deposed the military coup leaders of Haiti in 1994 is
another.
25. Evaluating the performance of a UN mission is a conceptually difªcult problem; for an excel-
lent discussion, see Downs and Stedman, “Evaluation Issues in Peace Implementation,” in
Stedman, Rothchild, and Cousens, Ending Civil Wars. Downs and Stedman code all these missions
as “successful” except for UNAVEM II, which they code as a failure. Other UN PKOs focused on
post–civil war state reconstruction issues that were regarded by our interviewees and by much of
the literature on PKOs as underreported success stories include ONUMOZ (Mozambique, 1992–
95) and UNTAES (Croatia and Eastern Slavonia, 1996–98). For a statistical analysis that ªnds a
positive effect of UN PKOs on democratization and postwar stability, see Michael W. Doyle and
Nicholas Sambanis, “International Peacebuilding: A Theoretical and Quantitative Analysis,” Amer-
ican Political Science Review, Vol. 94 (December 2000), pp. 779–801.
26. Under UNOSOM II, U.S. Special Operations forces led a mission in April 1993 to capture
Somali warlord, Mohammed Farah Aideed, who was responsible for the massacre of twenty-four
the Security Council and the UN Secretariat reigned in rather than expanded
UNAMIR (1993–96) in the face of the 1994 Rwandan genocide.27
The Clinton administration largely disengaged from UN peacekeeping
activity after the Somalia ªasco, and a protracted budget crisis induced by the
U.S. Congress’s refusal to pay UN dues hamstrung an already overtaxed
ªnancing system for PKOs. In the early and mid-1990s, the UN Department of
Peacekeeping Operations (DPKO) was trying to manage and service tens of
thousands of peacekeeping forces in the ªeld with only dozens of staff at its
headquarters in New York. Operational and chain-of-command problems were
abundant, including slow and inadequate initial deployments, PKO troop con-
tingents following orders from their home state capitals rather than from the
UN, and ªeld commanders complaining in one mission (UNOSOM II) that
they could not get in trouble on a weekend because there was no one in New
York to answer the phone when they called for assistance.28
In the late 1990s, the Clinton administration returned somewhat to the sup-
port and use of UN PKOs, favoring Security Council authorization of four
complex civil war–related missions: Kosovo after the U.S. and NATO attacks
on Serbian leader Slobodan Miloševib’s regime (UNMIK, June 1999), Sierra Le-
one (UNAMSIL, October 1999), Democratic Republic of the Congo (MONUC,
November 1999), and East Timor (UNTAET, October 2000). Except for
UNTAET, which has been successful so far, these missions have displayed
many of the same problems that beset earlier missions.29 In particular, the ini-
Pakistani UN peacekeepers. In that mission, eighteen U.S. soldiers were killed and eighty-four
were injured. Hundreds of Somalis also died. In its aftermath, President Clinton shut down the
Somali operation, and southern Somalia fell back into violent anarchy.
27. The problems and failures of UNPROFOR, the UN PKO operating for most of the Bosnian war,
are well analyzed in the UN report The Fall of Srebrenica, report of the Secretary-General pursuant
to General Assembly resolution 53/35, November 15, 1999, A/54/549. See also David Rohde, End-
game: The Betrayal and Fall of Srebrenica (New York: Farrar, Straus and Giroux, 1997). On Rwanda,
see Report of the Independent Inquiry into the Actions of the United Nations during the 1994 Genocide in
Rwanda, December 15, 1999, New York; United Nations, Department of Peacekeeping Operations,
Comprehensive Report on Lessons Learned from United Nations Assistance Mission for Rwanda
(UNAMIR), 1996; and Howard Adelman and Astri Suhrke, with Bruce Jones, The International Re-
sponse to Conºict and Genocide: Lessons from the Rwanda Experience (Copenhagen: Joint Evaluation of
Emergency Assistance to Rwanda, 1996). On Somalia, see Adelman and Suhrke, The International
Response to Conºict and Genocide; and Kenneth Allard, Somalia Operations: Lessons Learned (Washing-
ton, D.C.: National Defense University Press, 1995).
28. Remarks (now outdated) of Maj. Gen. Lewis Mackenzie, a Canadian and former head of UN
forces in Sarajevo, quoted in Ramesh Thakur, “From Peacekeeping to Peace Enforcement: The U.N.
Operation in Somalia,” Journal of Modern African Studies, Vol. 32, No. 3 (1994), p. 393.
29. UNMIK has been successful in some regards, and its design shows evidence of learning based
on coordination problems observed in post–Dayton agreement Bosnia. See Caplan, A New
Trusteeship?
tial deployment of UNAMSIL was disastrously inadequate, requiring rescue

by British troops and expansion in size. And despite the ofªcial rhetoric that
mass killings would never again be permitted in the wake of Rwanda and
Srebrenica, once again the blue-helmeted UN multinational soldiers had little
choice but to stand by watching as ethnic massacres took place in the eastern
Congo in early 2003.30
The Brahimi report, commissioned by the secretary-general in 2000, ana-
lyzed the sources of failure in past UN PKOs and proposed a broad-ranging set
of reforms. The report identiªes three core problems with UN PKOs in the
1990s. First, the Security Council often approved mission mandates that were
overly vague and optimistic about the situation on the ground, leading to in-
adequate deployments with unclear objectives. Second, the mission mandates
constructed within the Security Council frequently exceeded the resources pro-
vided by member states. There were “commitment gaps,” as the report puts
it.31 Together, these two problems produced gaps between what outside ob-
servers could reasonably expect a UN PKO to do, and what it was capable of
doing.
Although the report is not explicit about the roots of these failings, from its
text and our interviews we infer two political pathologies involving the ªve
permanent members of the Security Council (the P-5) and the UN that yielded
the gap between mission expectations and the resources provided.
The ªrst pathology ran from the P-5 to the secretary-general. In the 1990s,
Western governments sometimes wanted to appear to be taking action in hu-
manitarian crises over which they did not want to spill blood. Such tasks were
passed to the UN while the P-5 knowingly underfunded the missions relative
to their mandates. This allowed them to appear to be doing something, and if
there were failure, blame would later fall on the UN for its supposed
inefªciencies. Perhaps as a matter of default rather than design, the UN was
thus used as a scapegoat.
The second pathology ran in the other direction. Here the report intimates
that the secretary-general had incentives to undersell mission requirements to
the members of the Security Council because he wanted (or faced strong pres-
sures to get) a mission to country X and knew that the Security Council would
30. The massacres led to “mission creep,” a phenomenon we address in the next section. The ªrst-
ever EU-led multinational force was deployed in June as an interim measure, followed by an ex-
pansion of the mandate and size of MONUC (authorized by Security Council resolution S/RES/
1493 [2003]).
31. United Nations, Report of the Panel on Peace Operations, pars. 48–64.
not endorse a mission that was above some size threshold. This pathology was
based on a “foot in the door” logic, driven by the hope that a mission would be
expanded when it became obvious to all that it was failing.32 Both of these pa-
thologies caused Security Council mandates to exceed the resources provided
to PKOs in the 1990s—especially in Somalia, Bosnia, and Rwanda—and so
contributed to peacekeeping failures.
The Brahimi report outlines remedies to overcome the gap between mission
objectives and resources provided. It exhorts the Security Council to write
clear and achievable mandates that it is willing to back up and fund, and it rec-
ommends that mandates be left in draft form until the secretary-general
conªrms that the force levels they call for can actually be supplied.33 It exhorts
the secretary-general to tell the Security Council “what it needs to know, not
what it wants to hear” when formulating mandates.34 That is, he should not
make a mission seem easy if the result is likely to be a hypocritical mandate
from the Security Council.
We should ask why the P-5 need to be told that they should be willing to ad-
equately support the PKOs they endorse, and why the secretary-general needs
to be told to be straightforward with the Security Council. Surely the actors al-
ready have a subtle understanding of these political dynamics. The P-5 knew
they were underfunding and understafªng missions in Bosnia, Somalia, and
Sierra Leone, yet still ordered the secretary-general to organize and pursue
them.35
One possible explanation is that the Brahimi report provides the secretary-
general with a political instrument in his negotiations with the Security
Council. The secretary-general can now remind the P-5 that they enthusiasti-
cally endorsed Brahimi principles in Security Council resolution 1327 (2000),
so if the Secretariat says that it will take X troops and Y dollars to achieve
Z mandate goals, this is what the council must authorize. If the council is un-
willing to do so, it should deªne a less ambitious mandate or not undertake
32. This second pathology could involve members of the Security Council, who, like the secretary-
general, might want a more aggressive UN reaction and see a strong mandate as a means toward
this end, while council members who would prefer no involvement compromise on a strong man-
date with inadequate resources.
33. United Nations, Report of the Panel on Peace Operations, par. 60. This recommendation was not
accepted.
34. Ibid., p. 10. And similarly, “the Secretariat must not apply best-case planning assumptions to
situations where the local actors have historically exhibited worst-case behaviour.” Ibid., p. x.
35. This is particularly evident in United Nations, The Fall of Srebrenica.
the mission. The authors of the Brahimi report may have been giving him po-
litical cover for the rash act of telling the truth to the Security Council.36
One of main truths that needs to be told, the report suggests, is that Security
Council mandates must provide for robust rules of engagement by UN
forces—that is, the ability and permission to ªght back, deter or pursue spoil-
ers, and stop violence against civilians.37 If the Security Council is unwilling to
provide this up front for missions involving internal conºict, the report argues,
the PKO should not be approved.
The third set of problems the Brahimi report identiªed was marked in-
efªciencies and lack of coordination in the anticipation, planning, deployment,
and management of UN PKOs. The report recommended the creation of a new
intelligence and planning unit within the Secretariat; a set of reforms to make
rapid deployment feasible; more posts and funding for the DPKO; and internal
restructuring to improve information sharing, planning, and coordination in
support of missions. Although the General Assembly rejected the new intelli-
gence and planning unit, it has thus far approved some increased funding and
posts for the DPKO, as well as aspects of the rapid deployment and internal
structural reforms.38 Insiders believe that this is the backbone of the report and
the part that will have the greatest long-term impact.39 In the demands for new
budget items, an implicit deal was proposed. The UN Secretariat asks the Secu-
rity Council (and the General Assembly) for more resources to fund PKOs, and
the Secretariat then promises to use those funds more efªciently and effec-
tively, and avoid cases that are likely destined for failure.40
36. Interview UN3, April 3, 2001.

37. Stephen Stedman deªnes spoilers as “leaders and parties who believe that peace emerging
from negotiations threatens their power, worldview, or interests, and use violence to undermine
attempts to achieve it.” Stedman, “Spoiler Problems in Peace Processes,” p. 5.
38. Members of the Group of 77, which dominates the General Assembly, often see peacekeeping
operations as a Security Council (and hence largely rich-state) activity that threatens to take funds
from economic development and can threaten sovereignty as well. In addition, as principal troop-
contributing countries, they resent perceived lack of input into PKO mandate formulation and
mission planning. Regarding the proposed intelligence and planning unit, they were unenthusias-
tic about increasing the UN’s capability to monitor developments in member states that might oc-
casion international intervention (UN1, April 2, 2001; UN9, April 5, 2001). For an excellent analysis
of the prospects and problems with implementing the Brahimi report’s recommendations, see
Christine Gray, “Peacekeeping after the Brahimi Report: Is There a Crisis of Credibility for the UN?”
Journal of Conºict and Security Law, Vol. 6, No. 2 (December 2001), pp. 267–288.
39. Interviews UN3 and UN4, April 3, 2001.
40. Thus Brahimi insisted that the UN not be given the awesome job of peacekeeping in post-
Taliban Afghanistan, although he did take responsibility for heading a scaled-down UN operation
(UNAMA). Staff in the UN Department of Peacekeeping Operations were similarly unenthusiastic
Insurgency and the Logic of “Mission Creep”
The gap between mission needs and the resources supplied by international
interveners in collapsed states has another important source not clearly
identiªed in the Brahimi report. Once international forces are on the ground in
a country with a weak central state and local conditions that favor guerrilla
techniques, there can be a powerful internal dynamic for “mission creep,” or
escalation of mission goals and requirements. This dynamic grows from the
nature of the political and military problem that international interveners face
in a collapsed state, and it can operate whether the interveners are UN blue
helmets or the U.S. military. Neither UN nor U.S. political and military doc-
trine adequately addresses this problem.
The UN’s basic doctrinal orientation in peacekeeping operations is as a
mediator and monitor that facilitates the resolution of political differences di-
viding the parties. For civil war interventions, the DPKO’s ideal case is one in
which two sides have seriously committed themselves to a detailed peace
agreement (worked out with the help of UN mediators), and then jointly re-
quest a PKO for the purpose of monitoring and assisting with implementation.
As Jean Marie Guéhenno, under secretary-general for peacekeeping operations
from October 2000, put it, “Consent [is] still the principle under which the De-
partment operate[s], otherwise, it [is] simply not peacekeeping and the United
Nations [is] not the right organization to do it.”41
However, the UN and third parties can face strong incentives to intervene in
a civil-war-torn state even when there is no peace agreement, or when there is
a ºimsy peace agreement likely to be challenged by a return to violence by a
spoiler. Moreover, even with a relatively sincere peace settlement, if the state
apparatus of the country has largely collapsed as a result of years of mis-
governance, war, or foreign intervention, then international interveners will
about the prospect of a peacekeeping mission to postwar Iraq, in part because they see it as too
hard a case. See Michèle Grifªn, “Mediator, Midwife, or Manager? The United Nations and State-
Building,” paper presented at the Center for International Security and Cooperation workshop
“How to Build a State,” Stanford University, May 23–24, 2003.
41. This is the paraphrase of Guehénno’s remark reported in United Nations press release GA/
SPD/265, “Reform of UN Peacekeeping Operations a Real Process with Real Beneªts, Under Sec-
retary-General Tells Fourth Committee,” October 15, 2003. Similarly, as the primer on peacekeep-
ing at the DPKO website emphasizes, there can be no success until there is “a genuine desire on
the part of the warring parties to solve their differences peacefully.” An Introduction to United Na-
tions Peacekeeping, http://www.un.org/Depts/dpko/dpko/intro/3.htm, chap. 3.
not be able to leave without substantially rebuilding state security institutions

and providing security in the interim. Otherwise, local anarchy will encourage
guerrilla attacks, local warlords, and competitive attempts to capture the
(largely notional) state.
To its credit, the Brahimi report recognizes that “consent may be manipu-
lated in various ways,” that it may be withdrawn, and that factional splits may
yield new combatants who reject the original peace agreement.42 The remedy it
proposes is robust rules of engagement and initial deployments adequate to
face down any spoiler who might emerge.
In weak state conditions, however, this remedy at best lowers the odds of
humiliating international withdrawals in the face of attacks by warlords. When
past conºict or other factors have rendered the state apparatus too dysfunc-
tional to provide for domestic security, mission creep is highly likely. Addi-
tional forces may be required to deal with depredations by thugs and
guerrillas, and it will become increasingly clear that exit without a return to
war demands some level of sustained transitional administration by interna-
tional parties. The reality of state weakness means that peacekeepers need to
foster state building if there is to be any hope for exit without a return to con-
siderable violence. This Hobbesian logic applies whether the forces are UN
troops, ad hoc international coalitions, or the U.S. military in Afghanistan or
Iraq.
The argument is consistent with and supported by recent research on the
causes of civil war in the post–World War II period. In our study of 161 coun-
tries since 1945, the best predictors of a country’s civil war propensity are the
presence of factors that favor a particular form of military conºict—namely,
rural guerrilla warfare.43 Guerrilla warfare is overwhelmingly the most com-
mon form of violent civil conºict in this period, and it is a highly robust tech-
nology in conditions prevalent in postcolonial Africa, Asia, and the Middle
East. The steady increase in the number of ongoing civil wars shown in
Figure 1 is not due to more wars breaking out each year. Rather, while the rate
of outbreak has been fairly steady, the rate at which rural guerrilla wars in the
third world have ended has been lower. This has produced a gradual accumu-
lation of unresolved conºicts and increasingly incapacitated states. Across
countries, measures of low state capability relative to potential guerrilla
bands—such as low per capita income, high total population, recent decoloni-
42. United Nations, Report of the Panel on Peace Operations, p. 9.

43. Fearon and Laitin, “Ethnicity, Insurgency, and Civil War.”
zation, and mountainous terrain—are the best predictors of a state’s civil war
propensity.44
The implicit theory undergirding UN peacekeeping operations is that peace
depends primarily on the resolution of political differences and the alleviation
of political or social grievances. Surprisingly, then, after one controls for the
level of economic development, measures of social cleavage and the extent of
political grievances appear unrelated to a country’s risk of civil war outbreak.
For example, more ethnically or religiously diverse countries, or countries with
an ethnic majority and a signiªcant ethnic minority, were no more likely to
have a civil war start in this period, controlling for per capita income. Simi-
larly, the risk of civil war outbreak does not vary systematically with a coun-
try’s level of democracy, income inequality, or the presence of policies
discriminating against minority languages or religions. In a study of some 280
ethnic and religious groups over the same period, we ªnd that economically
disadvantaged minorities were no more likely to be involved in rebellion than
economically equal or better off minorities.45
It is possible that we ªnd no evidence of a relationship between civil war
and discrimination or broad social grievances because our measures are poor.
However, the absence of a clear pattern across a large number of indicators and
in several different data sets is striking.46 We suspect that the correct interpreta-
tion is not that social and political grievances are irrelevant to motivating
ªghters. Rather, grievances are lamentably common, so that knowing their
level does not help distinguish the countries that suffer from civil wars from
those that do not. Instead, conditions that favor rural insurgency mark the
states at highest risk for civil war in this period. Protracted civil war has been
on the rise ever since the transfer of independence to a swath of colonies
whose new leaders were unable to create the order and progress necessary to
44. Our results are not idiosyncratic. Per capita income and total population tend to be the most
powerful predictors in a number of similar cross-national studies, such as Paul Collier and Anke
Hoefºer, “Greed and Grievance in Civil War,” World Bank, 2001. The precise mechanism by which
low per capita income favors civil war is unclear. Fearon and Laitin, “Ethnicity, Insurgency, and
Civil War,” argue that it is mainly measuring state capability to ªnance and run a competent coun-
terinsurgency, whereas Collier and Hoefºer tend to see it as a measure of the opportunity cost of
becoming a rebel. Either way, it is a measure of state capability relative to potential guerrilla forces.
45. Fearon and Laitin, “Ethnicity, Insurgency, and Civil War”; James D. Fearon and David D.
Laitin, “Weak States, Rough Terrain, and Large-Scale Ethnic Violence since 1945,” paper presented
at the annual meeting of the American Political Science Association, September 2–5, 1999, Atlanta,
Georgia.
46. Broadly similar ªndings are reported in Collier and Hoefºer, “Greed and Grievance in Civil
War.”
dissuade insurgents from seeking the rewards of control over the reigns of
government either at the center or in a region.
These ªndings do not imply that civil conºicts can be peacefully resolved
without addressing the political grievances that the parties say are at issue. It is
hard to imagine, for instance, how to end the Palestinian insurgency in Israeli-
controlled territories without addressing at least some Palestinian grievances.
But the ªndings do suggest that a theory of international intervention in ser-
vice of peace in civil-war-torn states is misconceived if it focuses solely on me-
diating political differences or on providing space for and assistance with such
peace processes. Where state collapse and environmental conditions favor in-
surgency, counterinsurgency and state building will often be necessary even to
create the conditions needed to address grievances or to make political or eco-
nomic progress of any kind.
These tasks involve a different action plan than the classical notion of peace-
keeping. State building, for example, may require a long-term commitment to
rebuilding basic institutions for security and political stability. For its part,
counterinsurgency implies that the UN ethic of “impartiality,” and its tradi-
tional chapter 6 role of placing blue helmets between parties in need of assur-
ance that the other side will exercise restraint, is not directly applicable to
many post–World War II civil wars. Counterinsurgency often requires tactics
of intimidation and threat. As a former Pentagon ofªcial told us, teams sec-
onded from the Royal Ulster Constabulary (RUC) were probably the most ef-
fective police unit in Kosovo, though their methods may not always have been
fully in accord with human rights conventions.47 The reality of insurgency and
immediate postwar disorder suggests that a dominant military force will often
be essential to lead an effective PKO.48
These considerations do not apply to all cases of postconºict international
intervention. They are more relevant the lower the bureaucratic and political
capabilities of the state at issue. UN missions and those of other interveners
have tended to “creep” or be unable to exit where state administrative, police,
47. Interviews DC1, March 18, 2002; and DC4, March 20, 2002.
48. The UN system is inherently biased in favor of member states against rebel challengers, so that
legal authorization for intervention to support insurgents against a bad (but recognized) govern-
ment is bound to be exceptional. In some civil-war-torn states, however, the insurgents may offer
better prospects for state reconstruction, good government, and political justice than does the re-
gime that controls the capital. Although the anarchical conditions that favor insurgency pose un-
solved problems for PKOs, the best solution is not always to support a recognized government
against the insurgents. Still, developing workable standards for the international authorization to
support rebel groups (as in Kosovo, for example) remains a vexing problem.
and military capacities have been lowest, such as in Afghanistan, the Congo
(both in 1960 and today), Sierra Leone, Liberia, Somalia, Haiti, Bosnia
(UNPROFOR), and Kosovo. By contrast, there has been little or no mission
creep in cases where the state retained governing capabilities despite the war,
such as in Nicaragua, El Salvador, Guatemala, Croatia (UNTAES), and Cambo-
dia (UNTAC). In the latter set, the underlying problem more closely resembled
what received UN PKO doctrine imagines, and accordingly their PKOs were
more successful.49
Four Strategic Challenges for Neotrusteeship
Given the costs and difªculty of counterinsurgency operations and state build-
ing in collapsed states, the U.S. government’s and UN DPKO’s lack of enthusi-
asm for such enterprises is understandable.50 But this is no excuse for failing to
develop doctrine and mechanisms for collective action to address the problem
of how to intervene constructively in postconºict weak or collapsed states.51 In
this section we consider four major problems for the system as it stands and
suggest some possible solutions.
recruitment
In a world with WMD and relatively open borders, political order, democracy,
and competent government are international public goods. Who should pay to
provide these goods in collapsed states? How best to organize interventions to
make the costs manageable and equitably distributed?
Two general principles are commonly employed to answer the ethical ques-
49. In another set of cases, there were pressures for mission creep that, for various reasons, the
interveners resisted: Angola (UNAVEM II and III), Rwanda (UNAMIR), and Ivory Coast
(MINUCI, which could yet “creep”). UNTAET in East Timor is thus far an anomalous case where
prospects appear good for exit by the interveners, despite there having been little preexisting state
capacity. It appears to have been favored by East Timor’s extremely small size and the relatively
uniªed independence movement.
50. Grifªn, “Mediator, Midwife, or Manager?” discusses the deep unease and unhappiness of
DPKO staff, despite successes in Kosovo and East Timor, with transitional administration missions
that require the UN to take on governmental functions in postconºict situations.
51. More controversially, doctrine is needed for state reconstruction efforts in states whose re-
gimes were deposed as alleged threats to international peace and security. Whereas the UN may be
criticized for having a doctrine that does not adequately conceptualize or address the problem
posed by peacekeeping in a weak-state setting, the Bush administration may be criticized for not
seeing the problem at all. Administration ofªcials seem to have imagined that postconºict state
building in Iraq would largely take care of itself. See, for example, David Rieff, “Blueprint for a
Mess,” New York Times Magazine, November 2, 2003, pp. 28–33, 44, 78.
tion of who should pay for a public good. First, those most capable of paying
should pay the most. Second, those who beneªt most should pay the most.
The ªrst principle implies that the rich states of the West should bear most of
the ªnancial costs of peacekeeping operations. Regarding funding for UN
PKOs, this is already the case. As the Brahimi report suggests, however, the
major powers’ “lack of political will” in ªnancing PKOs has been a prime
source of the commitment gap. Furthermore, rich states have shown great re-
luctance to send their troops on dangerous PKO missions. In the violent mo-
ments of a civil war or peace process, the great powers have not put their own
soldiers at serious risk unless the conºict posed a direct threat within their
own region, or unless the failed state was identiªed as a potential harborer of
terrorists or concealer of WMD.
Along with many of our interviewees at the UN and in Washington, we
deplore the P-5’s (and particularly the U.S. government’s) reluctance to pro-
vide more funding and troops for PKOs in postconºict states. But to the extent
that this reºects a political judgment by Western politicians about how much
stability in collapsed states is worth to their voters, it is not a failure of collec-
tive action. To this extent, it is simply a constraint on who will pay, and how
much, that must be taken into account when organizing multilateral interven-
tions for collapsed states.
By necessity, the constraint is already reºected in evolving practices for
recruiting and structuring PKOs. The Brahimi report notes that the current era
of “complex” peacekeeping in civil-war-torn states has seen a dramatic shift
away from developed-country blue helmets—the standard practice during the
Cold War—toward blue helmets supplied by third world states.52 This is the
result of a natural trade between the rich, major powers and many developing
countries. The former have ªnancial resources but are highly sensitive to casu-
alties and often reluctant to put their forces under UN or international com-
mand. The latter may have large armies, weak tax revenues, or both. These
troop-contributing countries (TCCs) send forces for reasons of prestige or ex-
pected recompense, whether monetary or via special political deals with major
powers particularly interested in getting a mission to a particular country.53
Even though such trades make sense and should be encouraged for facilitat-
52. United Nations, Report of the Panel on Peace Operations, p. 17.

53. Two of our interviewees intimated that Pakistan had received a promise of sanctions removal
from the United States for an agreement to send PKO forces to Sierra Leone in 2000 (UN1 April 2,
2001; UN9 April 5, 2001), though neither was certain of this.
ing the supply of PKOs, they pose tricky political and operational problems.
First, members of the Group of 77, the principal suppliers of PKO troops in
recent years, are increasingly bitter about their lack of voice in the Security
Council deliberations that formulate and manage the PKO missions that em-
ploy their troops. Without reform of the Security Council, several ambassadors
emphasized, their countries’ willingness to supply troops will wane.54 Second,
the UN is continually in arrears to TCCs. Although all UN member states are
taxed for peacekeeping operations, the budget is perennially in deªcit, and
many TCCs hold extensive accounts owed for their services in past PKOs. De-
spite the Brahimi report’s valiant and partially successful efforts to put the UN
DPKO on more solid ªnancial footing, funding for particular operations is
often ad hoc and mostly paid by donor countries in the developed world. The
secretary-general and even SRSGs must often act like college presidents,
spending their time urging the rich to fund worthy special projects.
On the operational side, we have argued that in a country with a fractured
state and an environment conducive to guerrilla war, a successful PKO
requires a dominant military force that can act decisively, establish a clear
chain of command, and take responsibility. Many third-world troop contin-
gents have served with distinction in UN PKOs; however, they often lack the
equipment, advanced technology, training, and doctrine necessary to operate
as a dominant military force in the chaotic conditions of a collapsed state.55
Thus, there needs to be a lead state or regional organization with advanced
technical and organizational capabilities to be the principal contractor with the
UN for PKOs sent to collapsed states. On the next level, efªcient supply of
PKOs requires TCCs providing what Pentagon ofªcials call “critical mass
forces”—battalions assigned for speciªc tasks under the overall direction of the
lead state or organization. Once the United Kingdom took over the lead in
Sierra Leone, a PKO that had fallen into an abyss was restored to effectiveness.
54. Interviews UN1, April 2, 2001; and UN9, April 5, 2001. The Brahimi report refers to the need
for better consultation between troop contributors and the Security Council (p. 11, par. 61), and this
issue garnered attention and some action in the post-Brahimi debate within the UN. See Gray,
“Peacekeeping after the Brahimi Report.”
55. As the Brahimi report commented (par. 108), “Some countries have provided soldiers without
riºes, or with riºes but no helmets, or with helmets but no ºak jackets, or with no organic trans-
port capability (trucks or troops carriers).” Through its predeployment visits and other programs,
DPKOs Force Generation Service has gone some way to upgrade the personnel sent to missions by
TCCs. See William J. Durch, Victoria K. Holt, Caroline R. Earle, and Moira K. Shanahan, The
Brahimi Report and the Future of UN Peace Operations (Washington, D.C.: Henry L. Stimson Center,
2003), pp. 77–78. There is a long way to go, however, before many of these armies can realistically
take the lead in a collapsed-state PKO.
In the post-Soviet region, when Russia or the Commonwealth of Independent

States (CIS) acted as a lead state (in Pridnestrovie, in Abkhazia, in South
Ossetia, and in Tajikistan), the civil wars were short, with relatively few deaths
and quick armistices. The reason, our informants argued, was that Russian
peacekeepers were able and willing to (in the words of several informants)
“liquidate” spoilers. They were able, as in Tajikistan, to pick a warlord favor-
able to them and provide him the military support necessary to compel other
pretenders into negotiations.56
Why does the lead actor need to be a major power or a regional organization
such as NATO or the CIS? Why not the UN? Or, in this age of privatization,
why not contract out to private corporations such as Military Professional
Resources Incorporated (MPRI) or DynCorp, which already supply training
and a host of logistical services in combat zones around the world?57 In the
next section we argue that the UN is not a good candidate for lead agent of in-
terventions in weak state conditions. Both past experience and structural fea-
tures of the organization suggest that the UN is inappropriate for leading the
kinds of military actions that are often necessary when anarchy reigns. Where
there is a greater level of state coherence and when local conditions are less
conducive to guerrilla war, UN-led PKOs make more sense.
“Private military corporations,” by contrast, should have at best a logistical
role to play. MPRI-like ªrms currently train troops and provide management
services on defense matters, law enforcement, and leadership development in
both the public and private sectors. Indeed, they do virtually everything
but engage in combat (“trigger pulling” in military jargon).58 This is a critical
failing, however. Given the nature of the problem faced by international
intervention in a collapsed state—endemic conditions ripe for insurgency—
trigger-pulling authority is essential. States are understandably reluctant to
56. Russian ofªcials conºate the concepts of “peacekeeping” and “peacemaking” and use the lat-
ter as the generic term. This violates a certain political correctness norm in the UN Department of
Peacekeeping Operations. Russian usage implies that PKOs by deªnition involve a proactive, not
necessarily impartial, strategy.
57. On private contract armies see P.W. Singer, “Corporate Warriors: The Rise of the Privatized
Military Industry and Its Ramiªcations for International Security,” International Security, Vol. 26,
No. 3 (Winter 2001/02), pp. 186–220; and Deborah Avant, “The Market for Force: Exploring the
Privatization of Military Services,” paper prepared for the Study Group on Defense Industry Glob-
alization, Conversion, and Arms Trade (New York: Council on Foreign Relations, 2003).
58. In a few cases, such as Executive Outcomes in Sierra Leone and Sandline International in
Papua New Guinea, private military corporations pulled triggers on behalf of an embattled gov-
ernment willing to pay for mercenary support. But these cases involved contracts with a state
rather than with the UN or other international organization.
contract this out. A license to kill requires public authorization and account-
ability. An organization without operational command cannot direct states that
have operational capability. Thus, the UN system needs to recruit for leader-
ship among its member states, or among UN chapter 7–recognized regional or-
ganizations, to lead such operations.59
How then to motivate major powers or NATO to take on such roles? If an
effective PKO requires a lead state, this implies concentrated costs for this
country, especially in terms of casualty risks.
We see two natural, if partial, solutions. The ªrst is to encourage and facili-
tate interventions led by the major powers or regional actors with the greatest
national security or economic interest in restoring stability and democracy to
the collapsed state. The second is to develop international institutions and
legal arrangements whereby the costs of third-party peacekeeping and state
building are gradually shifted to the state being rebuilt. Both solutions are con-
sistent with the second normative principle stated above: that those who
beneªt most from provision of a public good should pay the most. There are
two chief beneªciaries of restoring political order in a state destroyed by civil
war: the residents of the collapsed state, and neighboring or other states that
have a particular security, economic, or historical interest in the stability of the
country in question. Our proposal is to develop international arrangements
that would more systematically take advantage of this congruence of interests,
while monitoring and controlling the potential for abuse in each path.
Powerful states are most easily recruited as lead agents for peacekeeping/
state-building operations when the results are linked to their security interests,
or when past relations imply a “special responsibility.” Examples include
NATO and the OSCE in Bosnia and Kosovo; the United States in Haiti (both in
1994 and in 2004), Liberia, Afghanistan, and Iraq60; Australia in East Timor;
Nigeria and later the United Kingdom in Sierra Leone; France in several PKOs
in francophone West Africa; and Russia in several Central Asian cases. These
examples indicate that the practice of international contracting to interested
parties for PKO provision became widespread in the 1990s. This is not surpris-
ing given the sharp increase in dangerous missions to failed states without a
proportionate increase in funding through the UN.
This practice, however, violates prior understandings of UN PKO doctrine.
59. Chapter 8 refers to recognized regional organizations supporting peace in their region.
60. At least, the Bush administration saw a close link between intervention in Iraq and U.S.
national security interests.
Christine Gray refers to bitter complaints by the Indian permanent representa-

tive, who argued that during the Cold War, “the UN understood quite rightly
that successful peacekeeping could be carried out only by countries that were
neutral and had no interests of their own to pursue. India complained that this
cardinal rule had been broken repeatedly in recent years, with the UN now co-
opting regional players, though it is in the nature of politics that they are often
part of the problem, not of the solution.”61 This is a valid concern, but if the
choice is between no PKO and one led by interested major powers or regional
actors, a better course may be to develop international arrangements to moni-
tor and manage the agency problems as best possible. We elaborate below in
the section on accountability.
Much less considered or employed is the second way to link a strong interest
in state reconstruction with provision of services: the use of loans and direct
transfers to pay for peacekeeping, to be repaid as the state’s economy revives.
In effect, this means taxing the citizens of the collapsed state in the future to
pay for peacekeeping and reconstruction in the present. There are some prece-
dents. In 1993, Cyprus began to pay one-third of the UNFICYP mission costs.
Kuwait paid one-half the budget for its monitoring force (UNIKOM). The
1963–64 Yemen operation was funded by neighbors (Saudis and Egyptians),
who had an interest in order and took responsibility for paying for operations
beneªting their neighbor (UNYOM). Furthermore, there was de facto taxation
of the beneªciaries in the case in Sierra Leone, where the Nigerians imposed
a heavy informal tax burden through unsystematic looting, and East
Timor, where Australians are well positioned to beneªt from involvement in a
developing offshore petroleum industry.62
Thus, the idea of intertemporal taxation to foster recruitment of peace-
keepers is neither completely new nor a violation of an international consen-
sus. Below, we argue that if properly designed, such arrangements could also
help provide the incentives for local parties to contribute to building a state
that can stand on its own without continued third-party support. But obvi-
ously there are major agency problems and questions of legitimacy that inter-
national institutional innovation would need to address, such as who can
61. Gray, “Peacekeeping after the Brahimi Report,” p. x.

62. In particular, having Australians closely involved in the transitional administration of East
Timor increases the likelihood that Australia will receive a favorable ruling on the ownership of
offshore oil ªelds where sovereignty is ambiguous. For a discussion of this issue, see La’o Hamutuk
Bulletin, Vol. 4, Nos. 3–4 (August 2003), available on the web at http://www.etan.org/lh/
bulletins/bulletinv4n34.html#update.
agree to such a deal on behalf of citizens of the collapsed state, how to make
the arrangement sufªciently transparent and secure as to avoid neocolonial
robbery, and how to calibrate the loans so as not to undermine economic rede-
velopment. We brieºy address these issues in subsequent sections.
coordination
The various actors involved in developing neotrusteeships pose an immense
coordination problem. Besides a lead state and TCC troops, peacekeeping in
weak state conditions will include a collection of other organizations such as
other states supplying police and administrative personnel, NGOs with a vari-
ety of functional specializations,63 intergovernmental agencies (the OSCE has
played a special role in this regard), agencies of the UN (e.g., the UN Depart-
ment of Political Affairs to help run elections), and private contractors (e.g.,
DynCorp) providing logistical services. Even if rationalizing the practice of
having a lead state would go some way toward coordinating action on the
security side, there remains the problem of how to coordinate international
action on the many other aspects of state building.
We see three avenues by which overall mission coordination in neo-
trusteeship ventures can be improved: by encouraging and rationalizing a lead
state system for interventions in weak-state conditions; by learning from and
replicating successful on-the-ground organizational innovations; and by pro-
moting common international PKO troop standards through international
training programs.
Above we argued that when the host state is barely functional, an effective
PKO requires a lead actor that can serve as a dominant military force. This pro-
posal would address the coordination failures that beset several UN missions
in the 1990s, especially UNPROFOR. More coordination is possible when the
Security Council designates a lead agent in its mandate, as it did with the
United States in the Korean War and with NATO in post-Srebrenica Bosnia.
UN agencies, TCCs, and NGOs participating on an ad hoc basis get from a lead
state a better sense of how they ªt into the organizational picture. With a
clearer assignment of responsibilities, various organizations will develop spe-
63. In an interview (with Laitin, in Washington, D.C., November 10, 2003), a U.S. Marine Corps
colonel in charge of civil-military affairs in Bosnia reported that in 1999 there were 458 NGOs reg-
istered by his ofªce. For a more general discussion of the proliferation of NGOs and regional orga-
nizations involved in postconºict operations, see also Krasner, “Troubled Societies, Outlaw States,
and Gradations of Sovereignty”; and Bruce D. Jones, “The Challenges of Strategic Coordination,”
in Stedman, Rothchild, and Cousens, Ending Civil Wars.
cialties within PKOs and learn how to coordinate with other organizations pro-
viding related services (such as fostering links between border protection and
tariff collection). The Pentagon, for example, bristles when asked to provide ci-
vilian services such as border protection in Bosnia and seeks to keep its dis-
tance from the slippery slope of providing tariff collection services.64 With
clearer organizational leadership, there would be less fear that accepting any
one task entails obligations for a whole set of related functions.
In weak state conditions, the DPKO is a poor candidate for coordinating a
complex international intervention. The Brahimi report suggests that the UN
can act as the lead if given brigade-level capabilities.65 We remain skeptical, for
two reasons. First, UN-led mediation efforts require neutrality, but neutrality is
virtually impossible to maintain in a PKO in a largely anarchical setting. In the
short term, robust rules of action pursued by UN blue helmets will work to the
advantage or detriment of some local parties. In the long-term, impartiality
requires that the UN support the forces most willing to work toward peace.66
To the extent the UN plays the short-term role of lead military actor, it loses its
credibility as a neutral party that can subsequently mediate peace agreements.
The UN can act as principal (authorizing the lead state) and as a subcontractor
for various functions (e.g., running elections); however, to maintain neutrality,
which can be useful and important, it cannot be the lead agent.
Second, despite the Brahimi report’s recommendation that the UN have ac-
cess to a brigade-level quick-reaction force, armies seconded to the UN inevita-
bly have dual command structures such that central decisionmaking in a crisis
becomes nearly impossible. The committee structure of the Security Council
and the lack of autonomy within UN agencies mean that there is no way to run
military operations under Secretariat management, despite the common need
for these in interventions into dysfunctional states. The “dual key” system
instituted by Secretary-General Boutros Boutros-Ghali in UNPROFOR (in
which the UN and NATO had to approve all Bosnian targets) worked very
64. Interview DC8, March 21, 2002.

65. See Brahimi report, pars. 110–116. The brigade is presented more as a quick reaction force than
as a lead force; however, because the report refers to “command-level planning” (par. 116), the bri-
gade would have had features, had it been accepted by UN members, that we have proposed for a
lead state.
66. Both the Brahimi report and our respondents at the UN made much of the doctrinal distinction
between neutrality, or equidistance from all parties, and impartiality, meaning that a UN PKO
seeks impartial implementation of the mandate and thus possible action against forces that are un-
dermining a peace agreement. See United Nations, Report of the Panel of Peace Operations, par. 50.
Given the UN system’s inherent bias in favor of recognized governments, it is not clear whether it
can really be impartial between government and rebel spoilers.
poorly.67 In interventions in collapsed states, the UN cannot serve as the coor-

dinator of peacekeeping operations because its organization is insufªciently
capable of quick and resolute action.
In practice, the many agencies involved in postconºict peacebuilding often
work out ad hoc organizational innovations to coordinate their activities.
There is a strong need to consolidate lessons learned and best practices from
such ªeld innovations.
For example, the failure to coordinate foreign organizations operating in
Somalia in October 1992 led to a completely counterproductive situation.
Many NGOs were spending a good part of their budgets rescuing expatriates
who had been kidnapped by local staff loyal to warlords. Ransom money was
then used by warlords to purchase arms, which hindered the relief operations
in the refugee camps. Building on an innovative governing mechanism from
Operation Provide Comfort that supported the Iraqi Kurds in the wake of the
1991 Gulf War, Ambassador Robert Oakley and Philip Johnston, president of
CARE, jointly set up a U.S.-led civilian-military operations center (CMOC).
Coordination meetings among the lead state, the NGOs, and the military oper-
ation were held daily. Governance was by committee, making daily readjust-
ments across incommensurate organizational structures. This example shows
how ad hoc committees can arise spontaneously to solve at least part of the
coordination problem. CMOCs are already embedded in U.S. military doc-
trine, and have been implemented in Rwanda and Haiti. They should be built
into operational standards for all newly created PKOs that link a military and
an NGO component.68
Common military standards would also favor better coordinated and thus
more effective PKOs. Much was done in the wake of the PKO failures of the
early 1990s to reduce these coordination costs, especially between lead states
and TCCs for UN chapter 7 operations. In 1996, U.S. Secretary of State Warren
Christopher, in coordination with the French and British (in what was called
the P3 Initiative of 1997), offered the services of U.S. Special Forces to train
eight to ten battalions drawn from participant countries as part of an African
Crisis Response Initiative (ACRI). The initiative envisaged a force to provide
peacekeeping or humanitarian relief throughout the continent. In the wake of
67. United Nations, The Fall of Srebrenica, pars. 109, 117–120.

68. See Maj. Aaron L. Wilkins, “The Civil Military Operations Center in Operation Uphold Democ-
racy (Haiti),” 1997, paper presented to the research department of the Air Command and Staff Col-
lege, http://www.au.af.mil/au/awc/awcgate/acsc/97–0086.pdf.
the failure of UN forces to maintain order in Sierra Leone in May 2000, Presi-
dent Clinton initiated a concurrent program, Operation Focused Relief, to
better prepare the Nigerian, Ghanaian, and Senegalese armies for peace en-
forcement. Under the Bush administration, training was further upgraded to
include offensive light infantry operations in a program called Africa Contin-
gency Operations Training Assistance. And, in the year after the September 11
terrorist attacks, U.S. Special Forces teams were commissioned to work in
Georgia and the Philippines, in part to convey international standards for
armies engaged in counterinsurgency operations.
These forays into training to set new international standards for military
activity in peacekeeping operations favor smoother coordination in PKOs. But
they are not without their own problems. For example, even though its Rein-
forcement of African Peacekeeping Capabilities program worked in conjunc-
tion with ACRI,69 France has jealously guarded its special role in francophone
Africa. French diplomats have thereby resisted the idea of U.S.-trained troops
working in Central African Republic. What looks to U.S. ofªcials like interna-
tional standards may look to others like U.S. control. More important, as dem-
onstrated by the case of Hussein Aideed, the son of Mohammed Farah Aideed
(the warlord who initiated the intra-Hawiye civil war in Somalia), those who
are trained for counterinsurgency at U.S. facilities can later employ those skills
quite effectively in insurgencies. Training the Georgian army in 2002 posed
similar risks, at least according to military ofªcials in Russia whom we inter-
viewed. They see the Georgian army as a patchwork of private militias, any
one capable of turning against the government or one another, depending on
circumstances.70 However risky, greater investment in international standards
of peacekeeping for countries in all regions will help to reduce the costs of in-
ternational coordination of peacekeeping activities.71
accountability
If the lead state serves as the agent of the UN, how does the UN play a role as
“principal,” the authority to which the agent is responsible? Once PKOs leave
the world of neutrality and enter the world of “partisan insurgency” and
“lethal peace enforcement” (PKO activities as described to us informally at the
69. On ACRI, see http://usinfo.state.gov/regional/af/acri/; the authors thank Joe Felter for his
information on the various training programs for peacekeeping in Africa.
70. Interview M5, June 26, 2002.
71. The Brahimi report stresses the need for better training of police and military personnel in sev-
eral places. For instance, United Nations, Report of the Panel on Peace Operations, p. 18.
Pentagon), the risk of crimes and human rights abuses committed by interna-
tionally sanctioned agents increases.
Take postconºict policing, for example. Britain uses the Royal Ulster Con-
stabulary in Kosovo to strike out at spoilers. The RUC backs up UNMIK police
in rounding up criminals and bringing them to justice. As noted earlier, the
RUC’s tactics, while effective, may not have been fully in accord with human
rights conventions. In broader terms, both the UN and DynCorp, when they
recruit police to work in postconºict settings, tend to attract applicants who
were dismissed from their home forces for good reasons. And even when the
problem is not competence and integrity, there can be problems of norms of
police action. According to one story we heard, a U.S. civilian police ofªcer in
Kosovo turned to her new counterpart from a Middle Eastern country and
asked what he did back home. “I am the chief torturer in my station house,” he
replied.72
In the old UN Trusteeship system, a UN council oversaw the activities of the
trust powers. Today, responsibility for oversight to see whether the trust con-
sortium is acting in accord with its mandate is unclear, especially since the U.S.
renunciation of the International Criminal Court (ICC) treaty. The Bush admin-
istration’s demand in 2002 at the UN that no U.S. soldiers involved in a PKO
would be under the jurisdiction of the ICC suggests great reluctance by the
world’s superpower to have such oversight.
Who will oversee the actions of the lead states, especially if the ICC is ex-
cluded from examining trustee abuses? In fact, any system of oversight runs
into the problem that if criminal acts of the trust power’s forces can be exposed
and those who violated the trust can be brought to justice, the ªrst strategic
problem, recruiting agents, becomes more difªcult. The incentive for any state
to free ride increases to the extent that participation entails potentially embar-
rassing costs.
Legal oversight can work only if there is some limitation of liability for the
trustee and its representatives. Otherwise, recruitment of lead states will be
72. There was no way to conªrm this story, but abuses by CIVPOL were mentioned by several in-
terlocutors concerning peacekeeping operations. See, for example, the website of Refugees Interna-
tional, http://www.effectivepeacekeeping.org/docs/hr1414-qa.pdf. It reports, “The UN has had
problems with the behavior of the police sent by member states to serve in UN peace operations.
There have been cases of illegal behavior, such as arms and drug smuggling, and sexual
trafªcking. In 2002, several CIVPOL ofªcers were found to be running a sex slave ring in the Bal-
kans, involving women and underage girls.”
well-nigh impossible. In a talk at Stanford University’s Center for International

Security and Cooperation in February 2002, Jean Arnault, second in command
in the post-Taliban UN hierarchy in Afghanistan, proposed that only the high-
est crimes of UN personnel (or agents of the UN) be subject to criminal prose-
cution. Given the complexities and requisites of peacekeeping in the face of
insurgency, this will be a hard line to draw. International oversight but with
limits to liability is a problem that international lawyers must address.73
Although little work has been done on this problem of authorization and
monitoring, there is at least one contemporary precedent that might be worth
developing. Due to skepticism about Russian intentions, the Security Council
has compelled Russia to work closely with UN observers (UNOMIG in Geor-
gia, UNMOT in Tajikistan) and with the OSCE (in Pridnestrovie, Georgia, and
Tajikistan) to provide an aura of international legitimacy to their operations as
a lead state. To avoid the opprobrium of their international overseers, Russia
has already abandoned its initial hopes of wresting Abkhazia from former
Georgian president Eduard Shevardnadze’s grasp. Also, to the extent that the
Russians can get support from other CIS countries as they have in Georgia and
Tajikistan (whether from the CIS itself or from its Collective Security Treaty),
their intervention has the imprimatur of a UN chapter 8 action.74 Russian-led
PKOs have been subject to more international oversight than those of other
lead states.
The monitoring of Russian-led PKOs might serve as a template for countries
held under less suspicion. Consider the so-called Moscow Agreement that led
to the creation of the UN Observer Mission in Georgia (UNOMIG) in 1993 (ex-
panded in 1994). The Security Council agreed to authorize forces from the CIS.
This authorization was contingent on the CIS accepting oversight from
UNOMIG, which would be protected by CIS forces, but would report to the
Security Council violations by the CIS of the Moscow Agreement. UNOMIG
provides funds to HROAG (Human Rights Ofªce in Abkhazia, Georgia),
which is responsible for the resettlement of refugees and such things as trans-
lating human rights law into Abkhaz. Established by Security Council resolu-
73. The use of the UN Trusteeship Council as a forum to heap scorn on the United States as it was
pulling out of the Marshall Islands and Palau in the late 1980s suggests that there is no easy an-
swer to the question of oversight without the overseer being subject to harassment.
74. The Collective Security Treaty was signed by Russia, Belarus, Armenia, Kazakhstan, the
Kyrgyz Republic, and Tajikistan in 1995. In a May 2002 summit, the presidents agreed to transform
the treaty into an international regional organization under chapter 8 of the UN Charter.
tion 1077 in 1996, HROAG is jointly staffed by the Ofªce of the High
Commissioner for Human Rights and the OSCE. It was an ad hoc arrangement
driven by distrust of the CIS peacekeepers as agents of the Security Council.
Although this outcome was unplanned, and although UNOMIG’s capacity
to monitor CIS forces has been greatly constrained by its dependence on these
same forces for protection, it suggests an interesting model for future peace-
keeping operations. In this model, the military and police agents in the PKO
would agree to submit regular reports and to some level of monitoring by
human rights observers, so that there would be incentives for them to avoid
violations and discipline violators.
Suppose the UN General Assembly played the role of principal for PKOs
once authorized by the Security Council, and suppose it held the lead state
responsible for making progress reports to ad hoc committees of the General
Assembly—in which TCCs would have preferred representation. The General
Assembly would need to have the investigative services of the Secretariat, the
seed of which is already planted in the UN Ofªce of Internal Oversight Ser-
vices (OIOS).75 This would be a modern equivalent of the Trusteeship Council.
Although it would not have the authority to criminally investigate PKO
agents, it would have the investigative power to expose agents who violated
the spirit of the mission or international human rights norms.
exit and transfer of authority

In sharp contrast to classical imperialists, neotrustees want to withdraw as fast
as possible. References to “exit strategy” have ªlled the policy discourse and
debate surrounding international and U.S. operations in the Balkans, East
Timor, Sierra Leone, Afghanistan, and now Iraq.76
Our analysis of the causes of civil wars in this period suggests that the
search for an exit strategy is delusional, if this means a plan under which full
control of domestic security is to be handed back to local authorities by a cer-
tain date in the near future. Exit requires a functioning state capable of provid-
ing order. To provide order, the state needs the capacity to prevent local
insurgents from wresting control of regions and the countryside. But often the
75. On the mandate of the OIOS, see http://www.un.org/Depts/oios, and especially the secre-
tary-general’s promulgation of the investigations division ofªce, ST/SGB/2002/7, sec. 7.
76. The report of the secretary-general, “No Exit without Strategy: Security Council Decision-
Making and the Closure or Transition of United Nations Peacekeeping Operations,” United Na-
tions Doc. S/2001/394, April 20, 2001, provides an analysis of the problem from the vantage point
of the UN Secretariat.
absence of such capacity either helps explain why the country was embroiled
in a civil war in the ªrst place or is the result of the war, or both. Low per capita
income predicts higher risks of civil war, we have argued, because it correlates
with conditions favorable to insurgency—pervasive poverty, poor roads,
rough terrain, and, most of all, state inability to perform rural police work and
counterinsurgency in a competent, non-self-defeating manner (an extremely
difªcult task even for strong states). This task is harder than it was in Japan
and Germany after World War II, countries that had the administrative compe-
tence to effectively wage total war.
To make matters worse, the agents of neotrusteeship lack any viable theory
about how to build a functioning state apparatus under these conditions. In
fact, there are reasons to believe that the very presence of international troops
may work against this end. Suppose you are the person or group favored for
control of the capital city by the UN, the United States, NATO, or whoever the
main external provider of force is. The more you invest in developing local ca-
pacities for self-government, the greater the risk that the external patron will
ramp down its mission and leave, and the greater the number of potential local
rivals empowered by the new security apparatus. Thus there can be a disincen-
tive to actively or successfully develop indigenous state capabilities, leading to
what President Bush’s former special envoy to Afghanistan, Zalmay Khalilzad,
has called “security welfare states.”77 In this context, the concern is reasonable,
though it remains unaddressed in the extensive literature on PKOs.
One possible solution would be to build into the initial agreements that
authorize international intervention (e.g., peace agreements, UN mandates) a
plan under which costs of international peacekeeping would increasingly be
borne by the state being reconstructed. To recover the tax revenue, the leaders
of the rebuilding state would have an incentive to establish legitimate domina-
tion so that the state could stand on its own without core functions being pro-
vided by outside forces. On the international side, such deals would help with
the recruitment problem by allowing interveners to put some bounds on the
costs of peacekeeping missions. They could expect that even in the worst
case—if the local capacity for stable self-government were to develop very
slowly—the peacekeeping mission would ultimately be self-supporting.
This mechanism could also be used to provide incentives for local actors to
ªnd durable, institutionalized ways to resolve their differences. For example,
77. Quoted in Pamela Constable, “U.S. Plans Greater Security Role to Help Curb Regional
Fighting,” Washington Post, February 25, 2002, p. A16.
in an article on the Nagorno-Karabakh problem, David Laitin and Ronald

Suny suggest that Azerbaijan use future oil revenues to pay for an interna-
tional peacekeeping force protecting Armenians in Nagorno Karabakh should
this region be returned to Azeri sovereignty. The force would remain until Ar-
menian residents of Azerbaijan vote to disband it. Thus, the costs of paying for
protection of Armenians living in Azerbaijan would provide incentives to the
Azeri government to credibly resolve Armenian insecurities, leading them to
work to end the arrangement in a mutually beneªcial way.78
It may seem callous to tax the citizens of a war-devastated country who have
suffered the most for its reconstruction. And such a tax paid to international
interveners would divert funds from use for economic development. Critics
might also doubt that war-ravaged places such as Afghanistan, Kosovo, and
Congo have the potential to produce sufªcient tax revenues to make such a
scheme feasible.
It would be callous, and often counterproductive, to demand such a price
right from the start. But we are not proposing this. The initial deal could have
the payments start at some reasonable time in the future, after basic security
and economic exchange had been reestablished. There is a solid ethical case for
such an arrangement. As the Brahimi report obliquely suggests and as numer-
ous PKO experts have observed, the central problem with post–Cold War
peacekeeping operations is that politicians in the rich, Western states do not
believe that their voters are willing to pay anything in blood and virtually any-
thing in money for peace and security in places they have never heard of.79 As
noted above, we think that the United States and other rich-country publics
should be willing to pay more, both on ethical grounds and in their own long-
term self-interest. But we doubt that exhortation, as in the Brahimi report, will
have much effect, so one must look for realistic alternatives.
The citizens of the postwar country have the strongest interest in a stable,
functioning state that can provide public goods, and so are plausibly the most
willing to pay. This suggests that an intertemporal scheme in which they pay
some of the costs by mortgaging some future revenues is an appropriate solu-
78. David D. Laitin and Ronald Grigor Suny, “Thinking a Way Out of Karabakh,” Middle East Pol-
icy, Vol. 7, No. 1 (October 1999), pp. 145–176.
79. September 11 clearly increased the U.S. public’s cost tolerance for such missions when they can
be linked to ªghting terrorism. Nonetheless, in Afghanistan, the Bush administration has acted as
if the tolerance level remains extremely low. The U.S. public’s patience was being seriously tested
in Iraq only months after Saddam Hussein’s fall.
tion.80 As noted earlier, this kind of arrangement has already been tried for sev-
eral UN PKOs. We are proposing that the model be developed and applied in a
more general fashion.
Regarding economic feasibility, the central activity of the modern state is to
tax its citizens and use the money to provide public goods. If, once recon-
structed, a government lacks the resources to do this, then it can only be a
ward of the de facto trusteeship system. There can never be effective gover-
nance unless the wealth of the country, however paltry, is subject to tax. In any
event, countries devastated by civil war often have signiªcant resources over
which the combatants fought, such as oil, timber, or diamonds.81 It is worth
noting that the two nonrecognized rump states of former Somalia (called the
Republic of Somaliland and the Republic of Punt) are not eligible for interna-
tional welfare. Yet they are able through taxation on the camel and goat trade
(along with donations from their diasporas) to run effective governing
institutions.
Perhaps a more important point for demanding intertemporal taxation is
that taxation of the citizenry motivates those who are taxed to monitor how
their money is being spent. As we know from the history of the modern state,
rulers are often forced into responsible governance due to pressures from their
taxed subjects. Successful exit from a state that has no taxing authority or capa-
bility is unlikely.
A ªnal objection to intertemporal taxation to help pay for the reconstruction
of collapsed states concerns monitoring and administration—who or what
would protect against neocolonial exploitation? Any such agreement would
require a high level of transparency from the outset and most likely new or
modiªed international institutional input to monitor its application. New or
modiªed arms of the World Bank or International Monetary Fund would be
natural candidates. The most straightforward implementation of the idea
would have these institutions empowered and funded to provide loans to col-
lapsed states for the purpose of paying for peacekeeping and the reconstruc-
80. Another reasonable objection is that such a deal requires a legitimate local authority to sanc-
tion it on behalf of the citizens of the war-torn country, which is precisely what is lacking after a
civil war. But the UN and other international agencies have shown themselves skilled at helping to
produce plausibly legitimate local authorities through elections (e.g., Cambodia) or other means
(e.g., Afghanistan). In contrast to classical imperialism, neotrusteeship’s focus on exit implies that
the trustees try to develop legitimate local authorities right from the start.
81. For example, oil exporting states have had more than twice the annual odds of civil war onset,
other things held equal. Fearon and Laitin, “Ethnicity, Insurgency, and Civil War,” p. 11.
tion of basic police and security functions. Because economic development and
ªnancial stability clearly depend on a state being able to provide basic physical
security for its citizens, such loans would be consistent with the mandates of
both organizations.
Even with cost-shifting to the reconstructing state, however, for some cases
complete exit by international interveners may never be possible. In Kosovo
and Bosnia, for example, it is not clear that transfer to full sovereignty is an
appropriate goal, because in both cases the problems include not only state
capacity but also intense disputes over the proper borders of the state.82 In rec-
ognition of this, one State Department ofªcial suggested that there needs to be
more thinking not about exit but rather integration of the neotrust territory
into a plethora of international organizations. The purpose is to provide con-
tinuous and unobtrusive monitoring of the peace and, in the longer run, to
make the national level of government irrelevant for people in comparison to
the local and supranational levels.83
In fact, UN ofªcials in the DPKO insisted in our discussions that in the com-
ing years in Kosovo, transitional administration has no end goal. By 2000, the
Kosovo mission did its own recruitment and thus did not need to go through
the UN pipeline or bureaucracy to get personnel. The implicit agenda is to
keep any talk of the international actors’ exit off the agenda. If any party to the
agreement states a ªnal goal openly (whether integration into Yugoslavia or
outright independence), one DPKO ofªcial warned us, the system will break
down. So all parties are advised to keep “eyes off ‘ªnal status’” with the vague
phrase of “substantial autonomy” (as in UN resolution 1244) for Kosovars.
Any party that would sign a clear ªnal status agreement would be threatened
by its own radical wing. A common view at the UN and in the State Depart-
82. We do not rule out the possibility that partition plans may in some cases be the best way to re-
construct states after an ethnic civil war, but we see these as having serious problems as well. See
James D. Fearon, “Separatist Wars, Partition, and World Order,” Stanford University, 2003; David
D. Laitin, “Partition and Territorial Concentration,” Stanford University, 2003; and David D. Laitin,
Identity in Formation: The Russian-Speaking Populations of the Near Abroad (Ithaca, N.Y.: Cornell Uni-
versity Press, 1998), chap. 9. For arguments in favor of partition as a solution to the problems of
post–civil war reconstruction, see Chaim Kaufmann, “Possible and Impossible Solutions to Ethnic
Civil Wars,” International Security, Vol. 20, No. 4 (Spring 1996), pp. 136–175; and Chaim Kaufmann,
“When All Else Fails: Evaluating Population Transfers and Partitions as Solutions to Ethnic
Conºict,” in Barbara F. Walter and Jack Snyder, eds., Civil Wars, Insecurity, and Intervention (New
York: Columbia University Press, 1999).
83. Interview DC6 at the State Department, Bureau of Intelligence and Research, March 21, 2002.
In effect, European supranational institutions are to play a role similar to that once played by the
Ottoman Empire.
ment was that Bosnia will remain a country and Kosovo a region in which
national government will be unable to maintain interethnic peace for a long
time.84
A partial solution, however, is already emerging in former Yugoslavia. For
all the anxieties about the future, there remain all kinds of nongovernmental
and intergovernmental civil servants in Bosnia who provide surveillance and
raise the costs for any local politician who might reinitiate hostilities. Con-
tinued international surveillance will be the local political price paid for failure
to provide security to a country’s civilian population. That Bosnia and Kosovo
are in Europe where the rich countries have many such institutions makes this
model—continued international surveillance through membership in a pleth-
ora of organizations—plausibly effective. It may allow for a peaceful transfer
after a long period of transitional surveillance and integration with Western
supranational institutions.
Conclusion
Major international interventions to prop up and rebuild failed states are not a
temporary aberration in the course of international politics. Rather, they reºect
more durable, even structural characteristics of the present international sys-
tem. Since the end of World War II, there has been a steady growth in the num-
ber of (mainly postcolonial) states rendered dysfunctional by years of rural
guerrilla war, corrupt rule, or both. For the major powers and many other
states, the biggest external threats now derive not from the risk of strong states
wanting to conquer and annex territory, but from diverse security, economic,
and even health consequences emerging from political conºict, state collapse,
and misrule in the third world. Moreover, independent of the various threats
posed to the North, persistent civil war and lack of economic development in
the South are two of the greatest sources of human suffering on the planet.
We have argued that collapsed states pose an international collective action
problem whose solution will involve multilateral interventions that share
the initial burden across a wide variety of states, international and non-
governmental organizations, and corporations. Although the UN system mo-
bilized an impressive number of such collective actions in the 1990s, its
approach to PKOs for very weak states remains ad hoc, underrationalized, and
84. Interview UN2, March 3, 2001.

inadequate. Where the state apparatus is largely notional and local conditions
favor guerrilla techniques, the pressures for mission creep in a PKO are typi-
cally powerful, as it becomes clear that exit without a return to war demands
sustained transitional administration by or with the armed support of interna-
tional parties. The reality of state weakness means that forces maintaining
peace need to do state building if there is to be any hope for exit without a re-
turn to violence. Adding more and better-armed peacekeepers, or holding
more talks on political reconciliation or constitutional design, will not solve
this problem quickly or even at all.
We identiªed four main problems that bedevil policy in regard to collapsed
states: recruitment, coordination, accountability, and exit. As for recruitment,
we emphasized the need for a layered set of participants, because the tasks in-
volved require the resources and capabilities of a variety of countries, NGOs,
and corporations. Recruitment of regional or major powers with particular
national security or economic interests in a collapsed state is easier and should
be encouraged; however, the potential for abuse by these agents should be
better monitored via new reporting requirements and possibly new institu-
tions developed within the UN.
With so many actors involved in the governance of collapsed states, severe
coordination problems inevitably arise. A lead state is therefore a sine qua non
for mission success. Although many UN organizations will be involved, the
UN is ill suited to be the lead organization for coordination purposes. Further-
more, because the lead state needs to act in ways that are not transparently im-
partial, were the UN to play that role, it would compromise its long-term
ability to act as an honest broker. A lead state or regional organization should
set the terms of coordination among its own agencies, those of the UN (which
provides not only functional support but also legitimation), and the host of
other organizations serving functional roles.
Giving a lead state coordinating power raises the issue of accountability.
Trigger-pulling power, even in the name of service to international peace, can-
not be allowed to stand unchecked. Whether or not the ICC survives, we rec-
ommend some newly constructed arm of the UN based on principles and
lessons learned by the Trusteeship Council.
Finally, due to the sources of the civil wars that lead to collapsed states, suc-
cessful exit from neotrusteeship will be extremely difªcult in most cases. We
have stressed the need to develop local tax-collecting capability as an incentive
to move the country from international welfare toward self-governance and, in
some cases, the notion of transfer not to full sovereignty but rather as a state
embedded in and monitored by international institutions.
It may be that de facto states or statelike organizations will eventually

emerge out of anarchy without the involvement of major powers or interna-
tional institutions. But the local and international costs and risks of such “natu-
ral” processes of state formation can be very high in a world with open
borders, weapons of mass destruction, and no internationally legitimate means
for redrawing state boundaries. Major powers and international organizations
will have little choice but to involve themselves in state building. It makes
sense, then, to construct new institutions and operating procedures that will be
effective and fair in dealing with the challenges posed by collapsed states. As
we have argued, the current, ad hoc and underrationalized arrangements
ought to be reformed in the direction of neotrusteeship.

Fearon, James D., and David D. Laitin. 2004. "Neotrusteeship and The Problem of Weak States." International Security 28 (4) - 5-43

Uploaded by

Copyright:

Available Formats

Fearon, James D., and David D. Laitin. 2004. "Neotrusteeship and The Problem of Weak States." International Security 28 (4) - 5-43

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fearon, James D., and David D. Laitin. 2004. "Neotrusteeship and The Problem of Weak States." International Security 28 (4) - 5-43

Uploaded by

Copyright:

Available Formats

Neotrusteeship and the Problem of Weak States

Neotrusteeship and the James D. Fearon and

International Security, Vol. 28, No. 4 (Spring 2004), pp. 5–43

Ironically, the Bush administration has since undertaken state-building pro-

What Is Neotrusteeship and Why Is It Happening?

Figure 1. Civil War and UN PKOs since 1945.

a UN peacekeeping operation—a good indicator of the spread of neotrustee-

and administration. Freetown, the capital of Sierra Leone, is swamped with

These differences stem from changed motivations for international interven-

The UN PKOs and the Brahimi Report

Collective action is no simple matter, however, especially absent a suprana-

rogue regime or a collapsed state, and the appropriate response—witness Iraq

tial deployment of UNAMSIL was disastrously inadequate, requiring rescue

36. Interview UN3, April 3, 2001.

Insurgency and the Logic of “Mission Creep”

not be able to leave without substantially rebuilding state security institutions

42. United Nations, Report of the Panel on Peace Operations, p. 9.

Four Strategic Challenges for Neotrusteeship

52. United Nations, Report of the Panel on Peace Operations, p. 17.

In the post-Soviet region, when Russia or the Commonwealth of Independent

Christine Gray refers to bitter complaints by the Indian permanent representa-

61. Gray, “Peacekeeping after the Brahimi Report,” p. x.

64. Interview DC8, March 21, 2002.

poorly.67 In interventions in collapsed states, the UN cannot serve as the coor-

67. United Nations, The Fall of Srebrenica, pars. 109, 117–120.

well-nigh impossible. In a talk at Stanford University’s Center for International

exit and transfer of authority

in an article on the Nagorno-Karabakh problem, David Laitin and Ronald

84. Interview UN2, March 3, 2001.

It may be that de facto states or statelike organizations will eventually

You might also like