. 3
( 10)


Libyan administration in the territory, a mission that neither followed a civil
con¬‚ict nor came close to meeting the de¬nition of a “major” operation.
Figure 3.2 lists three major peacebuilding operations that were deployed
in 1999 to East Timor, Kosovo, and Sierra Leone.16 I offer some preliminary
observations on these three missions in later chapters of the book, but I do
not include them as in-depth case studies because, at the time of writing,
all three were still too new to judge their effects with the same degree of
con¬dence. The eleven remaining missions in Figure 3.2 “ Angola, Rwanda,
Liberia, Cambodia, Bosnia, Croatia, El Salvador, Nicaragua, Guatemala,
Namibia, and Mozambique “ represent a rich and diverse set of cases. Five
of these missions took place in Africa, three in Latin America, two in Europe,
and one in Asia. They concern some of the most prominent and distressing
international events of the 1990s, including the war in the former Yugoslavia
and the genocide of ethnic Tutsis in Rwanda. Some of these missions have
been viewed as great successes, such as the operations in El Salvador and
Mozambique; others, like the mission in Angola, have been branded failures.
Given the diversity of these eleven cases “ and the fact that they include all
of the major international postcon¬‚ict peacebuilding operations launched
between 1989 and 1998 “ it should be possible to draw general conclusions
about postcon¬‚ict peacebuilding.

16 Another mission “ to the Democratic Republic of Congo “ was formally authorized in 1999
but did not deploy (more than 200 troops) until summer 2001.

Angola and Rwanda
The Perils of Political Liberalization

No peacebuilding operations have failed more miserably than the missions
in Angola and Rwanda: Both countries slipped back into violence before
peacebuilders could accomplish their tasks. The war in Angola continued
for several years at varying degrees of intensity, while Rwanda experienced
not only resurgent war but genocide. The question under consideration,
however, is not how these missions turned out in general but, rather, how
effective the peacebuilders™ liberalization efforts were as a remedy for civil
violence. Let us examine each of the cases in turn.

Angola, located on the southwestern coast of Africa, was ruled by Portugal
until 1975. As many as eleven national liberation movements sprang up
during the 1960s to ¬ght against the colonial government.1 By the early
1970s, three of these organizations had emerged as the strongest indigenous
groups: the Movimento Popular de Libertacao de Angola (MPLA), the Frente
¸˜ ˜
Nacional de Libertacao de Angola (FNLA), and the Uniao Nacional para
a Indepˆ ndencia Total de Angola (UNITA). These organizations not only
waged a guerrilla war against the Portuguese colonial authorities but also
battled among themselves for predominance.3 After the departure of the
Portuguese, the MPLA gained control of the Angolan capital, Luanda, and
decisively defeated the FNLA in 1976, leaving UNITA as its only major rival.
Fighting in Angola between UNITA and the MPLA continued for more than
¬fteen years, sustained in part by foreign military aid for both sides. UNITA
received support from South Africa and later the United States, while the
MPLA enjoyed the backing of the Soviet Union and its allies, including Cuba,

1 2 3
Papp 1993, p. 162. Hampson 1996, p. 87. Fortna 1993a, pp. 376“377.

The Peacebuilding Record

which deployed up to ¬fty thousand combat troops alongside MPLA forces
in Angola.4
During the early 1980s, international efforts to negotiate a peaceful set-
tlement of the con¬‚ict repeatedly failed.5 The ¬rst major progress came in
December 1988, with the signing of the agreement that ended the war in
neighboring Namibia.6 Among other things, this agreement provided for a
cease-¬re in Angola, the withdrawal of Cuban troops from the country, and
the termination of South African support for UNITA.7 The United Nations
subsequently deployed seventy military observers to monitor the phased de-
parture of Cuban troops.8 It soon became clear, however, that these measures
were insuf¬cient to stop the war between the MPLA and UNITA, in part be-
cause both parties continued to receive external support.9 Negotiations to-
ward a ¬nal resolution of the Angolan con¬‚ict resumed in 1989 and contin-
ued through to the spring of 1991, in the presence of a Portuguese mediator
along with American and Soviet observers.10
On May 2, 1991, the parties initialed a comprehensive peace accord in
Bicesse, Portugal, which included provisions for a full and permanent cease-
¬re, “free and fair” multiparty elections, respect for civil liberties including
freedom of association, and the integration of the two armies into a single
national army.11 These provisions were to be implemented by the Angolan
parties under the supervision of the United Nations and other international
observers. Soon after, the Security Council approved the creation of a new
UN veri¬cation mission in Angola (UNAVEM II) to ensure that the par-
ties carried out their responsibilities under the Bicesse accord.12 The pri-
mary responsibility for monitoring the implementation of the accords was
assigned to two commissions consisting of representatives from UNITA and
the MPLA, along with observers from Portugal, the United States, and the
Soviet Union. The mandate of UNAVEM II was to ensure that these joint
commissions carried out their monitoring responsibilities “ in other words,
to monitor the monitors. Several other public and private international agen-
cies provided humanitarian and logistical aid to the local parties, including
the International Committee of the Red Cross, the UNDP, and the UN High
Commissioner for Refugees.13
In November, the MPLA and UNITA settled on the details of the elec-
toral arrangements: The president of Angola would be elected by direct
and secret suffrage through a majority system “with recourse to a second
round, if necessary,” while the National Assembly would be elected through

4 5 Papp 1993, pp. 175“180; and Hampson 1996, pp. 89“91.
Krˇ ka 1997, p. 79.
6 See Chapter 8 for a discussion of the operation in Namibia.
7 8 Malaquias 1996, p. 92. 9 Papp 1993, p. 186.
United Nations 1996b, pp. 234“235.
10 For an account of these negotiations, see Rothchild 1997, pp. 120“135.
11 “Peace Accords for Angola [Acordos de la Paz para Angola],” UN Security Council document
S/22609 (May 17, 1991), annex.
12 13 Hampson 1996, p. 124.
UN Security Council Resolution 696 (May 30, 1991).
Angola and Rwanda 65

proportional representation.14 UNAVEM II™s mandate was subsequently ex-
panded to oversee the preparations and conduct of the elections, which were
scheduled for the fall of 1992.15
The agreement to hold elections and to transform the Angolan polity into
a “multiparty democracy,” as the Bicesse accord promised, represented a sig-
ni¬cant shift in the MPLA™s traditional preference for one-party government
on the Soviet model.16 The leader of the MPLA government of Angola, Jos´ e
Eduardo dos Santos, had insisted as recently as early 1990 that “only the
one-party system realistically serves our country.”17 It was not until peace
talks shifted to Portugal in the spring of 1990, and the United States (to-
gether with the Soviet Union) started applying pressure on both parties for
a settlement that included multiparty elections, that dos Santos accepted the
principle of holding an open election as part of a broader peacebuilding
Soon after, dos Santos also agreed to liberalize the Angolan economy.19
Like the shift to liberal democracy, of¬cial support for market-oriented
economic reforms contrasted with the MPLA™s long-standing espousal of
Marxist-Leninist economic policies (dating back to at least 1964), which
included central economic planning and state-run cooperative farms and in-
dustries.20 The Angolan government™s endorsement of market democratic
principles re¬‚ected the many pressures it faced in the early 1990s. Its princi-
pal foreign patron, the Soviet Union, was collapsing at a time when the war
with UNITA had reached a stalemate. The MPLA™s only realistic prospect
of gaining external investment and ¬nancial aid for Angola was to negoti-
ate a rapprochement with the United States and the international ¬nancial
institutions, which in turn required settling the con¬‚ict with UNITA and
implementing democratic and market-oriented reforms.21
It would, therefore, be misleading to suggest that the ensuing peacebuild-
ing operation simply monitored the implementation of a peace agreement
that the Angolan parties had themselves devised, since central components
of the agreement, including the commitment to multiparty democracy, had
been urged upon the parties by Western states involved in the negotiations
(the United States and Portugal) and by the very international agencies that
later undertook key peacebuilding tasks (the UN, the IMF, and the World
Bank). The peacebuilding operation was not merely an exercise in con¬‚ict
management, but part of a broader international effort to transform Angola

14 United Nations 1996b, p. 239.
15 UN Security Council Resolution 747 (March 24, 1992).
16 Krˇ ka 1997, p. 77; and Rothchild 1997, p. 113.
17 18 Walter 1994, p. 144. 19 Hampson 1996, p. 111.
Cited in McCormick 1991, p. 5
20 Lodico 1996, p. 105; Ciment 1997, p. 16; and Rothchild 1997, p. 113. The notable exception
was the oil sector, which remained open to Western corporations.
21 Papp 1993, p. 188; and Ciment 1997, pp. 16 and 169.
The Peacebuilding Record

from a warring one-party state into “ it was hoped “ a peaceful and prosper-
ous market democracy.
The peace process suffered from many delays. Such problems as lack of
of¬ce space and housing in Luanda, and UNITA™s initial refusal to admit
UN reconnaissance teams into its territory, slowed the initial deployment of
international observers.22 The original timetable called for UNITA and the
MPLA to transport their troops to designated “assembly areas” by August 1,
1991, and for the demobilization process and the formation of a new national
army to be completed before the holding of national elections in the fall of
1992. Both deadlines were missed. The arrival of troops in the assembly
areas was delayed for a number of reasons, including the lack of adequate
transportation, the shortage of food and medicine at assembly areas, and
the apparent reluctance of both parties to relinquish control over their best
military forces and equipment.23 In April 1992, the Angolan parliament
scheduled the elections for September 29 and 30, but by the end of September,
demobilization was still far from complete: Only 65 percent of the MPLA
and 26 percent of UNITA forces had been processed to return to civilian life,
while only eight thousand (of a planned ¬fty thousand) soldiers had been
integrated into the new Angolan army.24
Meanwhile, preparations for the elections moved forward under the su-
pervision of international monitors, and with the advice of experts from the
United Nations Development Program. Registration of voters took place be-
tween May 20 and August 10, 1992, and the electoral campaign extended
from August 29 to the eve of the election, September 28. Despite increased
reports of political violence as the voting neared, including several clashes
between UNITA and MPLA supporters,25 there were no major violations of
the cease-¬re.26 An estimated 92 percent of eligible voters cast ballots during
the two days of polling, which took place without serious incidents of vio-
lence, under the supervision of approximately eight hundred international
observers.27 Although some polling irregularities were noted, the special
representative of the UN secretary-general, Margaret Joan Anstee, reported
that there was “no conclusive evidence of major, systematic or widespread
fraud, or that the irregularities were of a magnitude to have signi¬cant ef-
fect on the results” “ a view that was echoed by most other international
On October 3, while the ballots were still being counted, UNITA of¬cials

22 Fortna 1993b, pp. 398“399; and Fortna 1995, pp. 290“291.
23 Fortna 1993b, pp. 399“400; Malaquias 1996, p. 95; United Nations 1996b, p. 242; and
Krˇ ka 1997, pp. 87“88.
24 Anstee 1996, p. 48; Lodico 1996, p. 111; and Krˇ ka 1997, p. 87.
25 Hampson 1996, p. 113.
26 United Nations 1996b, p. 240. There were, however, several reported skirmishes between
UNITA and MPLA forces between January and September 1992 (Stedman 1997, p. 37).
27 28 Anstee 1996, p. 205. See also Economist 1992, p. 51.
Krˇ ka 1997, p. 88.
Angola and Rwanda 67

complained that the election had been fraudulent.29 In an apparent effort
to counter rumors that the MPLA was leading in the vote count, the head
of UNITA, Jonas Savimbi, declared that the MPLA “is not winning and
cannot win” and insisted that, “in all provinces, UNITA is ahead both in
the presidential and the parliamentary results.”30 In fact, Savimbi and his
UNITA party were trailing in both the presidential and parliamentary races.
Final results were announced on October 17. The MPLA won 54 percent of
the seats in the legislature versus UNITA™s 34 percent, while dos Santos took
49.6 percent of the presidential vote compared to 40.1 percent for Savimbi.31
Neither candidate had received the minimum 50 percent support required for
a ¬rst-round victory, but instead of waiting for a second round of elections to
be held, as required under the Bicesse accord, Savimbi denounced the results
and launched UNITA on a nationwide campaign to occupy municipalities
and take over the MPLA government™s administrative structures. Fighting
¬rst broke out in areas of Huambo province and soon spread to other parts
of the country, including Luanda, in spite of international efforts to broker
a new cease-¬re. In the ¬rst week of November, MPLA forces had regained
control of Luanda and hunted down UNITA of¬cials and supporters who
were attempting to ¬‚ee the capital city. UNITA fared better in the countryside,
where its forces surrounded key provincial capitals and ports. By the start
of 1993, UNITA was estimated to control 50 of Angola™s 164 municipalities
and 75 percent of the country™s territory, including virtually all of the roads
and Angola™s major diamond-producing region.32
Fighting continued for the next two years, killing approximately three
hundred thousand people, or about 3 percent of the country™s population,
which exceeded the estimated number of Angolans killed in the eighteen
years of civil war leading up to the signing of the Bicesse accord in 1991.33
In November 1994, after the MPLA had reversed UNITA™s territorial gains,
the parties signed a new cease-¬re and peace agreement in Lusaka, Zambia,
which provided for the reintegration of the two armies and the formation of
a government of national unity with representation from both the MPLA and
UNITA, as well as a second round of presidential elections.34 But the Lusaka
agreement did not hold, and the Angolan civil war continued. Frustrated
with the recalcitrance of the local parties and fearing for the safety of its

29 United Nations 1996b, p. 244.
30 Anstee 1996, p. 201. Of¬cials from the National Electoral Commission, along with
UNAVEM II observers, investigated UNITA™s complaints and found no evidence of sys-
tematic fraud or fraud that would have signi¬cantly affected the outcome of the election
(United Nations 1996b, p. 244).
31 32 Meldrum 1993b, p. 45; and Anstee 1996, pp. 329“330.
Pereira 1994, pp. 16“17.
33 USIP 1996; and Tvedten 1997, p. 111.
34 “Lusaka Protocol,” November 15, 1994. For a description of the negotiations leading up to
the signing of the Lusaka Protocol, see Hare 1998. With the signing of the Protocol, the UN
transformed UNAVEM II into UNAVEM III.
The Peacebuilding Record

personnel, the United Nations ¬nally decided in February 1999 to terminate
its peace mission in Angola.35 Hostilities continued until April 2002 when
Savimbi was murdered, prompting the MPLA and UNITA to sign a new
There is no shortage of possible explanations for the failure of UNAVEM
II. Many commentators criticize the mission for failing to ensure that the
factions were fully demobilized before elections took place.36 Others submit
that the “winner-take-all” electoral system used in the Angolan elections,
which offered few incentives for the losing party to accept the results, was a
principal cause of the mission™s failure.37 Another interpretation blames the
design of the Bicesse accords themselves, suggesting that they were vague
and lacked “procedures for dealing with violations.”38 A variant of this
argument is that the peace agreement was ¬‚awed in assigning the primary
responsibility for the implementation of the accords to the Angolan parties,
rather than to the UN itself.39 Yet others contend that the United Nations
and the international community devoted insuf¬cient personnel and material
resources for overseeing the implementation of the accords.40
Perhaps the simplest and most obvious explanation for the failure of the
mission is that the Angolan parties themselves lacked suf¬cient political will
to achieve a lasting peace.41 Savimbi, in particular, was widely portrayed as
the “spoiler” of the peace process: a man who, despite his stated commitment
to peace, had no intention of demobilizing his forces or respecting the results
of the elections, unless they advanced his goal of becoming president of
Whatever the actual causes of renewed violence may have been, at least
one conclusion seems clear in the aftermath of UNAVEM II: The 1992 elec-
tion in Angola did not facilitate a reconciliation of the formerly warring
parties. On the contrary, the election forced the latent con¬‚ict between
Savimbi and dos Santos into the open, generating a climate of competi-
tion and hostility between their two parties, who approached the election
as “war by other means.”43 During the campaign period, both Savimbi and
(to a lesser degree) dos Santos used rhetoric that played upon Angola™s ex-
isting social and political divisions, including urban“rural and ethnic“tribal

35 Security Council Resolution 1299 (February 26, 1999). The UN mission had been reorga-
nized and renamed the UN Observer Mission in Angola in 1997. On the rationale for ending
the mission, see “Report of the Secretary-General on the United Nations Observer Mission
in Angola (MONUA),” S/1999/49 (January 17, 1999).
36 Meldrum 1993b, p. 24; Marcum 1993, p. 223; Bertram 1995, pp. 398“399; Fortna 1995,
p. 294; Vines 1995, p. 8; Malaquias 1996, p. 95; and Krˇ ka 1997, pp. 92“93.
37 Cohen 1993; Pereira 1994, pp. 15“16; Knudsen with Zartman 1995, p. 136; Ciment 1997,
p. 20; and Rothchild 1997, p. 135.
38 Fortna 1993b, p. 402; and Walter 1994, p. 152.
39 Walter 1994, p. 120; Malaquias 1996, pp. 94“95; and Krˇ ka 1997, p. 92.
40 Hampson 1996, p. 122; Rothchild 1997, p. 134; and Ciment 1997, p. 20.
41 42 Brittain 1998. On “spoilers,” see Stedman 1997.
Fortna 1995, p. 292.
43 Wesley 1997, pp. 98“109.
Angola and Rwanda 69

differences, in order to generate popular fear of the other party™s intentions.44
These verbal battles reduced the prospects of peaceful cooperation by making
it even more dif¬cult for the parties to accept any type of postelection demo-
cratic coexistence with their adversaries. Although it was ultimately Savimbi
who rejected the vote (and who, therefore, bore primary responsibility for
the collapse of the peace process), neither of the Angolan leaders seemed
willing to accept the possibility of electoral defeat in the weeks leading up
to the election or during the vote-counting period.45
The ordinary citizens of Angola may have had a clearer understanding of
the election™s likely effects than the international peacebuilders. “The people
are full of fear,” reported one Roman Catholic bishop in the days before
the vote, “and a lot of them are leaving their villages because they don™t
know what will happen during the elections.”46 International agencies, by
contrast, made unrealistic best-case assumptions about the elections. Indeed,
the entire plan for the democratization of Angola was based on the ¬‚imsy
hypothesis that encouraging democratic competition among recently warring
rivals would reduce the likelihood of renewed violence in Angola. In practice,
however, the elections brought the long-running contest between the MPLA
and UNITA to a head, with devastating results. Not only did the elections
fail to reconcile the formerly warring parties; they also apparently served
as a catalyst for renewed ¬ghting and helped to destroy the fragile peace
that had prevailed since the signing of the Bicesse accords. As one observer
wrote in the aftermath of these events, “No longer can it be assumed that a
cease¬re and internationally monitored elections are suf¬cient to establish a
lasting peace in a country torn by years of bitter civil strife.”47
This ¬ghting might have been avoided if the international peacebuilding
agencies had ensured that the parties were completely disarmed before the
elections took place “ certainly, UNITA™s military capacity gave Savimbi some
¬‚exibility in deciding whether or not to renew ¬ghting. Be that as it may,
the attempts to transform Angola into a democratic state in 1992 seemed to
have a detrimental impact on the country™s peace process.48

Rwanda is a small Central African country whose population is divided into
two main ethnic groups: According to the 1991 census, just over 90 percent

44 Wesley 1997, p. 102; Heywood 2000, pp. 216“218.
45 46 Quoted in Noble 1992. 47 Meldrum 1993a, p. 24.
Wesley 1997, pp. 102“105.
48 The economic dimension of peacebuilding was virtually nonexistent in Angola during the
1990s, and therefore is not addressed in this chapter. The World Bank conducted a few small
projects in the country, and the IMF did not commit any ¬nancial resources or negotiate an
adjustment program with Angola, due to the country™s failure to meet the most basic macroe-
conomic and policy “benchmarks” that the IMF required as a precondition for beginning
discussions of an adjustment program. See Hodges 2001.
The Peacebuilding Record

of the country™s inhabitants were ethnic Hutu, while roughly 8 percent were
ethnic Tutsi.49 During the period of Belgian colonial rule prior to 1959, vir-
tually all the chiefs and subchiefs who dominated Rwandan political and
economic life had been Tutsi. In 1959, however, Belgian of¬cials found
themselves under pressure, both internationally and from Hutus within
Rwanda, to “democratize” the country™s indigenous ruling structures, which
in practice meant replacing the Tutsi elite with Hutus.50 This goal was largely
accomplished by the time Rwanda formally became independent in July
Tutsi resistance to Hutu rule continued in the following years, which
in turn provoked waves of repression by the Hutu government against the
Tutsi population, spurring the emigration of tens of thousands of Tutsis
from Rwanda. Many of these Tutsi refugees settled in neighboring Uganda,
some achieving positions of in¬‚uence under the regime of Uganda™s presi-
dent, Yoweri Museveni. In 1990, approximately four thousand Tutsis who
had served in the Ugandan army established an armed wing of the expatri-
ate Tutsi organization, the Rwandan Patriotic Front (RPF). Together with
about three thousand other exiles, this force invaded Rwanda in October
1990 with the goal of overthrowing the Hutu regime.51 After the invasion
was repulsed by Rwandan government forces, the RPF reverted to guerrilla
tactics. Sporadic ¬ghting continued until July 1992, when a cease-¬re agree-
ment was reached with the assistance of mediation from the Organization
of African Unity. After a year of further negotiations, the RPF and the Hutu-
dominated Rwandan government signed a comprehensive peace accord in
Arusha, Tanzania, in August 1993.
The Arusha Accords provided for the creation of a transitional govern-
ment in which Tutsis would share power with Hutus, the integration of the
two armies, the return of refugees to Rwanda, and the organization of par-
liamentary elections for 1995, all of which was to be supervised by a neutral
international force.52 “The principal purpose of the Arusha accord,” writes
Samuel Makinda, “was to create a participatory, multi-party democracy in
which a government could be voted out of power and opposition parties
could function freely.”53 Indeed, the Hutu president of Rwanda, Juv´ nal e
Habyarimana, had been under considerable pressure to liberalize the coun-
try™s political system and to hold democratic elections “ pressure that came
not only from the RPF, which demanded a share of governmental power in
Rwanda, but also from opposition groups within Rwanda, as well as from
international donors, many of whom argued that democratization and a
multiparty system were necessary to end the war.54

49 50 Prunier 1995, pp. 41“54.
Sellstrom and Wohlgemuth 1996.
51 52 United Nations 1994, p. 2. 53 Makinda 1996, p. 562.
Vaccaro 1996, p. 370.
54 De Waal and Omaar 1995, p. 156; Adelman, Suhrke, with Jones 1996; Uvin 1998, p. 62;
Clapham 1998, p. 202; and Human Rights Watch 1999, p. 47.
Angola and Rwanda 71

Habyarimana and his supporters had long resisted calls for power sharing
with Hutu and Tutsi opposition groups. Habyarimana™s political party, the
Mouvement R´ volutionnaire National pour le D´ veloppement (MRND),
e e
was the only legal party in existence from 1975 until 1990, when the regime
was compelled by foreign donors and rising internal opposition to recognize
the existence of new parties. The clique that controlled the ruling MRND
party included Habyarimana™s relatives and closest political allies “ known as
the akazu or “little house” “ who were mostly Hutu and predominantly from
northwestern Rwanda, the president™s home region. As elsewhere in Africa,
control of the state itself was the main avenue of rapid wealth creation,
and members of the akazu enriched themselves by diverting state revenues,
doling out patronage, and administering international development aid.55
Many ordinary Rwandans also depended on the central administration for
employment and perquisites: Local administrators, village chiefs, security
agents, policemen, and local party cadres were all appointed by the central
The desire to hold onto these advantages likely contributed to the grow-
ing opposition within the MRND to the Arusha Accords and to the no-
tion of sharing power and democratizing Rwanda™s political system. By July
1993, however, international donors had lost patience with Habyarimana™s
recalcitrance and, working in combination with the World Bank, informed
the Rwandan president that international funds for his government would
be halted if he did not sign the Accords, including the provisions for
temporary power sharing with opposition groups during a transitional
period leading up to free and fair elections.56 Given that Rwanda was
relying on foreign donors to subsidize at least 70 percent of its public in-
vestment, the threat to cut off aid was effective: Habyarimana signed the
The Arusha Accords devoted little attention to the country™s economic
policies, but the Rwandan government had also been under pressure from
the international ¬nancial institutions to implement structural adjustment
programs. During the late 1980s, Rwanda™s ¬scal condition had deteri-
orated following a sharp decline in world coffee prices (coffee exports
provided more than two-thirds of the country™s foreign revenues) and a
severe drought.57 After resisting structural adjustment for several years,
Habyarimana ¬nally agreed in September 1990 to implement a package
of economic reforms in exchange for ¬nancial assistance from the IMF and
World Bank. The main elements of this package included measures to re-
duce the government budget de¬cit through a combination of improved rev-
enue collection and spending cuts; the liberalization of imports; the elimina-
tion of controls on domestic prices and other regulations impeding private

55 56 Human Rights Watch 1999, p. 124.
Uvin 1998, p. 21.
57 ¨
Sellstrom and Wohlgemuth 1996.
The Peacebuilding Record

enterprise; the elimination of subsidies for coffee producers; and the priva-
tization of state-owned enterprises.58 Although the immediate goal of these
reforms was to restore macroeconomic stability and competitiveness to the
Rwandan economy, a number of international donors argued that these
measures would also help to reduce political tensions within the country
by bringing the Rwandan economic crisis to an end.
In sum, the government of Rwanda committed itself to a dual process of
political and economic liberalization in the early 1990s, largely at the behest
of international agencies and Western states.59 The process was intended, at
least from the perspective of these external actors, to promote political and
economic stability in the country. Yet the faith of the international commu-
nity in the pacifying effects of democratization and marketization turned out
to be tragically misplaced.
In October 1993, two months after the signing of the Arusha Accords,
the United Nations approved the UN Assistance Mission for Rwanda
(UNAMIR) to oversee the implementation of the peace agreement.60 The
principal functions of UNAMIR were to ensure the security of the capital
city of Kigali, to monitor Rwanda™s border with Uganda, to verify compli-
ance with the cease-¬re agreement, and to oversee implementation of military
reforms, including the demobilization of both armies, as well as a number
of humanitarian tasks.61 In 1995, the operation was to shift its principal fo-
cus to preparing for municipal, parliamentary, and presidential elections.62
The entire complement of more than 2,500 UNAMIR military and civilian
personnel was fully deployed by the end of February 1994.63
During this period, however, tensions were mounting between Habyari-
mana and other members of his regime who opposed the peace settlement. Se-
nior ministers in Habyarimana™s own party, along with high-ranking military
of¬cers in the armed forces and Presidential Guard, issued veiled warnings
that attempts to implement the Arusha Accords would have dire results.64
Many of these individuals apparently had direct links to the Coalition pour
la D´ fense de la R´ publique (CDR), a new Hutu-based party that was formed
e e
in 1992 with an openly anti-Tutsi platform.65 In the months that followed the
signing of the Arusha Accords, Hutu hard-liners in the CDR and elsewhere
pushed ahead with plans to organize ordinary Rwandans into “self-defense”
units, nominally directed against the threat of invasion from abroad, but
also apparently in preparation for possible attacks upon Tutsi civilians in-
side the country.66 The CDR™s principal vehicle for spreading its concepts of
Hutu ethnic supremacy to the population of Rwanda was a private radio

58 59 Jones 2001, pp. 62“63.
60 UN Security Council Resolution 872 (October 5, 1993).
61 62 Sellstrom and Wohlgemuth 1996.
United Natioms 1996b, p. 343.
63 64 Adelman, Suhrke, with Jones 1996.
Vaccaro 1996, p. 382.
65 66 Human Rights Watch 1999, pp. 101“103, 126.
Prunier 1995, pp. 128“129.
Angola and Rwanda 73

station “ Radio-T´ l´ vision Libres des Milles Collines (RTLM) “ whose
broadcasts were, according to one Western observer, “the most virulent and
effective incitement to hatred and violence” against Rwanda™s Tutsi popula-
tion.67 During and immediately after the negotiation of the Arusha Accords,
the killing of Tutsi civilians became more frequent. Evidence mounted that
Hutu militia groups associated with the CDR and the ruling party were
responsible for some of these killings.68
On April 6, 1994, President Habyarimana died when his plane crashed
after being struck by a missile near the Kigali airport; within minutes, mem-
bers of the Presidential Guard and CDR militias established roadblocks
throughout the capital.69 Beginning that evening and extending over the
following thirty-six hours, members of the Guard conducted a series of
political assassinations that targeted Hutu and Tutsi politicians, activists,
journalists, and clerics who had voiced support for the planned democratic
transition or criticized the government for its delay in implementing the
Arusha Accords.70 Members of local militias and ordinary Hutu peasants
soon began killing Tutsi civilians throughout the country. Over the next
three months, somewhere between ¬ve hundred thousand and one million
Rwandan Tutsis were murdered by their Hutu compatriots, and an estimated
four million (roughly half of Rwanda™s prewar population) were dis-
placed from their homes in “one of history™s most intensely violent
Efforts to reinforce UNAMIR failed due to lack of interest from the in-
ternational community in deploying a larger force to Rwanda.72 Given the
inability of the existing UNAMIR force to carry out its responsibilities in
the midst of intense violence, the UN ordered the bulk of the force to leave
Kigali, leaving behind a token contingent of 450 soldiers.73 It was therefore

67 Chege 1996/97, p. 34.
68 ¨
Sellstrom and Wohlgemuth 1996; and Adelman, Suhrke, with Jones 1996.
69 Vaccaro 1996, p. 373. The identity of those who launched the missile against Habyarimana™s
plane remains unknown (Human Rights Watch 1999, pp. 181“185). However, the speed
with which the Presidential Guard and militias set up roadblocks following Habyarimana™s
assassination suggests that they may have had advanced warning.
70 Prunier 1995, pp. 230“231 and 242“243. See also the report of the UN Commission on
Human Rights (UN document E/CN.4/1995/7, June 28, 1994).
71 Jones 2001, pp. 1“2 and 43“44.
72 The deciding factor appears to have been the lack of support from the United States for
expanding the size and mandate of UNAMIR. See Burkhalter 1994/95.
73 Jones 1995, p. 230. The UN commander in Rwanda at the time, Major General Rom´ o e
Dallaire of Canada, explained in an interview why his force of 2,600 blue-helmeted troops
were unable to perform their mandate under these circumstances: “[T]hey [UNAMIR
personnel] couldn™t move all around. They couldn™t do their job. They did not have the
self-protection. They did not have the ammunition. They did not have the weapons” (CBC
1995). Dallaire has argued that the genocide could have been stopped if the UN contingent
had been reinforced with well-equipped and well-trained troops in the early days of the
killing (Buckley 1998; and Straus 1998).
The Peacebuilding Record

left to the RPF, which had resumed military operations on April 8, to stop
the massacres through force of arms. By the time the RPF gained control of
most of the country™s territory, however, an estimated 80 percent of Rwanda™s
resident Tutsi population had been killed.74
The mass murder of Rwanda™s Tutsis between April and July 1994
has been labeled a genocide because of the magnitude of the killings
and because a speci¬c ethnic group was targeted for extermination. Most
observers of Rwandan politics have concluded that the genocide was
carefully planned and orchestrated by a relatively small group of Hutus
belonging to the regime™s political, military, and economic elite, including
members of Habyarimana™s own party, who had created and were in control
of the network of local militias that conducted the initial wave of killings
(although the genocide became less organized as ordinary Hutu villagers,
spurred on by government-controlled radio stations, joined in the frenzy).75
The genocide effectively ended the UN peacebuilding mission in Rwanda,
as the international community™s attention shifted to the new refugee cri-
sis caused by hundreds of thousands of Hutus ¬‚eeing into neighboring
countries. Since that time, the RPF has remained in control of Rwanda™s
government and territory, and the ¬ghting has largely subsided, although
there is still disagreement about exactly what happened, who was respon-
sible, and why the international community failed to foresee and avert the
Why did Hutu extremists plan and conduct the genocide? The evidence
suggests that the mass killing of Tutsis was a last-ditch effort to block im-
plementation of the Arusha Accords.76 Plans for political liberalization,
including the transitional coalition government and democratic elections,
would have challenged the dominance of the Hutu clique that surrounded
Habyarimana and controlled the military. As Howard Adelman and Astrid
Suhrke put it, the accords “represented a frontal attack on the power base
erected by the Habyarimana regime during 20 years of rule “ a denial of au-
thoritarian rule, of ˜Hutu power,™ and especially Northwestern-based Hutu
power which was the regional constituency and political backbone of the
regime.”77 Hard-line members of Habyarimana™s party and the CDR had
openly displayed their opposition to the Arusha Accords by boycotting

74 Prunier 1995, p. 265.
75 ¨
Prunier 1995, pp. 239“248; Sellstrom and Wohlgemuth 1996; and Longman 1997, p. 300.
For ¬rsthand reports of the premeditated character of the genocide, see the testimonies cited
in Prunier 1995, p. 242, n. 51. See also the report of the UN Commission on Human Rights
(UN document E/CN.4/1995/70, November 11, 1994), which refers to “reliable testimony”
that orders to carry out the massacres were issued by the central government via local political
and administrative authorities.
76 Burkhalter 1994/95, p. 44; Jones 1995, p. 227; Prunier 1995, pp. 241“242; Makinda 1996,
p. 266; Sellstrom and Wohlgemuth 1996; Longman 1997, pp. 287“300; and Jones 2001.
77 Adelman, Suhrke, with Jones 1996.
Angola and Rwanda 75

cabinet meetings and organizing violent demonstrations, and early plans
for a “¬nal solution to the ethnic problem” had circulated in extremist cir-
cles in late 1992, while negotiations at Arusha were still ongoing.78 Faced
with the prospect of sharing power with their Tutsi adversaries, this group
apparently decided to assassinate its political enemies, including Hutu mod-
erates within the government, and then physically eliminate the Tutsi pop-
ulation of Rwanda. The enactment of these plans “marked the culmination
of a program of government-sponsored violence and intimidation intended
to frustrate efforts to change the structures of power and democratize the
political system.”79 Political liberalization, in other words, not only failed
to reconcile the warring parties in Rwanda but also apparently served as a
catalyst for the genocide by threatening Hutu elements with the prospect of
losing power.
The tragedy cannot be explained entirely as a response to the interna-
tionally sponsored democratization plan; the genocide clearly had multiple
and complex causes. The growing power of the RPF relative to Rwandan
government forces, for example, may have convinced some members of the
Hutu-dominated regime of the urgency of solving the “Tutsi problem” once
and for all.80 The regime also faced mounting internal political pressure
from disgruntled Hutus, particularly those who were not from the privileged
northwestern region of the country. Less obviously, the causes of the geno-
cide could be traced back to the colonial history of Rwanda, which helped to
create the very ethnic divisions and resentments that played themselves out in
April 1994; and to the perverse tradition of tit-for-tat civilian massacres by
Hutus and Tutsis throughout the postindependence period, which effectively
established the “strategy” of genocide as a conceivable option for members
of the embattled Habyarimana clique. Any of these factors may have con-
tributed to the conditions that produced the genocide, and should ¬gure in
a comprehensive explanation of the genocide.
However, our more limited task in these case studies is to evaluate the ef-
fects of liberalization on the prospects for stable and lasting peace in peace-
building host states. Although the internationally sponsored plans for de-
mocratization and power sharing do not provide a complete explanation for
the events of April 1994, they do appear to have provoked extremist mem-
bers of the regime to act quickly “ both in preparing for the massacres and
in initiating the genocide immediately after Habyarimana™s death “ in order
to prevent the Arusha Accords from being implemented. At the very least,
then, the effort to move Rwanda in the direction of democracy did not have

78 Prunier 1995, pp. 160“161, 166, 173, and 200.
79 Longman 1997, p. 287. Makinda (1996, p. 556) reaches a similar conclusion, arguing that
the Hutu leadership was “so strongly opposed to the competitive political system that they
tried to sabotage every effort to make it work.” See also Clapham 1998.
80 Destexhe 1995, p. 46.
The Peacebuilding Record

the pacifying in¬‚uence that international peacebuilders had hoped for and
apparently expected. The international community had presented the plan
for power sharing followed by democratic elections as a means of resolving
Rwanda™s civil war, but attempts to foster peace and stability in Rwanda
by promoting political liberalization ultimately back¬red, and in the worst
possible way. As one student of Rwandan politics notes: “The push for de-
mocratization had unforeseen and ultimately negative effects, and it was a
key tactical error.”81
Other dimensions of the Rwandan case also cast doubt on the Wilsonian
assumptions of peacebuilding. First, although some commentators have ar-
gued that greater political liberalization “ such as more freedom of association
and expanded press freedom “ might have helped to avert the genocide, there
is little reason to believe this is true. Many international development agen-
cies that were involved in Rwanda before the genocide worked to increase
the number of nongovernmental, voluntary organizations in the country on
the assumption that the latter would enhance the “social trust” and “civic
engagement” that are the pillars of a democratic, pluralistic society.82 Yet
this assumption depended on these local organizations™ choosing to embrace
democracy and human rights. As it turned out, however, many local NGOs
in Rwanda either were apolitical and thus disconnected from the events
of April 1994, or were virulently anti-Tutsi and antidemocratic.83 Efforts
to promote a more active “associational life” in Rwanda, in other words,
did not appear to further the goals of either democracy or peace because
many indigenous NGOs subscribed to the dominant bigotry of the larger
Similarly, the partial liberalization of Rwanda™s popular media in the early
1990s may have made the genocide more, not less, likely to occur: Although
the Hutu regime monopolized the radio, a vibrant but irresponsible press
came into existence after 1990. This press was highly polarized between
supporters and opponents of the Habyarimana regime, with some of the new
publications advocating the destruction of Rwanda™s Tutsi population and
others openly expressing sympathy for the RPF™s war effort.84 Jack Snyder
and Karen Ballentine have argued that in¬‚ammatory publications on both
sides of the con¬‚ict worked against the cause of peace: Some pro-government
newspapers worked in tandem with RTLM to incite the Hutu population to
violence against Tutsis, while virulent attacks on the Habyarimana regime in
some antigovernment publications may have reinforced the Hutu extremists™
determination not to accept the power-sharing provisions of the Arusha
Accords.85 Thus, the liberalization of the media did not appear to have the
moderating effects that some international agencies predicted.

81 82 Uvin 1998, chap. 8. 83 Ibid.
Jones 2001, p. 164.
84 85 Snyder and Ballentine 1996, pp. 30“34.
Prunier 1995, pp. 131“132.
Angola and Rwanda 77

A further aspect of the Rwandan case that raises doubts about Wilso-
nianism is the seemingly deleterious impact of economic liberalization on
the Rwandan peace process in the years preceding the genocide. As noted,
the government of Rwanda had committed itself to a structural adjustment
plan in September 1990 at the insistence of the international ¬nancial in-
stitutions, a plan that included government spending cuts, reductions in
price subsidies, and deregulation of the domestic economy. These policies
may have contributed to the worsening of economic conditions in the early
1990s, and thereby made the Rwandan population more susceptible to the
government™s hate propaganda. Some commentators put this case in very
strong terms. Michel Chossudovsky, for example, argues that “the imposi-
tion of sweeping macro-economic reforms by the Bretton Woods institutions
exacerbated simmering ethnic tensions and accelerated the process of politi-
cal collapse.”86 However, the precise impact of structural adjustment policies
on economic and political conditions in Rwanda is dif¬cult to discern, given
that the introduction of these policies coincided with the collapse in interna-
tional coffee prices, a prolonged drought, and the RPF invasion in October
1990 “ the combined effects of which resulted in an absolute decline in the
country™s GDP in every year from 1990 to 1994.87 Nevertheless, speci¬c
elements of the structural adjustment program did apparently contribute to
the impoverishment of many Rwandans in the early 1990s: Higher fees for
health and education, for example, added to the already heavy burdens of
Rwanda™s poor, while a freeze on public wages, combined with successive
currency devaluations, decreased the real income of many middle-class
Whether these conditions increased the willingness of ordinary Rwandans
to participate in mass murder remains a matter of debate, but it does appear
that the economic crisis placed additional stress on the social and political
fabric of the country in the lead-up to the genocide.89 Indeed, whatever
connection may exist between the economic liberalization policies and the
genocide, these policies clearly did not enhance Rwanda™s political stability
during the Arusha peace process, and they seem to have contributed to the
impoverishment of many Rwandans at a time of acute political and social
tension. Nor did the international community, including the World Bank
and IMF, consider the potentially explosive political consequences of these
policies when they designed the structural adjustment program.90

86 Chossudovsky 1997, p. 111.
87 Per capital GDP fell by almost 40% from 1989 ($330) to 1993 ($200), according to ¬gures
cited in Sellstrom and Wohlgemuth 1996.
88 Ibid.
89 Prunier (1995, pp. 284 and 243) points out, moreover, that there was an element of material
interest in the killings: Most of the members of the militias were poor, and looting of victims™
possessions was widespread.
90 Uvin 1998.
The Peacebuilding Record

All told, the peacebuilding experience in Rwanda challenges the liberal
peace thesis in a number of ways. Efforts to democratize Rwanda facilitated
the rise of political parties that were “masks for ethnic groups that organized
murderous militias”91 and provoked Hutu extremists to plan and launch a
genocidal attack on the country™s Tutsi population; the liberalization of the
media and civil society organizations did not produce political moderation
and may have simply offered extremist groups a means of organizing and
conveying their in¬‚ammatory messages; and market-oriented economic re-
forms seem to have worsened, not ameliorated, the climate of insecurity
in Rwanda that the perpetrators of the genocide were able to exploit. The
precise degree to which the liberalization process may have ultimately con-
tributed to the resurgence of violence in Rwanda has yet to be determined,
and responsibility for the horrors of April 1994 must rest with those who
perpetrated the genocide. But it seems safe to conclude, at the very least, that
international efforts to transform Rwanda into a market democracy did not
advance the goal of establishing a stable and lasting peace.

The two missions investigated in this chapter indicate that the liberaliza-
tion process in Angola and Rwanda did not have the pacifying effects that
peacebuilders had anticipated. Indeed, the cases also offer circumstantial ev-
idence that the liberalization process may have worked against the goal of
creating a stable and lasting peace in these two states. Nevertheless, Angola
and Rwanda represent only two of eleven operations that we will examine in
depth, and in many respects these two missions offer the worst outcomes of
internationally sponsored liberalization efforts in war-shattered states dur-
ing the 1990s. As we investigate more cases in the following chapters, our
conclusions about the Wilsonian assumptions of peacebuilding will therefore
need to be revised and elaborated.

91 Kaplan 1997.

Cambodia and Liberia
Democracy Diverted

In contrast to Angola and Rwanda, where the democratization process
might well have served as a catalyst for renewed con¬‚ict and the collapse
of peacebuilding efforts, the results of postcon¬‚ict democratization efforts
in Cambodia and Liberia were widely viewed as positive. Both countries
held largely free and fair elections under international supervision, and in
both cases peacebuilding agencies declared their efforts a “success” and ter-
minated their operations shortly after the elections. At ¬rst glance, then,
Cambodia and Liberia seem to corroborate the Wilsonian hypothesis that
liberalization fosters peace in war-shattered states. Closer examination, how-
ever, raises serious doubts about the pacifying effects of democratization in
these two cases. Even though both countries experienced a period of relative
political stability following the departure of international peacebuilders, the
adoption of democracy itself cannot be credited because the newly elected
governments of both countries immediately backed away from their com-
mitments to democracy and reverted to more autocratic forms of rule; and
one country, Liberia, ultimately slipped back into war.
The Wilsonian hypothesis holds that violent con¬‚ict can be transformed
into peaceful political competition when groups are allowed to pursue their
interests freely through democratic political institutions. While this may help
to explain the initial success of elections in Cambodia and Liberia, it does
not explain the ensuing period, during which there was little real freedom to
challenge the policies of governing cliques, and the behavior of both coun-
tries™ elected leaders “ Hun Sen in Cambodia and Charles Taylor in Liberia “
was more despotic than democratic. Their regimes do not lend support to
the proposition that free political competition fosters peace.

Cambodia enjoyed little peace in the years after it gained its independence
from France in 1953. During the 1960s, Vietnamese communist guerrillas
The Peacebuilding Record

used Cambodian territory to move supplies and establish bases in their ¬ght
against the forces of South Vietnam and the United States, and in response,
American forces began a secret bombing campaign of Cambodia in 1969.1
Soon after, American and South Vietnamese troops crossed the border into
Cambodia to destroy communist bases and supply lines, a campaign that
pushed the guerrillas and the ¬ghting deeper into Cambodian territory. As
the Vietnam war drew to a close in April 1975, Cambodian communist guer-
rillas known as the Khmer Rouge captured the country™s capital of Phnom
Penh. The leader of the Khmer Rouge, Pol Pot, immediately ordered the city
emptied of all of its inhabitants, many of whom were sent to labor camps
in the countryside for “reeducation.” Over the following three years and
eight months, the Pol Pot government orchestrated a perverse reengineering
of the society during which an estimated 1.7 million Cambodians “ or ap-
proximately 20 percent of the population “ died as a result of forced labor,
torture, execution, malnutrition, or disease.
In December 1978, Vietnamese troops invaded Cambodia and overthrew
the Khmer Rouge regime, driving Pol Pot and his supporters into jungle areas
close to the border with Thailand. Tens of thousands of Cambodian civil-
ians also ¬‚ed over the Thai border and found their way to sprawling refugee
camps. Meanwhile, the Vietnamese established a new Cambodian govern-
ment, which continued the military campaign against the Khmer Rouge
with the assistance of Vietnamese combat troops. With Chinese equipment
and funding, however, the Khmer Rouge was able to defend its jungle re-
doubts and wage a guerrilla war against the Phnom Penh government and its
Vietnamese backers for most of the 1980s. Two smaller Cambodian oppo-
sition groups also fought alongside the Khmer Rouge. The Front Uni Pour
Un Cambodge Ind´ pendant, Neutre, Paci¬que et Coop´ ratif (FUNCINPEC)
e e
was led by Prince Norodom Sihanouk, who had been Cambodia™s head of
state from its independence until he was deposed in 1970; the second group,
the Khmer People™s National Liberation Front (KPNLF), was led by former
prime minister Son Sann. Both of these noncommunist factions received sup-
port from Thailand and Western countries, including the United States and
The rise of Mikhail Gorbachev to power in the Soviet Union in 1985 initi-
ated a series of events that led eventually to the negotiated settlement of the
Cambodian con¬‚ict. Vietnam had relied on the Soviet Union for material
support since the relationship between Vietnam and China had cooled in
the late 1970s. When Gorbachev indicated that his government intended to
reduce its international spending and seek peaceful solutions to the regional
con¬‚icts in which it was involved, Vietnam™s policy toward Cambodia be-
gan to shift. In late 1985, the Hanoi government informed its Southeast
Asian neighbors that Vietnam might be willing to withdraw its troops from

1 Shawcross 1994, p. 7.
Cambodia and Liberia 81

Cambodia. In 1986, the member states of the Association of Southeast Asian
Nations (ASEAN) and Vietnam began to discuss possible compromise so-
lutions to the Cambodian con¬‚ict. One year later, Prince Sihanouk met for
the ¬rst time with the leader of the Phnom Penh regime, Hun Sen. After
several rounds of negotiations, the four Cambodian parties (the govern-
ment, FUNCINPEC, the KPNLF, and the Khmer Rouge) ¬nally agreed on
a framework for the comprehensive settlement of the Cambodian con¬‚ict
in September 1990.2 The framework had been devised by the ¬ve perma-
nent members of the Security Council “ China, France, Britain, the Soviet
Union, and the United States “ who met again in November to draft a com-
prehensive peace agreement, building upon the framework agreement.3 The
document ultimately produced by these states was, after several months of
further negotiation, formally accepted by the Cambodian parties at a peace
conference in Paris on October 23, 1991.4
The agreement set out a detailed plan for transforming Cambodia into a
peaceful liberal democracy. Factional armies would be disarmed and de-
mobilized, refugees returned to their homes, political prisoners released.
A Supreme National Council consisting of representatives from all four
Cambodian parties would act as the country™s sovereign authority until mul-
tiparty elections could be held for a constituent assembly. This assembly,
once elected, would draft a new constitution and then transform itself into
the country™s ¬rst postwar government. The peace agreement set out the
main principles for a new constitution: Cambodia would follow “a sys-
tem of liberal democracy, on the basis of pluralism,” including “periodic
and genuine elections” by secret ballot and universal suffrage, civil liberties
enshrined in a declaration of fundamental rights, and an independent ju-
diciary empowered to enforce these rights.5 In short, the formula that the
“permanent ¬ve” members of the UN Security Council presented to the
Cambodian parties, and which these parties accepted with slight modi¬ca-
tions, re¬‚ected the Wilsonian assumption that transforming the Cambodian
state into a liberal democracy would facilitate the transition from civil war
to lasting peace. Indeed, the Security Council explicitly justi¬ed its support
for the Cambodian peace accords on the grounds that “free and fair elections

2 For a description of the negotiations that led up to the 1990 framework agreement, see
Heininger 1994, pp. 12“22.
3 Indonesia was also present at the meeting, in virtue of its position as the co-chair (along with
France) of the most recent peace conference on Cambodia.
4 The text of the agreement is reproduced in UN document A/46/608-S/23177 (October 30,
1991), annex.
5 Ibid., section II, “Agreement on a Comprehensive Political Settlement of the Cambodia Con-
¬‚ict,” annex 5, “Principles for a New Constitution for Cambodia.” As Michael Doyle (1996b,
p. 500) notes, the peace agreement “speci¬ed all the elements necessary for a constitutional
democracy” in Cambodia, with free and fair elections as the “hallmark and linchpin” of the
The Peacebuilding Record

are essential to produce a just and durable settlement to the Cambodia
A new United Nations ¬eld operation “ the UN Transitional Authority
in Cambodia (UNTAC) “ was created to oversee implementation of the ac-
cord. UNTAC™s extensive duties included supervising Cambodia™s civilian
police, monitoring the cease-¬re and the demobilization of factional armies,
investigating human rights complaints, repatriating refugees, and coordi-
nating an international campaign to reconstruct Cambodia™s war-damaged
infrastructure, as well as organizing and conducting national elections.7 The
operation was also given a wide-ranging mandate to supervise the conduct of
Cambodia™s existing governmental administration in order to “ensure a neu-
tral political environment conducive to free and fair general elections.” This
meant, in practice, that UN of¬cials were to be involved in the day-to-day
operations of individual government ministries.
“Not since the colonial era and the post“World War II Allied occupations
of Germany and Japan,” writes Michael Doyle, “had a foreign presence held
so much formal administrative jurisdiction over the civilian functions of an
independent country.”8 James Schear elaborates this point:

By the latter stages of the operation, UNTAC personnel could be found doing such
things as probing into the country™s penal code, investigating its defense procurement
decisions, vetting editorials in state-run media, reviewing regulations on national her-
itage preservation, scrutinizing admissions policies at public educational institutions,
monitoring passport and visa procedures, managing monetary and ¬scal decisions,
and delving into a host of other civil administrative activities.9

Not only did the peace accords explicitly prescribe the remolding of the Cam-
bodian state into a liberal democracy, but they also empowered international
civil servants, working under the auspices of the United Nations, to oversee
and expedite these reforms by occupying positions within the Cambodian
government ministries themselves.
In the realm of economic policy, the accords carefully avoided endors-
ing any particular economic model for postwar Cambodia, and explicitly
warned international donors against interfering in the country™s economic
policymaking process: “No attempt shall be made to impose a develop-
ment strategy on Cambodia from any outside source.”10 In practice, how-
ever, the international ¬nancial institutions encouraged Cambodia to adopt

6 UN Security Council Resolution 745 (February 28, 1992).
7 The operation™s mandate is set out in detail in UN document S/23613 (February 19, 1992)
and its addendum, UN document S/23613/Add.1 (February 26, 1992).
8 9 Schear 1996, p. 158.
Doyle 1995b, p. 13.
10 UN document A/46/608-S/23177 (October 30, 1991), annex, part IV, “Declaration on the
Rehabilitation and Reconstruction of Cambodia,” para 2.
Cambodia and Liberia 83

market-oriented economic reforms.11 In mid-1992, the government of Cam-
bodia reached an agreement with the IMF to implement an economic stabi-
lization program that involved deep cuts in capital and social spending, along
with wage and salary limits for public employees, in order to restore balance
to the government™s ¬nances.12 This was followed, in October 1993, by the
negotiation of an IMF structural adjustment loan to Cambodia that com-
mitted the government to further market-oriented reforms, some of which
were implemented “under the tutelage of UNTAC of¬cials” working within
the Cambodian bureaucracy.13
Although external peacebuilding agencies played a central role in promot-
ing and effecting Cambodia™s transition to a market economy, this transition
does not appear to have been “imposed” on the Cambodian parties. After
all, these parties opted to pursue market-oriented economic reforms rather
than an alternative development strategy. But there is little doubt that the
principal peacebuilding agencies in Cambodia were predisposed in favor of
liberal economic policies, and that they communicated this predisposition
to the Cambodian government through the offer of concessional loans and
technical advice.
These observations lead to the following conclusion: The peacebuilding
operation in Cambodia, like the others examined in this volume, sought
to transform Cambodia into a particular kind of society “ a liberal market
democracy “ on the assumption that doing so would create the most propi-
tious conditions for a stable and lasting peace. As Joakim Ojendal argues,
“one of the intrinsic aspects of the operation certainly was the establishment
of a liberal order,” which in practice meant the promotion of democratic elec-
tions, civil liberties, and market-oriented economic reforms in Cambodia.14
The mission got off to a rocky start. Although UNTAC of¬cially be-
came operational on March 15, 1992, many of its of¬ces were not fully
staffed until December, just six months before the scheduled elections in
May 1993.15 The mission™s military component was not ¬elded in time to
begin the process of disarmament and demobilization scheduled for June
1992.16 Necessary vehicles, prefabricated housing, of¬ce and communica-
tions equipment, and other items were also slow to arrive in Cambodia.17

11 At a meeting of international donors on June 22, 1992, for example, the international ¬nan-
cial institutions “stressed the importance of market-based reforms in Cambodia to increase
the output in major sectors of the economy” (UN doc A/47/285-S/24183 of June 24, 1992,
12 Irvin 1993, pp. 128“132.
13 Ibid., p. 132. Indeed, UNTAC took over responsibility for ¬nancial and macroeconomic
operations of the Cambodian government in March 1992 (United Nations 1996b, p. 191).
14 15 Prasso 1995, p. 39. 16 USGAO 1993, p. 41.
Ojendal 1996, p. 194.
17 United Nations 1995, p. 16. Many reasons have been cited for UNTAC™s delayed deploy-
ments, including inadequate planning by UN headquarters (Heininger 1994, p. 85; and
Jennar 1994, p. 153), the organization™s cumbersome recruitment and procurement system
The Peacebuilding Record

While the UN scrambled to assemble teams of military and civilian per-
sonnel and transport them to Cambodia, the operation suffered another
serious setback: The Khmer Rouge simply refused to prepare its forces for
demobilization as required by the peace agreement.18 The cantonment of
the factional armies was to have been completed by the end of July, but the
three other Cambodian parties were reluctant to disarm their forces in the
face of Khmer Rouge recalcitrance. As a result, by mid-November, only some
55,000 troops had reported to the cantonment sites, most of whom appeared
to be untrained teenagers with antiquated weapons, while superior forces
and caches of weapons remained in the ¬eld.19 After several unsuccessful at-
tempts at convincing the Khmer Rouge to comply with the accords, UNTAC
abandoned its demobilization effort and allowed the soldiers who had al-
ready been cantoned to return to their respective armies on “agricultural
On November 30, the Security Council imposed a selective trade embargo
on the Khmer Rouge and ordered UNTAC to proceed with the planned elec-
tions with or without the cooperation of the Khmer Rouge.21 The operation™s
military component was subsequently redeployed “to protect voter registra-
tion and, subsequently, the electoral and polling processes, particularly in
remote or insecure areas.”22 Thus, after the failed demobilization effort,
UNTAC shifted its attention to the goal of ensuring a peaceful environment
for free and fair elections. But as the May 1993 elections drew closer, the
security situation in Cambodia deteriorated.
The opening of the campaign period in March unleashed a wave of po-
litical violence, which intensi¬ed as election day neared.23 Opposition po-
litical of¬ces were “attacked, ransacked, and burned, and party members
were beaten, kidnapped and killed.”24 In the ten weeks leading up to the

(USGAO 1993, pp. 35“41), the sheer size and complexity of the Cambodia operation (United
Nations 1995, p. 16), and the fact that UNTAC found itself in competition for scarce re-
sources with the UN operation in the former Yugoslavia (Schear 1996, p. 152).
18 To justify their noncompliance with the accords, Khmer Rouge of¬cials asserted that “a great
number” of Vietnamese military personnel remained in Cambodia in contravention of the
accords (United Nations 1995, p. 17; and Schear 1996, p. 157). In response, UNTAC con-
ducted an extensive investigation, which revealed no evidence of any organized Vietnamese
military presence in Cambodia.
19 20 Doyle 1995b, p. 35.
Berdal and Leifer 1996, p. 43.
21 The trade embargo included stopping the supply of petroleum products “to areas occupied
by any Cambodian party not complying with the military provisions” of the peace accords,
and a moratorium on the export of logs from Cambodia. Regarding the elections, the Council
determined that UNTAC should proceed with preparations for free and fair elections “in
all areas of Cambodia to which it has full and free access as at 31 January 1993.” Security
Council Resolution 792 (November 30, 1992), paras. 5, 10, and 13.
22 Boutros-Ghali™s third progress report on UNTAC, UN document S/25154 (January 25, 1993),
para. 41.
23 24 Doyle 1995b, p. 56.
Chopra 1994, p. 27.
Cambodia and Liberia 85

vote, political violence resulted in a reported 176 deaths, 316 injuries, and
67 abductions.25 Two groups were thought to be primarily responsible for
these attacks: Agents of Hun Sen™s incumbent regime apparently perpetrated
much of the violence against supporters of the two noncommunist oppo-
sition parties, FUNCINPEC and the KPNLF; while the Khmer Rouge was
held responsible for killing dozens of ethnic Vietnamese civilians and for sev-
eral attacks against UNTAC of¬cials in the Cambodian countryside, which
was apparently intended to disrupt the polling.26 Hostile intimidation also
interfered with the registration of eligible voters at refugee camps along the
Cambodian border.27 To many observers, the election-related violence rep-
resented a “looming disaster” for UNTAC.28
Nevertheless, preparations for the elections continued, and for reasons
that are still unknown, the Khmer Rouge did not ultimately carry out
its threat to launch a large-scale military offensive when polls opened on
May 29.29 The relative peacefulness of Cambodia during the two days of
voting came as an “astonishing, welcome surprise” after weeks of mount-
ing violence,30 with nearly 90 percent of registered voters casting ballots.31
FUNCINPEC, under the leadership of Prince Sihanouk™s son, Norodom
Ranariddh, took more than 45 percent of the votes; Hun Sen™s Cambodian
People™s Party (CPP) won roughly 38 percent; and the KPNLF came in a
distant third with 3.8 percent.32
However, Hun Sen™s reluctance to accept his loss to Ranariddh led to
further confrontations in the days after the vote.33 In an effort to create a
stable postelection government, Prince Sihanouk announced on June 16 the
formation of an Interim Joint Administration, with himself as head of state
and both Ranariddh and Hun Sen as co“prime ministers. Ministerial posts
were to be split evenly between the two leading parties.
From the vantage point of late 1993, the effects of political liberalization
on the prospects for lasting peace in Cambodia appeared to be mixed. On
the one hand, democratic elections seemed to provide a formula for three
of the four formerly warring parties to shift their political disputes from the
battle¬eld to the ballot box. In late 1992, when it became apparent that

25 Plunkett 1994, p. 71.
26 Boutros-Ghali™s fourth progress report on UNTAC, UN document S/25719 (May 3, 1993),
paras. 4“5, and United Nations 1995, pp. 41“44.
27 28 Doyle 1995b, p. 51.
Hampson 1996, p. 198.
29 At a public lecture at Yale University in 1996, I asked the former UN administrator for
Cambodia, Yasushi Akashi, why the Khmer Rouge did not seek to disrupt the elections, and
Akashi said that he did not know.
30 31 United Nations 1995, p. 46.
Shawcross 1994, p. 21.
32 Ibid. The remaining vote was shared among seventeen other political parties. The KPNLF
contested the election under the banner of its political wing, the Buddhist Liberal Democratic
33 Will 1993, pp. 399“400.
The Peacebuilding Record

important elements of the peace accords, including the planned demobi-
lization, would not be ful¬lled, the UN had redirected its efforts toward
organizing free and fair elections. This gamble had apparently paid off when
the new coalition government was formed, and Cambodia seemed to enter a
new period of political stability.34 On the other hand, the process of political
liberalization had also served to exacerbate tensions among the Cambodian
parties in the period leading up to the elections “ tensions which, according
to many observers at the time, had threatened to destroy the fragile truce
between FUNCINPEC and the CPP. James Schear underscores this point:
“The injection of political pluralism into the country, though welcome in
many respects, opened the door to an upsurge in violence that UNTAC was
ill prepared to handle.”35 Although UNTAC helped to stabilize the situa-
tion “ for instance, by posting guards at party of¬ces and campaign rallies
to deter attacks “ the fact remains that election-related violence could have
easily escalated into renewed civil war.
By the end of 1993, after the Cambodian parties had formally approved
the constitution and the new coalition government had been formed, the UN
Security Council declared UNTAC a “success” and ordered the withdrawal
of the operation.36 Relations between the coalition partners “ Ranariddh™s
FUNCINPEC and Hun Sen™s CPP “ were tense but peaceful. The Khmer
Rouge, though still an active insurgency, was beset by internal divisions
and weakened by a steady stream of defections. Most analysts agreed that
Cambodia was more at peace than it had been at any time since the early
But, in fact, the situation was fundamentally unstable. The quick depar-
ture of the UN mission and the precarious power-sharing arrangement “set
the course for the inevitable overthrow of the democratic process four years
later,” as UNTAC™s military commander later publicly acknowledged.38 In
theory, Ranariddh was “¬rst prime minister” because his party had won
more votes in the election, and Hun Sen was “second prime minister.” In
practice, however, Hun Sen continued to control the apparatus of the state
and the largest armed force in the country, and was unwilling to cooperate
with Ranariddh.
Anticipation of new elections scheduled for 1998 soon led to a deterio-
ration in relations between Cambodia™s two co“prime ministers. Ranariddh
was determined to bolster the military forces loyal to FUNCINPEC before

34 This was certainly the opinion of the UN™s point man in Cambodia, Yasushi Akashi, who
declared that the elections had provided “an important basis for eventually consolidating
peace and preparing for national reconciliation” in Cambodia (Akashi 1994, p. 258).
35 Schear 1996, p. 174.
36 “Statement of the President of the Security Council Concerning the Successful Completion
of the Mandate of UNTAC,” UN document S/26531 (October 5, 1993).
37 Mabbett and Chandler 1995, p. 257.
38 Sanderson 2001, p. 165. See also Sanderson and Maley 1998.
Cambodia and Liberia 87

the 1998 elections in order to deter CPP-sponsored violence and intimida-
tion, which he and his supporters had faced during the 1993 campaign.
Hun Sen was similarly determined to weaken FUNCINPEC before the
1998 elections in order to avoid another embarrassing loss at the polls.39
By early 1996, political violence was once again on the rise.40 Later that
year, Ranariddh began to pursue contacts with the remaining leaders of
the Khmer Rouge in the hopes of enlisting their support against the CPP.41
Hun Sen saw these moves as a threat and responded by launching a mil-
itary campaign against Ranariddh™s supporters in July 1997, which led
to renewed ¬ghting between forces loyal to the two governing parties.42
Although the bulk of Ranariddh™s forces were defeated in two days, ¬ghting
between the CPP and individual FUNCINPEC units continued for several
Ranariddh was outside Cambodia at the time and escaped arrest and pos-
sible assassination, but approximately forty of his supporters in Phnom Penh
were executed and hundreds arrested.44 According to the United Nations,
moreover, at least ¬fty opponents of Hun Sen were killed in the subse-
quent eight months from August 1997 to March 1998.45 The underlying
cause of this renewed violence was the ongoing rivalry and distrust between
FUNCINPEC and the CPP, but the approaching elections appear to have
been the major precipitating factor that brought this rivalry to a head.46
As one Western journalist, who was in Cambodia throughout this period,
commented: “The only reason that there was a coup [in 1997] was that Hun
Sen saw himself as being politically out¬‚anked and realized that he would
have lost the election [if he had not taken action].”47
This is not to say that the democratization process was solely responsible
for the resurgence of political violence in Cambodia. Nevertheless, political
competition and violence among the Cambodian parties did intensify in the
period leading up to both the 1993 and the 1998 national elections, sparking
new ¬ghting between the CPP and FUNCINPEC in the latter period. While
it is true that the 1993 vote permitted FUNCINPEC and the CPP to estab-
lish a new relationship as partners in a governing coalition, this partnership
was untenable, and the political violence of the 1993 campaign period fore-
shadowed the more serious ¬ghting that erupted in the lead-up to the 1998
elections, when both parties attempted to outmaneuver the other in advance
of the vote. UNTAC, it seems, took advantage of the moment of relative

39 40 Ott 1997, p. 434.
Karniol 1997; and Jeldres 1997.
41 42 Mydans 1997b.
Ott 1997, p. 435; Karniol 1997; and Roberts 2001, p. 140.
43 44 Ott 1997, p. 435. 45 Reuters 1998.
Mydans 1997a; Associated Press 1998.
46 Doyle 2001, p. 92.
47 Nate Thayer, in a personal communication to the author on May 15, 1998. Others have
reached a similar conclusion, including Richard Solomon, former U.S. assistant secretary of
state for East Asia (Newshour 1997 [July 14]).
The Peacebuilding Record

calm immediately after the 1993 election to declare victory and withdraw
from the country before the chaos returned.48
Although Cambodia has experienced greater political stability since the
signing of the Paris Peace Accords, the traditional sources of violent con-
¬‚ict in the country “ including the recurring pattern of authoritarianism and
rebellion “ seem not to have been fundamentally altered by the peacebuild-
ing mission. The country™s internal politics have historically been marked
by distrust and lack of cooperation among charismatic leaders, who have
generally sought absolute power and engendered similarly intolerant and
extreme forms of opposition. As David Roberts writes: “No Khmer leader
since independence, whether regal, communist, republican or former peas-
ant, has accepted without resistance a challenge to the absolutism of their
authority.”49 Longtime Cambodia observer Sorpong Peou concurs: “From
Sihanouk to Hun Sen, the state leadership always seemed interested in pre-
serving or enhancing its hegemonic status quo, even if it later grew vulnera-
ble to challenges from within and/or from without the state. As these leaders
tightened their grips on power, they succeeded in turning foes into friends,
but later risked turning friends into foes.”50
Both Roberts and Peou express doubt that the post-1993 “democratiza-
tion” substantially changed this pathological characteristic of Cambodian
politics. While the country was more democratic after 1993 than in the past,
and the elections of 1993 and 1998 were freer and fairer than many observers
had predicted, Hun Sen also used a strategy of violence and intimidation to
undermine the ability of his political opponents to challenge his authority.51
Starting immediately after the 1993 elections, he maneuvered himself and
the CPP into the dominant position in the coalition with FUNCINPEC, even
though they had received fewer votes. The 1997 coup d™´ tat seriously weak-
ened FUNCINPEC “ the CPP™s only major rival “ and represented “a glaring
example of violence against the democratic spirit.”52 Those who harmed or
threatened opposition elements escaped prosecution by a police and court
system that remained dominated by Hun Sen supporters, and the CPP lead-
ership “used every occasion to make sure that anyone accused of harming
its political interests or its party leaders was punished “ severely.”53 By the
end of the decade, Cambodia had only the “veneer” of democracy, and Hun
Sen ruled “by virtue of a monopoly of muscle, the readiness of thuggish
subordinates to use it and a tight grip on the machinery and resources of the
Indeed, one could argue that Cambodia™s relative stability after 1997
was due primarily to Hun Sen™s largely successful but patently illiberal
maneuverings to reestablish his de facto hegemonic control of Cambodian

48 49 Roberts 2001, p. 171. 50
Fleitz 2002, p. 129. Peou 2000, p. 427.
51 52 Ibid., p. 298. 53
Sanderson and Maley 1998. Ibid., p. 304.
54 Roberts 2001, p. 202; Economist 2002 (February 9).
Cambodia and Liberia 89

politics.55 In other words, it is unclear to what degree Cambodia™s relative
stability can be attributed to the internationally sponsored democratization
process, or instead to Hun Sen™s efforts to suppress political opposition in
the country. From this perspective, the 1997 coup d™´ tat and its aftermath
represented a return to traditional yet fundamentally undemocratic methods
for establishing political order in Cambodia: Through violence and intim-
idation, the CPP undercut the ability of opposition parties to participate
effectively in the democratic process.56
It is doubtful that this form of authoritarian quasi-democracy will offer a
lasting solution to the cycle of violence that the country has experienced since
its independence, since the ascendancy of Hun Sen seems to have followed the
long-standing tendency toward zero-sum competition for absolute control
in Cambodian politics “ a tendency that has provided the country with brief
periods of stability, but ultimately has always elicited violent counterreac-
tions from opposition groups. Put differently, the partial democratization of
Cambodian society that took place during the early 1990s under the super-
vision of international peacebuilders did not appear to alter the “underlying
tenor” of the con¬‚ict.57
This is not to suggest that the peacebuilders™ efforts in promoting polit-
ical liberalization left Cambodia in a worse condition than before. On the
contrary, whatever the weaknesses of the peacebuilding mission, it helped
to implement a peace settlement that brought an end to a prolonged period
of pervasive violence and brutality. The fact that Cambodia is no longer a
“killing ¬eld” is something to be celebrated, and the UN and other interna-
tional peacebuilders deserve partial credit for this accomplishment. But the
question we are investigating is more speci¬c: Does Cambodia corroborate
the assumptions of Wilsonianism? To be more precise, did internationally
sponsored liberalization efforts in Cambodia help to create the conditions for
a stable and lasting peace? On the basis of record to date, the answer must be
both yes and no. On the one hand, multiparty elections provided a formula
that the warring parties were willing to adopt as the basis for determining
who would rule Cambodia in the postcon¬‚ict period. Further, elections have
been generally free and fair, and in 1998 the populace expressed its support
for Hun Sen and the CPP. But on the other hand, the transition to liberal
democracy was very limited, and to some extent the modest achievements
of 1993 were reversed in the ensuing years as Hun Sen undermined his po-
litical opponents and consolidated his near-absolute powers using decidedly
illiberal means.

55 Freedom House, a U.S.-based organization that analyzes the political and civil liberties of
every country in the world, downgraded Cambodia from “partly free” to “not free” in
the mid-1990s: http://www.freedomhouse.org/research/freeworld/FHSCORES.xls, accessed
in August 2002.
56 57 Roberts 2001, p. xv.
Fleitz 2002, p. 128.
The Peacebuilding Record

At this writing, the government continues to control the police and ju-
diciary, and there is effectively no opposition in the country. As noted in
Chapter 1, the liberal peace thesis “ of which Wilsonianism is one variant “
is based on the premise that peace derives not only from elections but also
from real political contestation within the rule of law, respect for political
and civil liberties, constitutional limitations on the exercise of governmental
power, and the maximization of individual freedom. Cambodia has had elec-
tions but lacks many other characteristics of liberal democracy.58 We must
therefore treat with skepticism any suggestion that the Cambodian expe-
rience corroborates the Wilsonian hypothesis. Indeed, the country became
more stable in the late 1990s as Hun Sen increasingly backed away from his
earlier democratic commitments.

In the early nineteenth century, the American Colonization Society purchased
land that would later became part of Liberia, a small country on the west
coast of Africa that shares a border with Guinea, Ivory Coast, and Sierra
Leone, and began resettling freed American slaves to the territory.59 When
Liberia gained independence in 1847, it was governed by a small group of
Americo-Liberians, who subjugated the indigenous population in a series of
wars. Ironically, given the new arrivals™ origins as former slaves themselves,
the Americo-Liberian elite created a system of forced labor that continued
well into the twentieth century. (The League of Nations condemned it as
virtual slavery in 1930.)60 Approximately three hundred closely knit families
formed the ruling elite, which preserved its “feudal oligarchy” intact for over
a century.61
In 1980, a small group of indigenous army of¬cers overthrew the Americo-
Liberian regime in a violent coup and installed Master Sergeant Samuel Doe
as Liberia™s new head of state. At ¬rst, the native population welcomed
the coup and the new regime, believing that their “needs and interests,
which had long been overlooked, would now claim the full attention of their
government.”62 But their hopes soon met the reality of Doe™s increasingly
authoritarian, self-serving, and nepotistic rule. Like the Americo-Liberians
before him, Doe created a governmental system that bene¬ted a small mi-
nority within the country “ in this case, his own ethnic group, the Krahns,
who constituted about only 4 percent of the population.63 Many of the
president™s non-Krahn supporters were gradually eliminated and all forms
of opposition were suppressed.64 To the surprise of many, Doe agreed to

58 59 Conteh et al. 1999. 60 Dalton 1965.
Sanderson 2001.
61 Alao et al. 1999, pp. 12, 14.
62 W. Nah Dixon, Great Lessons of the Liberian Civil War, quoted in Alao et al. 1999, p. 18.
63 64 Sesay 1996; Alao et al. 1999, p. 18.
Alao 1998, p. 11; Ero 2000, p. 197.
Cambodia and Liberia 91

hold general elections in 1985, but after early returns indicated that he was
losing the vote, he brazenly manipulated the vote count and announced his
victory.65 Critics of these results faced “brutal reprisals.”66
Animosity against the Krahn hegemony “ and against the person of Doe “
was widespread in Liberia by the late 1980s, and it created propitious condi-
tions for a former Doe ally, Charles Taylor, to launch an insurrection against
the government in December 1989. Taylor™s rebel force, the National Patri-
otic Front for Liberia (NPFL), gained the support of peasants in the hinter-
land as Doe™s army terrorized non-Krahn ethnic groups that he suspected
of treason.67 The insurgents advanced quickly, reaching the outskirts of
the Liberian capital, Monrovia, in July 1990. In response, the Economic
Community of West African States (ECOWAS), led by Nigeria, sponsored
peace negotiations and deployed a cease-¬re monitoring group (ECOMOG)
comprised of approximately four thousand troops from Gambia, Ghana,
Guinea, Nigeria, and Sierra Leone. However, ¬ghting in Liberia did not
stop. President Doe was captured and killed in 1990, and opposition forces


. 3
( 10)