<<

. 7
( 10)



>>

circumstances that I shall describe in a moment “ partition raises serious
and ultimately insurmountable problems as a peacebuilding strategy in most
postcon¬‚ict situations.
One dif¬culty is the role that peacebuilders would be called upon to
play in its implementation, particularly when rival communities are not
already completely separated. In such circumstances, Kaufmann advocates
“separation campaigns” in which international intervenors would forcibly
transfer civilian populations to create ethnically homogeneous regions. This
procedure would involve several steps.11 First, international forces would

6 Kaufmann 1996b, p. 137.
7 In addition to the references in n. 3, see also Pape 1997; and the comments of U.S. Senator
Kay Bailey Hutchinson (1997) and Washington Post columnist Robert Novak (1997).
8 Pape 1997, p. 27.
9 This may explain Kaufmann™s (1998, pp. 152“155) ¬nding that the successor states of parti-
tions tend to be no less democratic than their predecessors, and in some cases are considerably
more democratic.
10 Sambanis 2000. For an overview of the debate, see Kumar 2000.
11 Kaufmann 1996a, pp. 91“92; and Kaufmann 1996b, pp. 164“166.
Institutionalization Before Liberalization 183

have to intervene militarily in support of “the weaker side” in the local con-
¬‚ict. Second, they would need to decide upon the geographic boundary line
for the partition. Third, the international forces would need to take physical
control, through a conventional military operation, of the territory on the
side of the boundary line and in the process drive “enemy” forces out of the
target region. Fourth, once the territory was occupied, civilians of the enemy
ethnic groups who remain behind would be interned, to be exchanged af-
ter the war. Finally, intervention forces would withdraw after guaranteeing
the separation lines, by international agreement if possible, or by ongoing
military assistance if necessary.
Kaufmann should be lauded for carrying through his prescriptions to their
logical conclusions, but in doing so he exposes a deep ¬‚aw within the parti-
tion strategy: namely, the requirement that international intervenors adopt
the role of “ethnic cleansers.” Kaufmann himself notes that the strategy is
“not pretty,”12 but he fails to spell out its full implications. He assumes, in
particular, that “enemy” civilians will be cowed into giving up their homes
without a ¬ght.13 This assumption is dubious: International forces would
almost certainly ¬nd themselves in situations in which they would be re-
quired to use deadly force (or threats of deadly force) in order to remove
civilians who adamantly refused to abandon their homes. If civilian resis-
tance to mandatory resettlement is widespread, or if it sparks violent riots,
peacebuilders may be called upon to kill large numbers of noncombatants
in the name of effecting a partition.14 So long as international actors remain
unwilling to perform such odious tasks, partition strategies will be impracti-
cal for warring states in which rival communities are not already separated
into homogeneous, or nearly homogeneous, enclaves.15


12 Kaufmann 1996a, p. 92.
13 “At ¬rst glance internment of enemy forces may seem to risk in¬‚icting a heavy loss of
life, but it need not. While forcible relocation is distasteful for the intervenors and very
unpleasant for the internees, given adequate planning, discipline, and resource it need not
be very dangerous” (Kaufmann 1996a, p. 95).
14 Indeed, partitions have historically tended to be quite messy, particularly when the rival
groups are geographically intermixed and must therefore be separated. In the violence that
accompanied the partition of India, for example, hundreds of thousands of people were
killed, with some estimates citing more than a million deaths (Hardgrave 1994, p. 71; and
MacFarquhar 1997, p. 26, n. 3). For a recounting of the “sordid history” of partitions,
see Kumar 1997. In the particular case of Bosnia, American mediator Richard Holbrooke
(1998, p. 363) offered this analysis of the partition option: “Dividing the country along ethnic
lines would create massive new refugee ¬‚ows. Serbs, Croats and Muslims who still lived as
minorities in many parts of the country would be forced to ¬‚ee their homes, and ¬ghting
would be certain to break out as the scramble for land and houses erupted again. Thus,
contrary to the arguments of the partitionists, the chances of ¬ghting would be increased,
not decreased, by partition and the relocation that would follow.”
15 As Sisk (1995, p. 32) writes, “Secession and partition can work only in societies where
communal groups are homogeneously concentrated in territories.”
Problems and Solutions
184

The third problem with the partition strategy ¬‚ows directly from the
second: If the international community were to adopt partition as a stan-
dard peacebuilding formula, it would almost certainly create incentives for
violent nationalists in other states to launch ethnic cleansing campaigns in
the hopes of gaining international support for the creation of new, ethnically
based political units. Kaufmann assumes that this would not happen, arguing
that local actors would be dissuaded from launching such campaigns on the
grounds that “secession attempts are very costly,”16 but he appears to miss
the point that international peacebuilding strategies in support of partition
would reduce the expected costs of a secession attempt and might therefore
make it a more attractive option to local actors. This is not to argue that the
international community should never support the partition of war-shattered
states: In countries where warring communities are already physically sepa-
rated, such a strategy may be appropriate, particularly if it does not require
international agencies to perpetrate or endorse ethnic cleansing. The point,
rather, is that even in these rare circumstances, the partition strategy should
be used as a last resort because of the “moral hazard” that disaffected groups
in other countries will launch their own ethnic cleansing campaigns in order
to secure international assistance in carving out their own, ethnically pure
states.
More generally, Kaufmann espouses partition because he believes that
the alternative is continued ¬ghting and “intercommunal slaughter.”17 This
proposition is rooted in yet another dubious assumption: that ethnic identi-
ties cannot “soften” in the period following the termination of hostilities. I
argued in Chapter 9 that we should expect social cleavages in war-shattered
states to be deep and cumulative because of the polarizing effects of civil
con¬‚ict. On this point, I agree with Kaufmann. However, I disagree with his
assessment of how durable these cumulative cleavages will be in the postcon-
¬‚ict period. According to Kaufmann, once the genie of ethnic mobilization
is out of the bottle, there is no way to put it back in; ethnic identities that are
hardened by civil war will remain hard inde¬nitely.18 Only a physical parti-
tion, or an overpowering and permanent military occupation by an outside
power, can prevent such a state from experiencing renewed violence. Yet, if
one accepts that ethnic identities are in some measure “socially constructed”
rather than innate, ¬xed, and timeless phenomena, there is at least a possi-
bility that hardened communal identities will soften over time.19 Kaufmann
responds to this argument by claiming that conditions of continuing distrust
in states that have experienced intense ethnic violence will tend to reinforce
the hardness of these communal identities ad in¬nitum.


16 17 Kaufmann 1996a, p. 99. 18 Kaufmann 1996b, p. 153.
Kaufmann 1996b, p. 170.
19 On the social construction of ethnic identities, see Brass 1974; Eller and Coughlan 1993;
Eriksen 1993; Little 1994; and Nagel 1994.
Institutionalization Before Liberalization 185

This seems to be an overly deterministic argument.20 As I will argue in the
next section, peacebuilding policies can be designed in ways that promote
the gradual reintegration of rival groups in a war-shattered state, together
with a concomitant softening of communal identities.


Rethinking the Wilsonian Approach to Peacebuilding
Neither outright authoritarianism nor routine partition offers a viable alter-
native to the Wilsonian peacebuilding formula of the 1990s. A more sensible
option is to modify only those aspects of Wilsonianism that are problem-
atic, rather than rejecting Wilsonianism in its entirety. To this end, we need
to distinguish once again between the ends and means of the Wilsonian
approach. The end goal is to consolidate peace in war-shattered states by
transforming these states into liberal market democracies. This aim is not, in
itself, unsound, for well-established market democracies tend to be peaceful
both in their internal affairs and in their relations with other democracies.
Furthermore, there are other reasons (aside from peace) to promote the prin-
ciples and practices of market democracy, since the political and economic
freedoms that underpin market democracy have their own intrinsic value as
human rights.
The problem with Wilsonianism, rather, lies in the methods that it em-
ploys to achieve this goal, and the assumptions upon which these methods
are based. It is the process of liberalization, rather than the end point of mar-
ket democracy, that has generated the destabilizing side effects described in
the previous chapters. The challenge for peacebuilders, then, is to devise
methods of placing war-shattered states on a long-term path toward market
democracy, while avoiding the pathologies of liberalization itself.
How can this be done? Revisiting Samuel Huntington™s work on political
development offers a useful starting point. Huntington argued in the 1960s
that the very process of political and economic development could have dis-
locating effects on developing states, due primarily to social changes that
accompany the modernization process, such as increased levels of urbaniza-
tion, industrialization, and literacy, all of which serve to disrupt traditional
mechanisms of social control.21 By creating new demands within a society
while simultaneously undermining existing sources of political authority,
the modernization process produces “political instability and disorder” in

20 For an excellent critique of Kaufmann™s “essentialist” view of identity in relation to Bosnia,
see Campbell 1998.
21 Huntington 1968. For a similar argument emphasizing the potentially destabilizing effects
of the modernization process, see Deutsch 1961; Pye 1963; Apter 1970; Connor 1972;
Packenham 1973; Ake 1974; and Smock and Smock 1975. As Horowitz (1985, p. 99) points
out, Deutsch (1953 and 1961) appears to have been the ¬rst scholar to suggest that the
modernization process itself could engender social con¬‚ict.
Problems and Solutions
186

many developing states and thus derails efforts to transform these states into
market democracies.22 Huntington™s solution to this problem was to argue
in favor of restricting political participation and political mobilization in
the early phases of modernization, if necessary through the establishment
of authoritarian military governments in developing societies. Military rule,
he claimed, often represents the only viable method of managing the most
destabilizing effects of the modernization process, while creating effective
political institutions that could, at a later date, be placed under civilian and
democratic control.23
Huntington™s analysis is relevant to peacebuilding in many respects, but
his conclusions and recommendations are problematic. Particularly trouble-
some is his suggestion that military of¬cers should be relied upon to serve as
midwives for democracy “ a suggestion that he makes on the grounds that
military organizations are “cohesive, bureaucratized, and disciplined” and
therefore uniquely equipped to govern states undergoing political and eco-
nomic change.24 Experience in many parts of the developing world suggests
that military rulers are often reluctant to give up their power voluntarily to
civilian authorities.25 In addition, his analysis presupposes the existence of
a coherent national military, which rarely exists in countries that are just
emerging from civil con¬‚icts. War-shattered states tend to lack even this
minimal degree of central authority.
On the other hand, several features of Huntington™s analysis (in contrast
to his speci¬c recommendations) do relate directly to the problems of post-
con¬‚ict peacebuilding. His main argument is that “political order depends in
part on the relation between the development of political institutions and the
mobilization of new social forces.”26 In the absence of political institutions
capable of processing societal demands and ensuring compliance with the
rule of law, the stimulating of political mobilization can lead to civil strife
and unrest. Modern liberals, particularly in the United States, tend to view
the task of government building as a process of limiting and dividing power,
but Huntington points out that political stability is a precondition for polit-
ical liberty, and that political stability requires the existence of effective and
authoritative political institutions. “Authority has to exist before it can be
limited,” he writes.27

22 Huntington 1968, p. 5. Other commentators placed greater emphasis than Huntington on the
psychological sources of social unrest during the modernization process. For example, Apter
(1970, p. 159) wrote: “As societies modernize, the normative integration of the previous
system begins to weaken,” which leads to “less predictability in social action” and “greater
uncertainty by individuals both of themselves and of the anticipated responses of others.”
For a similar view, see Pye 1963, pp. 54“55.
23 24 Ibid., p. 239.
Huntington 1968, chap. 4.
25 Huntington (1968) notes this danger (on pp. 233 and 237) but remains unaccountably opti-
mistic about the prospects for military of¬cers to give way to democratic rule.
26 27 Ibid., p. 8.
Ibid., p. viii.
Institutionalization Before Liberalization 187

Despite surface appearances to the contrary, this argument is not an illib-
eral one. Huntington does not reject the importance of constitutional con-
straints on the exercise of governmental power. Rather, he argues that a func-
tioning liberal democracy must have both limited and effective government.
His argument thus recapitulates the classical liberal formula for a peaceful
(and just) polity: that there be respect for individual freedoms, protected by
limitations on the arbitrary exercise of governmental power, and that a cen-
tral authority exist in order to maintain a basic regulatory framework for
political and economic life.
Peacebuilders, in other words, have been insuf¬ciently attuned to the
“Hobbesian problem” that the classical liberals addressed and that modern
scholars of the liberal peace have largely assumed away. Peacebuilders have
sought to transform war-shattered states into liberal market democracies by
drafting rights-based constitutions, holding multiparty elections, liberalizing
the popular media and political activity, and implementing market-oriented
economic reforms. But these measures have, in several cases, spurred societal
competition before effective governmental structures could be established to
regulate this competition.
Yet it is not only the absence of effective governmental institutions that
makes war-shattered states vulnerable to the pathologies of liberalization. As
noted earlier, these states also tend to lack other domestic characteristics that
might otherwise help to moderate the intensity of societal con¬‚icts, includ-
ing a tradition of peaceful dispute resolution and a cross-cutting cleavage
structure. Putting more emphasis on the construction of effective institu-
tions is only one part of the solution; other measures that aim more directly
at limiting the intensity of societal con¬‚icts should also be part of the re-
vised peacebuilding strategy. Among other things, this means rejecting the
“shock therapy” approach to economic liberalization, which is ill suited to
the task of minimizing societal con¬‚ict in the immediate postwar period. It
also means that the guiding doctrine of peacebuilding should emphasize the
gradual and controlled liberalization of political and economic life.


Institutionalization Before Liberalization
The peacebuilding strategy I propose would preserve the Wilsonian goal
of transforming war-shattered states into liberal market democracies in the
long run, while minimizing the destabilizing effects of the liberalization pro-
cess in the short run. I call the strategy Institutionalization Before Liber-
alization (IBL) because the central recommendation is that peacebuilders
should concentrate on constructing a framework of effective institutions
prior to promoting political and economic competition. What is needed in
the immediate postcon¬‚ict period is not democratic ferment and economic
upheaval, but political stability and the establishment of effective adminis-
tration over the territory. Only when a working governmental authority has
Problems and Solutions
188




1. Wait Until Conditions Are Ripe for Elections

2. Design Electoral Systems That Reward Moderation

3. Promote Good Civil Society

4. Control Hate Speech

5. Adopt Con¬‚ict-Reducing Economic Policies

6. The Common Denominator: Rebuild Effective State Institutions

¬gure 10.1. Key Elements of the IBL Peacebuilding Strategy


been reestablished should peacebuilders initiate a series of gradual demo-
cratic and market-oriented reforms. Put another way, peacebuilders should
delay liberalization and limit political and economic freedoms in the short
run, in order to create conditions for a smoother and less hazardous transi-
tion to market democracy “ and durable peace “ in the long run.
The main elements of this strategy include the following: 1) postponing
elections until moderate political parties have been created, and mechanisms
to ensure compliance with the results of the election have been established;
2) designing electoral rules that reward moderation instead of extremism;
3) encouraging the development of civil-society organizations that cut across
lines of societal con¬‚ict, and proscribing those that advocate violence; 4) reg-
ulating incendiary “hate speech”; 5) promoting economic reforms that mod-
erate rather than exacerbate societal tensions; and 6) developing effective se-
curity institutions and a professional, neutral bureaucracy. (See Figure 10.1.)
Some of these elements have been implemented in missions launched after
1998, but, as we shall see later, these newer operations have continued to
emphasize rapid liberalization at the expense of institutionalization and have
consequently suffered from many of the weaknesses that characterized ear-
lier missions.

Wait Until Conditions Are Ripe for Elections
Peacebuilders have sponsored elections in war-shattered states as a means
of facilitating the peaceful management of societal con¬‚icts through compe-
tition at the ballot box, rather than through combat on the battle¬eld. But
elections do not always foster peaceful forms of competition, nor do they
necessarily produce governments committed to resolving disputes through
negotiation and compromise “ or, for that matter, governments committed
to preserving democracy. If the parties that win elections are dedicated to
the violent destruction of their rivals, or if they seek to undermine the very
democratic institutions that brought them to power, elections may actually
Institutionalization Before Liberalization 189

work against the goal of establishing a stable liberal democracy. Moreover,
if the parties that contest elections attempt to build popular support by
appealing to intercommunal fears and hatreds, the election campaign can
itself rekindle the very con¬‚icts that peacebuilders seek to mitigate.
Rather than recognizing that elections can sometimes serve as focal points
for destructive forms of competition, peacebuilders in operations conducted
between 1989 and 1998 tended to portray elections as simply serving the
cause of peace. In all eleven cases, preparations for elections began immedi-
ately, and they were held (or were about to be held, in the case of Rwanda)
within three years of the deployment of the operation. Peacebuilders also
typically presented the holding of an election as a signal of “success” and a
rationale for the termination of the operation. In some instances, postcon¬‚ict
elections facilitated a peaceful transition of government and an important
development in the transition to stable democracy. But in other cases, elec-
tions spurred new violence or installed leaders who immediately began to
reverse the movement toward democracy. Peacebuilding agencies™ seemingly
doctrinaire belief in the pacifying effects of elections allowed little room for
serious consideration of the potentially damaging consequences of rushed
elections.
So the ¬rst element of the IBL strategy is to delay the holding of elections
until conditions are propitious for a successful vote “ a vote that not only is
conducted in a free and fair manner but also furthers the development of sta-
ble democracy and diminishes the risk of renewed violence. Judging whether
such conditions exist requires an assessment of 1) the political parties that
are likely to contest the election, and 2) the institutional setting in which the
election will take place.
With regard to the parties, the most dangerous circumstances are those in
which the principal contenders for election are the very individuals or orga-
nizations that recently fought the civil war, particularly if these individuals
and groups exploit intercommunal fears and hatreds in order to build elec-
toral support. Although involvement of former belligerents in postcon¬‚ict
elections may be unavoidable because they often remain the most prominent
leaders in war-shattered states, peacebuilders can nevertheless use a variety
of means to promote greater moderation in the parties contesting elections.
By resisting calls for early elections, peacebuilders can allow passions to
cool with the passage of time. International agencies can also use induce-
ments and punishments “ carrots and sticks “ to encourage the rise of new
moderate parties and leaders in the period leading up to elections. These
inducements may include signi¬cant ¬nancial support to parties that pub-
licly eschew violence and violent rhetoric and make concerted efforts to gain
popular support from voters across communal lines. Punishments may in-
clude the banning of parties that preach violence or hatred from participating
in elections. Peacebuilders should proceed with elections only when there is
evidence that “moderate” parties “ or those that seem genuinely committed
to resolving disputes through peaceful negotiation and to intercommunal
Problems and Solutions
190

reconciliation “ have suf¬cient popular support (as demonstrated by inter-
nationally sponsored opinion surveys) to prevail over “immoderate” parties
at the polls.
Although international peacebuilding agencies have assisted in the draft-
ing and oversight of electoral laws in war-shattered states, they have gen-
erally been reluctant to become directly involved in regulating the activities
of political parties. In only one mission “ Bosnia “ did international of¬cials
prohibit certain individuals from contesting public of¬ce: namely, individu-
als indicted for war crimes. As we shall see, the willingness of peacebuilders
(in Bosnia and elsewhere) to regulate participation in postcon¬‚ict elections
has increased since 1998, but only marginally. Holding elections soon af-
ter the termination of hostilities is still treated as a top priority that trumps
virtually all others, including the question of whether peace can survive the
pressures of electoral mobilization, given the character of the political par-
ties that are likely to contest the election. If local parties preach intolerance
and hatred toward their rivals, or display little commitment to sustaining
democracy once in power, there is little to be gained by proceeding with
elections. Instead, peacebuilders should use the period leading up to elec-
tions to promote moderation within existing parties, to foster the growth
of new democratic and moderate parties, and if necessary, prevent the most
intolerant individuals and parties from running for public of¬ce.
The second consideration when judging the ripeness of conditions for
successful postcon¬‚ict elections is whether the country™s governmental in-
stitutions are capable of resolving disputes arising from the election and of
enforcing compliance with the election™s outcome. Before holding elections,
peacebuilders should ensure that a functioning, professional, and neutral
judicial mechanism to rule on election-related disputes is already in place.
Achieving this objective would typically require training or retraining judges
and legal advocates, and creating a working high court or constitutional
court to render authoritative judgments. (The existence of an effective consti-
tutional court would also diminish the threat of an unconstitutional usurpa-
tion of power by the newly elected government.) The court would also require
reliable enforcement mechanisms for upholding the court™s decisions “ at the
very least, a police force that is not beholden to any of the formerly warring
groups or to any individuals or parties running for election.
Although peacebuilders have worked toward the reconstruction of a func-
tioning judiciary and police force in many missions, in practice elections
have generally taken place prior to the establishment of effective judicial and
police institutions. The danger of conducting elections without such institu-
tions was most graphically illustrated in Angola, but problems of resolving
election-related disputes also arose, and threatened to escalate into violence,
in Cambodia and Mozambique. Moreover, in the absence of effective mech-
anisms to enforce constitutional norms, the undemocratic maneuverings of
newly elected leaders in Liberia and Cambodia went largely unchecked.
Institutionalization Before Liberalization 191

To date, peacebuilders have been reluctant to delay elections or to slow
the pace of democratization, even in the most adverse circumstances and in
spite of mounting evidence that hasty elections can have pathological effects
in war-shattered states. Minor scheduling adjustments have occurred in a
few cases: Mozambique™s ¬rst national elections were delayed from October
1993 to October 1994; and Bosnia™s ¬rst municipal elections were deferred
from September 1996 to September 1997 because of “massive registration
irregularities.”28 The IBL strategy, by contrast, would permit the postpone-
ment of elections for as long as it takes to establish the right institutional
and political conditions for a constructive vote that advances, rather than
hinders, efforts to consolidate a stable and lasting peace.


Design Electoral Systems That Reward Moderation
If and when the decision is made to proceed with elections, the design of
the electoral system itself could help to promote reconciliation among for-
merly warring parties. “For most leaders, most of the time,” writes Donald
Horowitz, “there are greater rewards in pursuing ethnic con¬‚ict than in pur-
suing measures to abate it. One of the great challenges of political engineers is
to make moderation rewarding and to penalize extremism.”29 Recent studies
have shown that certain electoral systems can indeed “encourage moderate,
centrist forms of political competition, rather than the polarizing extremes
and centrifugal patterns that characterize so many divided societies.”30
Presidential elections in Nigeria, for example, have previously required
presidential aspirants to win not only an absolute majority of national votes
but at least 25 percent of votes cast in no fewer than two-thirds of the nine-
teen states “ a requirement that, in practice, encouraged serious candidates
to “reach out and conciliate and propitiate the interests of groups other than
the ones [that they were] accustomed to appealing to.”31 The Nigerian ex-
periment lasted only from 1979 to 1983. It ultimately failed not because
of the presidential election formula, which successfully reduced the ethnic
partisanship of presidential candidates, but because the country™s national
legislature continued to be dominated by ethnic parties, and because vote
rigging and other forms of corruption undermined popular support for the
regime as a whole.32

28 29 Horowitz 1990b, p. 452.
Pomfret 1996.
30 Sisk and Reynolds 1998; and Reilly 2001, p. 7 and passim.
31 Horowitz 1990a, p. 127. See also Esman 1994, pp. 43“44. The Nigerian arrangements bore
a close resemblance to a scheme that was proposed, but not implemented, by the Muslim
leader Mohamed Ali in India in 1935. According to Ali™s plan, Muslim and Hindu electorates
in India were to be kept separate, but no candidate would be declared elected unless he had
secured at least 40% of the votes cast by his own community and 5% of the votes cast by
the other community (Laponce 1957, p. 324).
32 Nmona 1985, pp. 324“327; and Horowitz 1985, pp. 636“638.
Problems and Solutions
192

Contrast the Nigerian presidential election formula with the arrangements
governing Bosnia™s presidency. Both systems were designed to produce po-
litically moderate chief executives, but they sought to achieve this goal in
very different ways. In Nigeria, candidates were constitutionally required to
gain support from rival communities in order to win election. This require-
ment created a strong political incentive for aspiring candidates to promote
policies that would appeal to moderate voters across lines of social cleavage.
The Bosnian constitution adopted at Dayton in 1995, on the other hand,
provided for a tripartite presidency, with one representative from each of
the three major ethnic communities. The constitution called upon the co-
presidents to “adopt all Presidency Decisions . . . by consensus,” although
the legal requirement for measures to pass was only two af¬rmative votes,
not three.33 In order to prevent two members of the presidency from gang-
ing up on the third, the constitution also permitted each co-president to
exercise a veto over speci¬c decisions that were deemed to be “destructive
of a vital interest.”34 These arrangements, it was hoped, would encourage
the Bosnian presidency to devise policies that balanced the interests of the
country™s constituent communities.
However, the Bosnian constitution provided few incentives for aspiring
presidential candidates to solicit political support in different ethnic groups.
Indeed, quite the opposite: The population was divided into separate ethnic
electorates, with voters permitted to cast ballots only for the presidential
seat that corresponded to their individual ethnic af¬liation. This electoral
system reinforced the propensity for political mobilization in war-shattered
states to follow the existing lines of communal cleavage. With the electorate
thus divided, and with no electoral advantages to be gained by campaign-
ing for votes in more than one ethnic community, it is little wonder that
the candidates who won the 1996 presidential election gained power pre-
cisely because they appealed to ethnic nationalist sentiments within their re-
spective constituencies. The duly elected co-presidents thus came into of¬ce
with little interest in compromising with the other members of the presi-
dency, since it was on the basis of their credentials, not as conciliators but
as defenders of their own group™s interests over and above the interests of
their ethnic adversaries, that they had each gained power. The result was
the virtual paralysis of the presidency as a decision-making institution, and
the perpetuation in power of groups that are determined “to prevent their


33 Constitution of Bosnia-Herzegovina, Article V, Section 2(c).
34 Ibid., Section 2(d). The veto procedure is actually quite involved: Brie¬‚y, vetoes must be
recon¬rmed by a two-thirds vote of those members of the entity-level legislature who belong
to the same ethnic group as the presidency member who initially employed the veto. If
such a vote does not pass within ten days of the matter™s being referred to the relevant
entity-level legislature, then the veto would not take effect, and the original measure would
pass.
Institutionalization Before Liberalization 193

separate power structures from being subject to constitutional and demo-
cratic control.”35
The shortcomings of Bosnia™s presidential power-sharing formula appear
to be shared more generally by “consociational” models of governance.
Consociationalism, according to Arend Lijphart, who ¬rst coined the term,
describes a form of democratic government that has been employed in sev-
eral countries with well-de¬ned cumulative cleavages, most notably Austria,
Belgium, the Netherlands, and Switzerland.36 Consociational democracies
have four distinguishing characteristics: 1) Government consists of “a grand
coalition of the political leaders of all signi¬cant segments of the society,”
meaning that all the major social groupings share power in all the institu-
tions of government; 2) representation in these institutions is proportionate
to each segment™s share of the population as a whole; 3) policymaking is
subject to a “concurrent majority rule,” requiring joint decisions on matters
of common interest to be approved by all major segments in the society;
and 4) each segment retains exclusive decision-making authority over mat-
ters that are of concern only to that segment.37 Although Bosnian political
institutions do not display all the characteristics of consociationalism, the
institution of the presidency rests on many of the same principles, including
the notion that government should consist of a “grand coalition” of the ma-
jor social groupings in a country, and that each of these groupings should
have the power to veto joint decisions (which is the logical corollary of the
“concurrent majority rule”).
Lijphart promotes consociationalism as “the best type of democracy that
can realistically be expected” in deeply divided societies,38 but the record of
Bosnia™s multiheaded presidency suggests that consociationalism is ill suited
to the domestic conditions of war-shattered states. Both the Bosnian pres-
idency and the consociational model presume that government institutions
will be comprised of individuals who, in Lijphart™s words, “have the ability
to transcend cleavages and join in a common effort with the elites of a rival
subculture.”39 Yet for reasons I have already described, such individuals are
unlikely to achieve electoral victory in the highly polarized environment of
a war-shattered state “ unless, of course, the electoral rules themselves re-
quire candidates to secure a minimum level of support in communities other
than their own. Lijphart™s consociationalism and the Bosnian presidential
formula call for the division of the electorate into separate voting constituen-
cies according to communal af¬liation, which makes cross-factional voting
impracticable. Nor should we expect nationalist political parties in war-
shattered states to be willing to form alliances with parties in rival ethnic
camps (which, in theory, might be another way of moderating the policies of

35 Cox 2001, p. 8.
36 The classic work on consociationalism is Lijphart 1977. See also the essays in McRae 1974.
37 38 Ibid., p. 48. 39 Lijphart 1969, p. 216.
Lijphart 1977, pp. 25“52.
Problems and Solutions
194

these parties in the absence of cross-factional voting). If individual politicians
in these circumstances have little to gain and much to lose by appearing to be
“soft” on their ethnic adversaries, political parties would presumably also
face strong political incentives to maintain a hard line toward rival groups.
The challenge for peacebuilders, therefore, is to devise electoral and con-
stitutional rules that compel serious candidates to secure signi¬cant political
support across different communal groups. There is no magic formula to
achieve this goal “ electoral arrangements must be adapted to the circum-
stances of each state. In Bosnia, for example, an alternative to the Dayton
plan might have been a one-person presidency with a dual requirement for
victory: a plurality of votes cast in the entire country and a given percentage
of votes cast in each of the ethnic communities or geographical regions of
Bosnia.40 Like the Nigerian election rules, these requirements would have
likely created stronger incentives for candidates to appeal to moderate voters
in each ethnic camp, which might have ultimately encouraged the growth
of cross-factional political movements. Whatever speci¬c formula is chosen,
the broader point is that peacebuilders should deliberately design electoral
institutions to elicit moderation and cross-factional compromise from the
parties vying for election and from governments seeking reelection.


Promote Good Civil Society
Not all civil-society associations are conducive to peaceful democratic poli-
tics, particularly those that espouse violence against members of other groups
or that reject the idea of democracy itself.41 Promoting the development of
civil-society associations can therefore have quite different results depend-
ing on the nature of the associations created. At one extreme, pluralistic
organizations can help to break down social barriers between formerly war-
ring communities and provide grassroots support for political parties that
support intergroup accommodation. At the other extreme, chauvinistic or-
ganizations can reinforce existing social cleavages, exacerbate tensions, and
discourage cross-communal compromise. The challenge for peacebuilders,
therefore, is to promote “good” civil society while simultaneously restraining
its “bad” variant, particularly during the early phases of a peacebuilding mis-
sion when governmental institutions are still being constructed.

40 In addition, using an “alternative vote” system could have helped preclude the possibility
of no candidate meeting the minimum requirements for electoral victory. This system “ the
current method of electing Australia™s House of Representatives “ requires voters to rank
candidates on the ballot in order of preference. If no candidate meets the conditions for
election after all the ¬rst preferences have been counted, then the last-place candidate is
eliminated and all of his or her votes are transferred to the remaining candidates according
to second preferences indicated on the ballots, and so on, until a winner emerges with the
required levels of support. See Rae 1967, p. 24; and Palley 1978, pp. 16“17.
41 See the section on Bad Civil Society in Chapter 9.
Institutionalization Before Liberalization 195

Peacebuilding agencies have begun to experiment with programs aimed
at promoting the development of cross-factional social groups in countries
that recently experienced civil con¬‚icts. The Organization for Economic
Cooperation and Development, for example, has conducted “culture of
peace” programs in El Salvador and Rwanda that are designed to increase
grassroots dialogue and cooperation among members of formerly warring
communities.42 Similar initiatives have also been undertaken in Tajikistan
and the Philippines.43 Yet these programs remain largely experimental and
peripheral to the activities of the principal peacebuilding agencies.
A stronger commitment to nurturing the development of cross-factional
social groupings in war-shattered states should become a central element of
a broader strategy aimed at preventing the process of political liberalization
from intensifying intercommunal con¬‚ict. In particular, peacebuilding agen-
cies should offer greater ¬nancial and logistical support to cross-factional
associations in different areas of social, economic, and political life, from
political lobbying groups to trade unions to private social clubs.44 Clear
criteria should be established for groups seeking international support: For
example, groups might be required to have a cross-factional membership,
to hold regular meetings, to adopt a set of guiding principles for their or-
ganization that includes the goal of enhancing intercommunal cooperation
and tolerance, and to revoke the membership of any individual who openly
espouses violence against another group.
Peacebuilders must also be prepared to shut down organizations that
openly and repeatedly advocate violence against other groups in the society
when such behavior poses a threat to the consolidation of peace and democ-
racy. The precedent for such action is postwar Germany, in which the occupy-
ing Allied powers sponsored the reconstruction of political parties and civil-
society groups on the condition that these groups repudiate nationalism and
embrace democracy and tolerance. Organizations that failed to comply with
these guidelines were banned. This “denazi¬cation” process initially tar-
geted not only organizations but individuals as well: The Allies agreed to re-
move former Nazi Party members from “public and semi-public of¬ce, and
from such positions of responsibility in important private undertakings,”
and to replace them with persons “deemed capable of assisting in devel-
oping genuine democratic institutions in Germany.”45 This form of social
regulation may seem heavy-handed, but it was also crucial in restoring ci-
vility to Germany and in creating conditions for a stable, free, and peaceful
democracy in the long term.

42 See UN General Assembly document A/51/395, September 23, 1996, annex, for an overview
of OECD activities in this area.
43 See Slim and Saunders 1996, and Saunders 1999, chap. 7, for descriptions of the program
in Tajikistan. On the Philippines, see Garcia 1989.
44 45 From the Potsdam Agreement of 1945.
Prendergast and Plumb 2002.
Problems and Solutions
196

The regulation of extremist groups and individuals “ that is, those who
preach hatred and violence “ is one means of placing limits on the con¬‚ict-
exacerbating effects of political liberalization, while at the same time foster-
ing the development of civil-society associations that support democracy and
cross-factional compromise. In practice, however, peacebuilders have been
hesitant to exercise such powers, even when they have explicitly reserved
for themselves the right to dismiss of¬cials or to prevent certain individuals
from serving in government. In Cambodia, for example, UNTAC had for-
mal powers to “require the reassignment or removal of any personnel” in the
Cambodian government46 but UN of¬cials never actually used this power,
despite “numerous cases” of bureaucrats acting illegally in support of one of
the Cambodian political parties while undermining the others.47 The Bosnia
operation was the only other mission deployed between 1989 and 1998 in
which international personnel had the formal authority to remove local of-
¬cials who obstructed the peace process “ a power that international of¬cials
began to use, with great effect, in 1999. In general, peacebuilders could be
much more assertive in regulating the conduct of groups and individuals in
war-shattered states, focusing on those that promulgate ethnic hatred “ just
as nationalist and antidemocratic groups and members of the former Nazi
regime were proscribed from public life in postwar Germany.


Control Hate Speech
A free press, which provides a forum for debating public issues and criticizing
government policies, is vital to a functioning democracy.48 Encouraging the
development of free media should therefore be one of the central goals of
peacebuilding, as it has been in many operations. But the liberalization of
the media in war-shattered states can also lead to the proliferation of “news”
outlets that deliberately seek to incite hatred and violence against particular
groups within the society. When these messages ¬nd receptive audiences “ as
they did in Rwanda, for example “ new rounds of civil violence, including
attacks on innocent civilians, may result. As one observer writes: “A biased or
hatemongering media can sabotage almost any other peacebuilding effort.”49
Peacebuilders should pursue a two-track media policy. The ¬rst track
involves the development of responsible news outlets and the provision of
reliably accurate sources of information to the inhabitants of war-shattered
states; the second track involves regulating the activities of news media that
incite hatred and violence against particular groups. The UN and other inter-
national agencies have actively pursued the ¬rst track since the earliest post“
Cold War peacebuilding missions were launched.50 In Cambodia, for exam-
ple, Radio UNTAC transmitted a mixed format of news and entertainment

46 47 McLean 1994, pp. 55“56. 48 Lichtberg 1990.
United Nations 1995, p. 100.
49 50 For an overview of these efforts, see Lehmann 1999.
Howard 2001, p. 12.
Institutionalization Before Liberalization 197

programming, which became very popular in the country and provided
a neutral forum in which the political factions contesting election could
communicate their policy positions. At the same time, a panoply of gov-
ernmental and nongovernmental agencies “ from the United Nations Edu-
cational, Scienti¬c and Cultural Organization (UNESCO) to the U.S.-based
Asia Foundation “ have offered technical and ¬nancial assistance to local
newspapers and radio stations with the goal of promoting “independent,
open, and accountable” sources of information to the public, while also
seeking to protect responsible journalists from government intimidation.51
Similarly in Bosnia, funding from the European Union and the Open Society
Foundation (among other donors) supported the creation of the Open Broad-
casting Network, a group of local television stations that receive interna-
tional ¬nancial and training support in exchange for their compliance with
an editorial charter that mandates factual accuracy in reporting and prohibits
the use of any “language or representation which incites discrimination, prej-
udice or hatred.”52
While peacebuilders have offered extensive assistance to independent local
media in war-shattered states, they have been reluctant to pursue the sec-
ond track, which involves suppressing publications or broadcasting outlets
that disseminate maliciously distorted and in¬‚ammatory “news” directed
against a particular group. This reluctance is understandable: Censorship,
even for the purposes of promoting a free and responsible press, sets a trou-
bling precedent that could make it easier for subsequent governments to
suppress legitimate political dissent. But when “hate media”53 incite geno-
cidal violence or rekindles civil con¬‚icts, peacebuilders should not adhere
so blindly to the principle of press freedom that they allow greater harm to
occur through their inaction.
The argument that jamming incendiary broadcasts is never permissible54
ignores the role that Radio-T´ l´ vision Libres des Milles Collines and other
ee
radio stations played in orchestrating the Rwandan genocide of 1994. The
UN commander in Rwanda, General Rom´ o Dallaire, called for the jamming
e
of these broadcasts, but UN headquarters refused,55 a decision that was ap-
plauded by Western anticensorship groups, including the London-based Ar-
ticle 19 organization.56 The rigid anticensorship position is rooted in the no-
tion that promoting unconditional freedom of expression is the key to coun-
tering in¬‚ammatory propaganda,57 but as Jack Snyder and Karen Ballentine
point out, “from the French Revolution to Rwanda, sudden liberalizations


51 For example, see Sharpe 2001.
52 Open Broadcasting Network™s website: http://openbn.hypermart.net, accessed in July 2002.
53 Reporters Sans Fronti` res 1995.
e
54 Jeffery Heyman, the head of the UNPROFOR Radio Unit, is quoted making this argument
in Lehmann 1999, p. 100.
55 56 Article Nineteen 1995. 57 Human Rights Watch 1995.
Metzl 1997, p. 15.
Problems and Solutions
198

of press freedom have been associated with bloody outbursts of popular
nationalism.”58 Given the costs of inaction, peacebuilders would be well
advised to shut down newspapers and jam broadcasts that openly incite
violence in the fragile circumstances of a war-shattered state.59 Only once
have peacebuilders done so: In 1997, NATO soldiers in Bosnia took over
a key transmitter belonging to a Bosnian Serb radio station that had been
broadcasting in¬‚ammatory propaganda. NATO acted under a provision of
the Dayton Accord, which authorized peacebuilders to “suspend or curtail
programming that is hostile to the spirit” of the Accord. But all peacebuild-
ing missions should be given both the means and the mandate to block the
dissemination of hate propaganda.
One method of regulating hate speech is by establishing enforceable
“codes of conduct” for print and broadcast media, along with a licensing
system. As we shall see, peacebuilders in Kosovo have adopted this approach
with considerable success. Codes of conduct should include a positive com-
mitment to accuracy in reporting the news and prohibitions on the publica-
tion of material that incites hatred or violence. International of¬cials should
interpret these codes, rule on alleged violations, and impose sanctions, if nec-
essary, within a framework that allows for appeals. In Western democracies,
broadcast media are normally regulated and licensed by government agen-
cies, given the need to allocate broadcast frequencies across a limited spec-
trum. Print media, by contrast, tend to regulate themselves through “press
councils” and the like, and both print and broadcast journalists are typi-
cally also subject to libel laws. Some Western democracies, such as Germany
and Canada, have also criminalized the willful incitement of hatred against
particular groups. Thus, neither the licensing of broadcast media nor the
existence of rules governing the content of media are unknown in Western
democracies. What makes such regulation controversial in the context of
peacebuilding is the degree to which foreign of¬cials ought to impose rules
on the indigenous media. However, given the absence of domestic institu-
tional structures to manage hate speech in most of these countries, and the
potential dangers of allowing hate media to ¬‚ourish, there is little alternative
to doing so.
As in other areas of social, political, and economic life, the rapid and
unfettered liberalization of the media in war-shattered states does not nec-
essarily aid in the consolidation of peace or the transition to stable market
democracy. Creating institutional structures to manage the potentially patho-
logical effects of media liberalization is a sensible strategy for promoting a
free and responsible press. Over time, mechanisms for the self-regulation of
the media should largely replace those operated by peacebuilders. Indepen-
dent press councils, for example, might be created to interpret the codes of
conduct and impose sanctions on violators, if necessary. Initially, perhaps,

58 59
Snyder and Ballentine 1996, pp. 5“6. See Metzl 1997 for a similar argument.
Institutionalization Before Liberalization 199

peacebuilding agencies should retain a ¬nal veto over the operation of these
press councils, and ultimately this ¬nal veto power should be phased out
or transferred to the civilian court system once such a system is established
and functioning. At that point, media regulation in peacebuilding host states
would more closely resemble the regulatory arrangements that exist in many
Western democracies. But the key to this strategy is that it emphasizes in-
stitutionalization before liberalization and a managed transition to a free
press.

Adopt Con¬‚ict-Reducing Economic Policies
In Chapter 9 we examined some of the dangers and pathologies of rapid eco-
nomic adjustment in war-shattered states. Although “orthodox” structural
adjustment programs may create the conditions for sustainable economic
growth in the long term, they usually impose signi¬cant short-term social
costs. The economic reforms typically required by the IMF and World Bank
tend not only to lower the living standards of certain groups within states
undergoing adjustment but also to worsen the overall distribution of wealth
in these states.
Proponents of orthodox structural-adjustment models do not deny the
short-term social pains of economic reform, but they argue that temporary
dislocation is justi¬ed in order to create the conditions for sustainable eco-
nomic growth in the long term. There are, in fact, excellent reasons to encour-
age war-shattered states to pursue a market-oriented growth strategy, rather
than protecting uncompetitive local industries from domestic or interna-
tional competition. The second half of the twentieth century demonstrated
that centrally planned and state-dominated development strategies “ includ-
ing not only Soviet-style communism but also import substitution strategies
pursued in many parts of Latin America and Africa “ generally produced
lower levels of economic growth than market-oriented development strate-
gies. Although debates continue over the appropriate balance between the
market and the state in economic development, there is near-universal agree-
ment today that non-market-oriented economic policies (that is, those that
do not give the market the primary role in allocating scarce resources) are
too inef¬cient to generate sustained economic growth. This is not to say that
market-oriented economic policies are a suf¬cient condition for sustained
economic growth, but that these policies appear to be a necessary condition.
Nevertheless, there are different methods of promoting market-oriented
reform, some of which may be more suited to the needs of postcon¬‚ict
peacebuilding than others. The prevailing model of structural adjustment
is a problematic approach to economic reform in war-shattered states, for
several reasons. First, as we noted earlier, the economic hardships that rapid
adjustment tends to impose on speci¬c population sectors, along with the
regressive widening of distributional inequalities, have been generally asso-
ciated with higher levels of political unrest and violence in states undergoing
Problems and Solutions
200

these reforms. Given the fragile political conditions that exist in most war-
shattered states, particularly in the period immediately following the termi-
nation of hostilities, it is unwise to pursue a strategy of rapid and immediate
adjustment in these states. A better approach would be to delay reforms until
political conditions are less fragile, or to stretch the reforms over a longer
period in order to temper the disruptive effects of adjustment.60 Some com-
mentators argue that rapid reform is essential because slow reform “allows
the [domestic] opposition to coalesce, often in defense of the disproportion-
ate bene¬ts that it receives from public expenditure.”61 While this may be
true, the requirements of successful adjustment must be balanced against the
need for political stability in states just emerging from civil wars. As Thomas
Callaghy puts it, “economic liberalization without attention to domestic po-
litical stability . . . is likely to prevent successful economic and political ad-
justment.”62 This is particularly important for war-shattered states in which
economic deprivation or distributional inequalities have fueled violent con-
¬‚ict in the past.
Second, orthodox structural-adjustment policies make it dif¬cult to pro-
vide the inhabitants of war-shattered states with a “peace dividend,” or tan-
gible improvements in economic conditions, which could help to reinforce
popular support for peace. If peace comes along with material rewards, or-
dinary citizens may be less willing to heed extremist members of their own
community who seek to scuttle the peace. But orthodox adjustment mod-
els are based on a different set of priorities: namely, establishing conditions
for economic growth in the long term, even at the expense of an initial
economic downturn. The IMF and World Bank typically require states
undergoing adjustment to implement ¬scal austerity policies aimed at es-
tablishing budgetary balance, along with reductions in the money supply
designed to keep in¬‚ation in check. But ¬scal retrenchment and monetary
contraction tend to suppress economic growth in the short run, in part be-
cause they increase the cost of credit and reduce the amount of disposable
income in the hands of consumers. As a result, the use of ¬scal or monetary
policy to stimulate immediate economic growth in the aftermath of a civil
con¬‚ict becomes dif¬cult under prevailing models of adjustment. Rather than
enjoying a peace dividend, the inhabitants of war-shattered states that em-
brace orthodox adjustment are assessed what amounts to a “peace penalty”
in the form of policies that arti¬cially suppress growth in the short term, in

60 Paul Collier and Anke Hoef¬‚er (2002) argue, for example, that special postwar recovery
funds should be channeled to war-shattered states for at least a decade, and that placing
emphasis on “social policies,” including funding for education and health care, as opposed
to focusing primarily on macroeconomic reforms, is especially important for preserving peace
in countries that have emerged from civil con¬‚ict. These arguments are elaborated in Collier
et al. 2003.
61 Graham 1995, p. 144. See also Bhattacharya 1997, p. 1047.
62 Callaghy 1993, p. 165. See also Boyce 2002, pp. 48“50.
Institutionalization Before Liberalization 201

exchange for assurances about the possibility of economic improvement in
the long term. By delaying the implementation of these reforms or spreading
them out over a longer period, peacebuilders would have greater ¬‚exibil-
ity to pursue policies that stimulate economic growth in the crucial period
immediately following the war™s end.63
Third, war-shattered states typically lack the institutional capacity to
successfully manage market-oriented reforms. One of the most important
functions that governments can perform in market systems is to maintain a
sound legal framework that is capable of upholding property rights, resolv-
ing business disputes impartially, enforcing contracts, protecting consumers
from fraud, collecting taxes, and regulating the banking system. At the most
fundamental level (as Adam Smith and other classical liberal economists
recognized long ago), a successful market economy presupposes the rule of
law “ that is, an environment in which the “rules of the game” are for the
most part predictable, clear, and enforced in a consistent and disinterested
manner. Promoting economic liberalization in the absence of a sound legal
framework is a recipe for a malfunctioning and inef¬cient market economy in
which the boundaries between business and crime are blurred: for example,
the Russian economy in the 1990s.64
Partly in response to the problematic marketization of Russia and other
former Soviet bloc states, the IMF and World Bank have since the mid-1990s
acknowledged the importance of effective governmental and legal institu-
tions to the success of market-oriented adjustment programs.65 Under the
rubric of “good governance,” the Bretton Woods institutions (along with
major national aid donors) have increasingly made their ¬nancial assistance
conditional on recipient states implementing institutional reforms, including
not only measures to ensure a sound legal framework for the market econ-
omy but also improvements in the management of public resources. The
IMF in particular has emphasized the importance of reforming public-sector
institutions in recipient states in order to increase the ef¬ciency and trans-
parency of their operations and reduce the problems of wastage, corruption,
and misallocation of resources. The World Bank has also pointed out that
administrative competency facilitates the provision of public services upon
which the market economy depends, including education, health care, and
physical infrastructure.
War-shattered states, however, represent a special category of underinsti-
tutionalized states. Within the larger class of countries undergoing transi-
tions to capitalism under the guidance of international donor agencies, war-
torn territories tend to be particularly lacking in functioning legal systems
and governmental institutions. Despite the World Bank and IMF attention
to “good governance,” these organizations remain overly con¬dent in the

63 Jeffrey Sachs (1994) and Jenny Pearce (1998) have made similar arguments.
64 65 IMF 1997; World Bank 1997; James 1998.
Nellis 1999.
Problems and Solutions
202

ability of markets to organize themselves spontaneously, and have tended
to push for economic liberalization in war-shattered states prior to the con-
struction of necessary governmental and legal institutions.66
Nowhere is this problem more apparent than in Bosnia, where liberaliza-
tion programs undertaken in the absence of effective institutions have served,
as Michael Pugh puts it, to “reinforc[e] the dominance of clientelistic and
ma¬a political economies.”67 Following the signing of the Dayton Accord
in 1995, international donors immediately launched economic reforms that
included a far-reaching privatization program aimed at selling off inef¬cient
government-owned enterprises. But in Bosnia™s “especially acute institutional
vacuum,” there was little to prevent the dominant nationalist parties from
manipulating the selling of these public enterprises “to themselves or to their
allies through shady and non-transparent privatization deals.”68 This process
had adverse effects both for the economic and political reconstruction of the
country. Economically, it reinforced the underlying corruption and crony-
ism of the Bosnian economy, which “remains controlled by a political elite
at odds with the very reform policies that would lead to greater openness.”69
This elite reportedly misappropriated as much as $1 billion of international
assistance during the ¬rst four years of peace.70 Politically, it strengthened
the power of the very nationalist groups who are least interested in achiev-
ing interethnic reconciliation in Bosnia. While peacebuilding agencies have
made some strides in building public-sector institutions to govern the mar-
ket economy “ including the capacity of the central and regional govern-
ments in Bosnia to collect revenues, perform audits, and conduct economic
forecasts “ the fact that these agencies launched a comprehensive economic
liberalization program in the underinstitutionalized setting of post-Dayton
Bosnia did little to spur economic growth or promote cross-factional recon-
ciliation.71
Attempts by the Bretton Woods institutions to moderate the social costs
of structural adjustment in war-shattered states have also been inadequate.
In response to criticisms that orthodox adjustment models impose dispro-
portionate costs on the poorest and most vulnerable sectors of transitional
states, both the IMF and World Bank have made much-vaunted but essen-
tially minor changes in their lending policies. These initiatives included not
only the introduction of “poverty assessments” for each recipient state to
evaluate the effects of economic reforms on the poor, but also support for
“social investment funds” in a number of countries undergoing economic
adjustment. Many of these funds mirrored Bolivia™s Emergency Social Fund,
which was established in 1986 in response to the economic and social turmoil

66 67 Pugh 2000a, p. 2.
This critique draws generally on Kolodko 2000.
68 69 O™Driscoll, Holmes, and O™Grady 2001, p. 121.
Donais 2002, pp. 11, 7.
70 71 Donais 2002, p. 2.
Hedges 1999.
Institutionalization Before Liberalization 203

that accompanied structural adjustment reforms in that country, and which
disbursed money to local projects that sought to “create temporary employ-
ment, build social and economic infrastructure, and provide basic social ser-
vices to low-income populations.”72 Among the countries that subsequently
introduced similar social funds are Brazil, Chad, Chile, Ghana, Honduras,
Mexico, Peru, Poland, Senegal, Uganda, Uruguay, and Zambia, as well as El
Salvador and Nicaragua.73
Some commentators argue that these funds have helped to “dull the pain”
of economic adjustment by providing temporary assistance to vulnerable
populations,74 but there is also widespread agreement that they have had
little or no discernible impact on overall rates of poverty or distributional
inequality.75 In fact, one of the World Bank™s vice presidents, Shavid Burki,
acknowledged in 1996 that the Bank™s efforts to minimize the ill effects of
adjustment, including the social investment funds, had largely failed.76 With
the possible exception of the Chilean program, which was “big enough to
reverse the regressive impact of recession and adjustment on the income of
the poor,”77 most social investment funds have lacked the ¬nancial backing
necessary to make substantial headway against the poverty problem in states
undergoing economic adjustment.78 In the period 1988“1993, for exam-
ple, the World Bank committed $4 billion for adjustment programs, while
programs aimed at moderating the ill effects of adjustment received only
$141 million.79 The government of Denmark concluded in 1995 that social
investment funds have been “of little ¬nancial signi¬cance compared to the
overall magnitude of adjustment lending.”80
It should be noted, however, that this state of affairs re¬‚ects the broader
priorities of the international ¬nancial institutions: Social investment funds
do not impel developing country governments to “get the fundamentals
right” or to open their economies to market forces, and are therefore viewed
as peripheral to the adjustment process. The only long-term solution to
poverty and distributional inequality, both the IMF and World Bank ar-
gue, is sustained economic growth; and efforts to alleviate poverty in the
short run, no matter how well intended, risk diverting resources from the
adjustment process and perpetuating inef¬ciencies.
The Bretton Woods institutions have also been criticized, more speci¬-
cally, for treating war-shattered states as any other transitional countries,
rather than as special cases that require policies speci¬cally designed to


72 Jayarajah, Branson, and Sen 1996, p. 105.
73 For a description of the various social funds that have been instituted in these states, see
Glaessner et al. 1994; Graham 1994 and 1995; Vivian 1995; and IMF 1995.
74 75 See Vivian 1995.
Morley 1995a, p. 69. See also Glaessner et al. 1994.
76 77 Morley 1995a, pp. 68“69.
Cited in Cau¬eld 1996, p. 163.
78 79 Denmark 1995, p. 56. 80 Ibid.
Sandbrook 1997, p. 504.
Problems and Solutions
204

moderate, not exacerbate, societal tensions.81 Both agencies responded in
the late 1990s and early 2000s by devising a new set of lending principles
for postcon¬‚ict countries and by pledging to disburse emergency funds to
such countries in an expedited manner.82 The Bank™s new goals were to
“jump-start” the economy through investment in key productive sectors;
to support the reconstruction of governmental institutions, the rule of law,
physical infrastructure, and basic social welfare programs; and to facilitate
the demobilization of combatants and the removal of land mines.83 The
IMF™s revised goals were more limited: to help war-shattered states cope with
urgent macroeconomic imbalances and to assist in rebuilding “the adminis-
trative and institutional capacity required to put a comprehensive economic
program in place.”84 How these policy statements will ultimately be imple-
mented, and whether they will lead to substantive changes in lending prac-
tices for war-shattered states, remain to be seen. But the apparent willingness
of the Bretton Woods institutions to place a high priority on institutional
reconstruction in countries just emerging from civil con¬‚icts is a welcome
development.
Ideally, comprehensive economic liberalization should be delayed while
the governmental and legal frameworks needed to regulate a market econ-
omy are being established. Moreover, economic reforms should generally be
implemented in a gradual and phased manner, rather than all at once in the
orthodox “shock therapy” style, which tends to be politically destabilizing.
Schedules for achieving ¬scal balance and low rates of in¬‚ation should also
be extended in order to reduce pressures on social spending and the funding
of rehabilitation projects during the fragile period immediately following the
termination of hostilities.
Finally, international donors should be more responsive to the fact that
economic reforms that worsen income inequalities can work against the
consolidation of peace in countries with a history of civil violence arising
from distributional grievances. Contrary to the assertions of IMF and World
Bank of¬cials, doing more to promote income equity in such countries need
not involve a trade-off with economic growth: Several recent econometric
studies have concluded that societies with more equitable distributions of
income tend to experience higher levels of economic growth over the long
term.85 Even if greater emphasis on distributional issues did ultimately re-
duce the rate of economic growth, this trade-off would still be sensible if it
diminished the risk of renewed civil war. Furthermore, as we noted earlier, a
strong relationship exists between the distribution of wealth and the stability
of democratic regimes: Democracies with higher levels of income equality


81 82 IMF and World Bank 2001. 83
For example, Boyce 2002. World Bank 1998.
84 IMF and World Bank 2001, pp. 8“9.
85 For a review of this literature, see Birdsall and Jaspersen 1997.
Institutionalization Before Liberalization 205

tend to be more stable than democracies with lower levels of income equality,
independent of their level of economic development.86
At bottom, managing this problem in the context of peacebuilding re-
quires reordering funding priorities and redirecting some donor resources
away from traditional adjustment projects in order to provide much-
expanded support to such redistributive programs as the aforementioned
safety net funds, along with public education and health-care services, start-
up money for “microenterprises” and other job-creation programs in poorer
communities, and perhaps even temporary price controls for food staples.
Whatever speci¬c devices are selected, they should be part of a more serious
effort to moderate the polarizing effects of market-oriented adjustment on
the income distribution of war-shattered states, particularly those in which
distributional inequities have fueled violent con¬‚ict in the past. Attempts to
effect such changes in the Guatemala operation represented a step in the
right direction but were incomplete, as we have already noted.


The Common Denominator: Rebuild Effective State Institutions
All of the recommendations outlined in this chapter are based on the obser-
vation that promoting democratization and marketization in institutionally
weak, con¬‚ict-prone environments is an unreliable and potentially coun-
terproductive approach to peacebuilding. Democratic politics and capital-
ist economics are not self-organizing; they depend on public institutions to
uphold basic rules, to maintain order, to resolve disputes impartially, and
to regulate behavior incompatible with the preservation of market democ-
racy itself. At the most basic level, democracy and capitalism presuppose a
functioning state apparatus. I have described several institutional elements
that peacebuilders should seek to construct before they launch comprehen-
sive political and economic liberalization programs in countries that are
just emerging from civil wars. These elements include a constitutional court
to resolve disputes surrounding elections and to uphold the articles of the
constitution against challenges; a reliable police force to maintain domes-
tic order and enforce the rulings of the constitutional court and any other
court legally constituted under the constitution; a procedure for regulating
hate speech in the media; a system for overseeing the conduct of political
parties and civil-society organizations; electoral rules speci¬cally designed
to reward moderation; a legal framework capable of regulating the market
economy; and a redistributive mechanism to protect the welfare of the most
vulnerable sectors of the population. The common denominator across all
of these recommendations is that institutionalization should precede liberal-
ization to a considerable degree in order to limit the destabilizing effects of
the liberalization process itself.

86 Muller 1988 and 1995. See also Diamond, Linz, and Lipset 1990, p. 19.
Problems and Solutions
206

The ultimate success of the IBL strategy, however, will depend not only
on the willingness of peacebuilders to delay liberalization. The proper se-
quencing of reforms is only part of the prescription for more effective peace-
building. Just as importantly, peacebuilding organizations (and the states
that support these organizations) need to commit the time and the political
and ¬nancial resources necessary to restore functioning central governments
in war-shattered states, sometimes from scratch. If the peacebuilding record
of the 1990s demonstrated anything, it was this: There is no easy, quick,
and cheap method of establishing stable and lasting peace in war-shattered
states. Rebuilding effective governmental institutions, managing a phased
and gradual transition to market democracy, and ensuring that the rule of
law is suf¬ciently strong to defend the new state against inevitable chal-
lenges (including challenges posed by the competitive character of democ-
racy and capitalism themselves) require a more interventionist and long-term
approach to peacebuilding than that which has been practiced to date. They
require international peacebuilders to take on the role of nation builders “
to serve as surrogate governing authorities for as long as it takes to imple-
ment the liberalizing reforms that the peacebuilders themselves prescribe for
war-shattered states.
The direct international administration of war-torn territories is war-
ranted where no functioning governmental institutions exist and have to
be created from scratch, and where institutions do exist but have been co-
opted and corrupted by one of the political factions that fought the war,
such as in Cambodia.87 International administration is the key to creating
a competent, professional, law-abiding bureaucracy. This entails staf¬ng
governmental institutions with international personnel, and then gradually
replacing these of¬cials with adequately trained and politically nonparti-
san locals. Particular attention needs to be paid to the justice and security
sectors. In circumstances where local judges are nonexistent, incompetent,
corrupt, or lacking independence, courts should be operated by international
jurists and lawyers while a new cadre of local judges is being trained. If there
is no criminal and civil law in existence, peacebuilders should create suit-
able legislation and promulgate it. If, at the outset of a mission, serious and
widespread violence is occurring in the absence of a working justice system,
peacebuilders should be prepared to temporarily impose martial law “ as
they are permitted to do under international law.88 The core function of

87 Henry Kamm (1998, p. 252), a longtime observer of Southeast Asia, argues, for example,
that Cambodia™s state and society are still so dysfunctional that the international community
should go back into Cambodia, directly administer the country, and “gradually hand it back
to a new generation of Cambodians, who will have matured with respect for their own
people and will be ready to take responsibility for them.”
88 Article 4 of the International Covenant on Civil and Political Rights permits certain civil
rights to be suspended during a “time of public emergency which threatens the life of the
nation and the existence of which is of¬cially proclaimed.” See O™Neill 2002, p. 96, n 1.
Institutionalization Before Liberalization 207

any state is, as Max Weber put it, the monopoly over the legitimate use of
physical force within a particular territory. The ¬rst task of peacebuilding
is to restore this monopoly as a foundation and precondition for all further
institution-building efforts. As the early liberals recognized, a peaceful and
limited state presupposes an effective state that can, at a minimum, ensure
public security.
Finally, adopting the IBL strategy would require peacebuilding missions
to remain in place for a longer period than was typical for operations con-
ducted between 1989 and 1998. Most of these operations lasted from one
to three years “ the sole exception being the Bosnia mission, which was
initially deployed for one year, but whose mandate was subsequently ex-
tended when it became obvious that ¬ghting would likely resume if peace-
builders departed. Ideally, no time limits should be placed on peacebuilding
missions; they should remain in place until governmental institutions have
been constructed and the process of democratization and marketization is
well under way. Translating the terms of a democratic settlement into insti-
tutional roles and routines is, in the words of Larry Diamond, “gradual,
messy, ¬tful, and slow.”89 The formal U.S. occupation of Germany and
Japan after World War II lasted four years and seven years, respectively,
during which time the institutional structures of democracy and capitalism
were built, but informally, the presence of U.S. troops in these countries
continued to exert a stabilizing in¬‚uence for several decades after the war
was over. The challenges of peacebuilding are in some respects even greater:
Instead of facing a defeated enemy, peacebuilders must typically deal with
formerly warring parties that may still be politically and militarily intact.
Most missions should be planned for at least ¬ve years, with the expecta-
tion that they may last longer. Peacebuilding agencies and their sponsors
should not fool themselves into believing that holding hasty elections, par-
ticularly in the con¬‚ict-prone and underinstitutionalized conditions of most
war-shattered states, will create the necessary conditions for self-sustaining
peace.


Possible Criticisms of IBL
Having set out the main elements of the IBL strategy, the remainder of this
chapter considers a number of critiques that might be leveled against the
strategy. In particular, some observers might argue 1) that peacebuilders
who follow the IBL approach will become bogged down in endless missions;
2) that direct international administration of war-shattered states will create
a “culture of dependency” among local people, who might come to rely on
international of¬cials and lose interest in governing themselves; and 3) that

89 Diamond, cited in Decalo 1992, p. 35.
Problems and Solutions
208

the IBL strategy would be too costly and is therefore impracticable.90 I re-
spond to each of these concerns below.

The Endless-Mission Critique
Could the emphasis on institution building and gradual liberalization result
in operations that effectively never end? This danger certainly exists. Any
postcon¬‚ict operation runs the risk of prolonged deployment if it encoun-
ters problems that impede the ful¬llment of its mandate; and when peace-
building mandates become more complex, the potential obstacles to their
realization increase correspondingly. Because IBL calls upon peacebuilders
to begin the construction of domestic institutions before launching full-scale
political and economic liberalization, the danger of getting bogged down
is probably higher for missions that pursue the IBL strategy than it is for
“quick and dirty” missions that emphasize rapid liberalization and speedy
withdrawal.
There are two responses to this concern. First and most important, the
“quick and dirty” approach to peacebuilding is fundamentally ¬‚awed, as I
have argued. It is less likely to achieve the central goal of peacebuilding “
the creation of a stable and lasting peace “ than the IBL strategy. Indeed,
international agencies in Bosnia discovered this lesson the hard way: They
began under the assumption that the mission could be accomplished in one
year, but soon realized that their departure would likely result in a resump-
tion of ¬ghting and decided to extend the mission. But in the meantime, they
pressed forward with premature elections that reinforced the power of the
least conciliatory elements in the society, which ultimately increased the dif-
¬culty of establishing an effective Bosnian state. If the operation had started
with the goal of reconstructing the country™s political and economic institu-
tions and promoting the growth of cross-cutting civil-society organizations
before holding elections, the task of building a functioning Bosnian state
might be further along than it is today “ even though, paradoxically, pur-
suing the IBL strategy would have entailed a more open-ended commitment
to peacebuilding in Bosnia than the international community was willing
to make at the time. The danger of prolonged deployment is one that must
be faced and accepted if the goal of peacebuilding is to foster a stable and
lasting peace, since the alternative “ starting a mission with the intention of
completing it within a year or two “ creates incentives to rush liberalization
and to terminate peacebuilding prematurely.
Second, the IBL strategy does not preclude the judicious use of timetables
to maintain the forward momentum of a peacebuilding mission. Targets

90 Two additional concerns are addressed in Chapter 11: that IBL presupposes an unrealisti-
cally high degree of coordination among peacebuilding agencies, and that it requires more
“political will” than currently exists within the international community to implement the
strategy.
Institutionalization Before Liberalization 209

for political and economic reforms should be established (subject to revi-
sion, if necessary), and international of¬cials should report regularly on their
progress, explaining failures to meet these targets. Yet such schedules should
be based primarily on the requirements of institutionalization before liber-
alization, rather than on unrealistic demands for the speedy termination of
an operation.


The Culture-of-Dependency Critique
David Chandler has written extensively on the danger of fostering a
“culture of dependency” in populations under international administra-
tion.91 He argues that international authorities in Bosnia, in particular, have
limited the development of indigenous democracy by suppressing local in-
volvement in policymaking decisions and by creating the expectation within
parts of the indigenous population that international of¬cials will continue to
manage their affairs for them. Chandler believes that this approach, which
he views as unduly paternalistic and authoritarian, will ultimately “make
it more dif¬cult for the international community to exit from Bosnia” be-
cause it is paradoxically blocking the development of local capacities for
self-government.92
Chandler™s critique is biting, and it fundamentally challenges the IBL strat-
egy set out in this chapter. It is paradoxical, and in some respects unfortunate,
that international of¬cials must suppress certain forms of political expres-
sion in order to build the foundations for a stable and peaceful democracy.
But the problem with Chandler™s analysis is that he does not explain how
else to reach the end point of stable and lasting peace without managing and
controlling the destabilizing effects of political and economic liberalization
in the short run. I have argued that international peacebuilders have little
choice but to act “illiberally” in the earliest phases of a postcon¬‚ict transition
for precisely this reason, and that the IBL strategy is more likely to produce
a durable peace “ and a stable democracy “ than a strategy that promotes
political and economic competition in the immediate aftermath of civil war.
Encouraging maximum freedom in the short term may be philosophically
more comfortable for peacebuilding agencies and for Chandler, but it ignores
the “Hobbesian problem” of building effective institutions to contain free
competition within peaceful bounds “ just as scholars of the liberal peace
thesis have taken the existence of such institutions largely for granted. This
is the uncomfortable reality that peacebuilding agencies (and the states that
sponsor these agencies) must embrace if they wish to prevent the pathologies
of rapid liberalization from recurring in future operations.
That said, the IBL strategy does not preclude local involvement in the
international administration of a war-torn territory. Indeed, peacebuilders

91 92
Chandler 1999a, 1999b, and 2001. Chandler 1999a, p. 65.
Problems and Solutions
210

could go considerably further than they have in previous missions to in-
corporate local consultation mechanisms into the apparatus of governance.
In both Bosnia and East Timor, for example, international of¬cials created
unnecessarily opaque governance structures that permitted too little indige-
nous involvement,93 yet there is room for much greater consultation and
for experimentation with different forms of local participation in interna-
tional administration,94 which would help to diminish the reality and the
perception of dependency. Furthermore, as noted, decisionmaking authority
should be transferred to indigenous institutions as quickly as possible, and
locals should be trained in public administration and prepared to take over
the management of governmental agencies immediately. In the end, however,
the most sweeping demands for immediate democratization can only be par-
tially addressed, since gradual and phased liberalization in war-torn states
seems better suited to the establishment of a stable market democracy in the
long run.

The Excessive-Costs Critique
Many commentators point to the problem of “donor fatigue,” or the limited
willingness of states to contribute resources to peace operations and human-
itarian endeavors, as a restraining factor in the international community™s
ability to conduct such operations.95 One could argue that IBL™s demand
for longer-lasting and more extensive forms of peacebuilding would be too
costly to be practicable, given this problem of donor fatigue. But, in fact,
IBL would be a more economical investment of scarce resources than mis-
sions that appear cheaper but allow civil violence to rekindle. Scrimping on
peacebuilding offers false economies if operations fail to create the condi-
tions for durable peace. Whether the states and agencies that sponsor peace-
building can be convinced of the need for the up-front investment in IBL
is a different question, and one that I cannot answer, although one of the
purposes of this book has been to make precisely this case.
In addition to highlighting that IBL would, in the long run, be less ex-
pensive than the “quick and dirty” approach “ both in terms of human lives
and money “ it is worth noting that peacebuilding missions to date, includ-
ing the lengthier and more complex missions, have cost only a tiny fraction
of the money spent on national armed forces during the same time. By one
calculation, the total UN expenditure for all of the peace operations in the
1990s (including not only peacebuilding missions but those launched dur-
ing ongoing hostilities) was $19.9 billion, while world military expenditures

93 For example, Chandler 1999b; and Chopra 2000 and 2002.
94 See Beauvais 2001; Chopra 2002; and Hohe 2002a and 2002b. See also Korhonen 2001
on the importance of involving indigenous populations in transitional administrations as a
means of enhancing the perceived legitimacy of these arrangements.
95 For example, Kapp 2001.
Institutionalization Before Liberalization 211

during the same period was an estimated $6.9 trillion.96 For the United
States alone, total federal budget outlays for national defense between 1991
and 1999 was approximately $2.5 trillion, or 125 times greater than the
total UN peace operations budget, to which the United States contributes
25 percent.97 Given the additional costs to the United States of increasing
“homeland defense” in the wake of the September 2001 terrorist attacks,
and the fact (as acknowledged by the U.S. government) that states suffering
from chronic civil violence have become safe harbors for transnational ter-
rorist groups, investing in more effective peacebuilding would make sense
even from the purely self-interested standpoint of American national security,
while it would also enhance regional security and address one of the principal
causes of humanitarian crises in the world: pervasive civil con¬‚ict.


Conclusion
The IBL strategy represents an improvement over the pursuit of rapid liber-
alization, an approach that dominated peacebuilding missions in the 1990s.
Unlike the alternatives of authoritarianism or partition, IBL does not reject
the goals of Wilsonianism. Rather, it seeks to achieve these goals through
different means: gradual and controlled liberalization, combined with the
immediate construction of domestic institutions that are capable of manag-
ing the destabilizing effects of democratization and marketization.
But what has happened with regard to peacebuilding since 1998? Thus far,
I have deferred discussion of operations in Kosovo, East Timor, and Sierra
Leone, all of which were launched in 1999, because their recentness made
the outcomes of these operations more dif¬cult to evaluate than previous
missions. As we shall see in the next chapter, however, these operations have
embraced a few of the elements of the IBL strategy.

96 97
Hentges and Coicaud 2002. Ibid.
11

Lessons Learned and Not Learned
Kosovo, East Timor, Sierra Leone, and Beyond




The peacebuilding operations launched after 1998 in Kosovo, East Timor,
and Sierra Leone have moved gradually in the direction of the IBL strategy:
They have devoted more attention and resources to the construction of ef-

<<

. 7
( 10)



>>