Can Foreign Aid Moderate Ethnic ?

Milton J. Esman

UNITED STATES INSTITUTE OF PEACE CONTENTS

Summary v

Preface vii

1 Introduction 1

2 The Ethnic Context and Consequences of Foreign Aid 2

3 The Principal Actors and Components of Development Assistance 4

4 The Culture and of Development Assistance 6

5 The Goals of Intervention 9

6 Policies to Avert or Moderate 11

Notes 15

About the Author 17

About the Institute 18 v

and economistic . Economic resources and their efficient utilization have long been seen as the proper concerns of donors; attention to political issues and impacts has been strongly discouraged, with interfer- ence in the internal political affairs of member states expressly forbidden in the case of multilateral organi- zations such as the World Bank. Yet although most agencies (especially at their headquarters) remain SUMMARY allergic to political concerns in general and to ethnic realities in particular, some departures from this apolitical approach have recently been made, with organizations such as USAID promoting democratic development. Alongside concerns to promote democ- racy and , a concern to address the eth- nic dimension of foreign aid is also developing, albeit slowly and cautiously. The donors’ dilemma—whether to adhere to an apolitical stance and ignore ethnic problems or, by ignoring them, to compound them—is complicated by their uncertainty about the goals they should pursue and the methods they should employ. Should they aim through their aid to contribute to peaceful inter- ince World War II, a complex network has ethnic coexistence, should they strive to produce a emerged of bilateral and multilateral agencies sense of interethnic fairness in the distribution of re- Sthat manage economic assistance to low- sources and opportunities, or should they be content income countries in the form of investment projects, that their interventions inflict no harm on any ethnic policy advice, and technical assistance. Although each community? Should they promote allocations on the of these agencies has its distinctive personality, most basis of individual merit and competitive perfor- of them have avoided facing up to the post– mance, of interethnic proportionality, or of compen- reality of burgeoning ethnic conflict. satory preferences to members of disadvantaged This reluctance to deal with the violent conse- communities? There are no simple answers to these quences of ethnic pluralism is often self- defeating, for dilemmas. such violence can destabilize the environment within If development assistance agencies are to address which the agencies operate and thus undermine their the conflict potential of ethnic pluralism construc- efforts. The same reluctance can also blind agencies to tively, they must recognize that there are no standard the damage that the ill-considered provision of foreign formulas for managing ethnic conflict and that con- aid can inflict on ethnic relations within aid-receiving text conditions the effectiveness of interventions. . Some interventions may have neutral or even Thus, agencies must first equip themselves with as positive sum outcomes. Too many, spurred by the much reliable information about a host country’s eth- logic of “developmentalism” that ranks economic nic dynamics and relations as they routinely gather growth and modernization above values such as about local economic conditions. For each proposed democracy and interethnic fairness, have distribu- policy or project intervention, they should prepare an tional effects that generate or aggravate ethnic conflict, “ethnic impact statement” similar to the environmen- as one community is perceived by its rivals to be bene- tal impact statements that are now required. The fiting at their expense. Host governments—the critical ethnic impact statement should detail the ethnic land- intermediaries between assistance organizations and scape of the host country and assess the likely effects the local communities targeted to receive assistance— on ethnic relations of the proposed intervention. The often ensure ethnically destabilizing outcomes by de- statement should be prepared after negotiating with liberately skewing the benefits of development aid in governments to attach and enforce conditions to the favor of a particular ethnic community. resources the donors provide, so that the interven- The culture predominant among development as- tions do not worsen and may in fact contribute to the sistance institutions has a pronounced technocratic reduction or moderation of ethnic-based conflict. vii

attempt to explain why foreign aid agencies have been so reluctant to address this problem; the authors also suggest measures that might be employed in the fu- ture to mitigate the severity of ethnic conflicts in the countries that receive foreign aid. Although the views expressed in this short mono- graph were inspired by deliberations and debates at the 1995 conference, the author is solely responsible PREFACE for the opinions here advanced. I am especially in- debted to my colleague and partner in this enterprise, Ronald J. Herring.

his paper represents the author’s conclusions from a conference held in October 1995 at TCornell University on the effects of develop- ment assistance on ethnic conflict. Revised versions of the papers presented at that conference, which was financed in part by the Institute of Peace, are being prepared for publication by a univer- sity press under the title Development Assistance and Ethnic Conflict. Academic observers have begun to focus on the in- ternational dimensions of ethnic pluralism since the collapse of the multiethnic Soviet empire and Yugo- slavia drew worldwide attention to this phenomenon and its potential for violent and destabilizing conflict. The Carnegie Corporation has supported a series of conferences at Cornell on various aspects of this sub- ject. Papers presented for the first of these confer- ences were gathered together in a volume edited by Milton J. Esman and Shibley Telhami; entitled Inter- national Organization and Ethnic Conflict, that vol- ume was published in 1995 by Cornell University Press. In reviewing the literature, it became clear that the implications of international development assistance for ethnic conflict have not been systematically exam- ined, even though development assistance, from bilateral and multilateral sources, has become an im- portant factor in . The contribu- tors to Development Assistance and Ethnic Conflict 1

on weak and vulnerable . As a result, “so- cial soundness” or social-cultural compatibility was incorporated into the guidelines of several develop- ment assistance agencies.1 Projects and policies were expected to take account of the values, preferences, life-styles, and capabilities of the publics they affected; the harm inflicted on any group should be held to a minimum. Although social soundness is now a recog- INTRODUCTION nized concern of development assistance agencies and is included in their operating instructions, it has 1 not been fully institutionalized and is honored as much in the breach as in the observance. It has sel- dom been extended to such “political” factors as ethnic solidarities and their implications for develop- ment assistance.

n most countries that receive development assis- tance—in Asia, Africa, and Latin America, in East- Iern Europe and the successor republics of the former —ethnic pluralism has become an important dimension of . It is, indeed, often the most salient of factors in the political equation. Does this saliency present a special challenge to the providers of development assistance? If so, how should they deal with it? Where ethnic divisions have become politicized, can donors continue to assume the existence of an integrated national economy where economic growth will raise all ships, or of a so- ciety of individualistic economic maximizers willing and able to participate in market competition, or of a government committed to equity among its diverse citizens and subjects? Or must they reach beyond technical rationality, beyond macroeconomic vari- ables, to the actual structures, values, and political dy- namics of the societies in which they intervene? Can economic development proceed under conditions of political turbulence induced by ethnic conflict? If not, must foreign assistance directly confront this reality? And has foreign assistance the capacity to prevent, alleviate, or help to resolve such disputes? Two decades ago, the reluctance of development assistance agencies to consider the social impact of their interventions was held responsible for the fail- ure of many projects and for inflicting needless pain 2

agencies can equip themselves to estimate the impact of proposed interventions on ethnic politics. By intent or by inadvertence, development assis- tance produces changes, including changes in aspira- tions and expectations that can increase competition and hostility between ethnic communities. Contex- HE THNIC ONTEXT tual evaluation can reduce the uncertainties inherent T E C in foreign assistance and in induced societal change. AND CONSEQUENCES It can enable the adjustment or revision of proposed intervention strategies so that the resulting changes 2OF FOREIGN AID prevent unintended harm or even ameliorate intereth- nic relations. It is a precondition for the conversion of ethnic sensitivity to viable intervention strategies.

DISTRIBUTIONAL CONSEQUENCES

Externally introduced resources and externally in- duced policy changes have distributional conse- quences; they affect ethnic communities differently. During the 1950s and 1960s the test of effective for- eign assistance, whether based on market processes THE IMPORTANCE OF CONTEXT or on state planning, was macroeconomic expansion, measured as growth in gross national product (GNP). In developing sensitivity to ethnic pluralism, context Beginning in the late 1960s, it became evident that the is critical. Simplifying abstractions such as those that benefits of aid-induced growth had failed to trickle facilitate macroeconomic analysis and prescription down to large numbers of people, in many cases even are not useful for evaluating social and political reali- to majorities. In spatial terms, “urban bias” was found ties. Every societal environment is distinctive, and to privilege urban centers to the detriment of rural that distinctiveness at national, regional, and local lev- societies. In class terms, the urban and especially the els must be appreciated if development assistance is rural poor had been mostly bypassed.2 to have beneficial rather than detrimental effects on To correct these distributional , development interethnic relations or on relations between ethnic assistance was expected to target the poor majority communities and governments. Among the signifi- and to satisfy basic human needs, even if this necessi- cant contextual factors are the identities and de- tated some sacrifices in macroeconomic growth.3 mographies (numbers and geographic distribution) This antipoverty strategy influenced development of the principal ethnic communities; their relative assistance providers throughout the 1970s. The anti- power, economic roles, social status, and relations poverty thrust that explicitly recognized class and re- with government; whether interethnic relations are gional, but not ethnic factors was displaced during the stratified or segmented (in other words, are members 1980s. Forced to confront the debt crisis that disabled of particular ethnic communities confined to particu- so many Third World economies and under pressure lar socioeconomic levels or are all ethnic communities from the Thatcher-Reagan brand of economic liberal- represented at all socioeconomic levels?); the extent ism, the aid community reverted to neoclassical eco- to which ethnic communities are politicized, mobi- nomic orthodoxy, promoting stabilization and lized, or passive; divisions or factions within the growth, marketization, privatization, and minimal ethnic communities; their dominant values and capa- government. Distributive concerns were superseded bilities; evidence of interdependency and of crosscut- during the era of structural adjustment and economic ting affiliations and memberships; and the recent liberalization—an era that persists to this day. history of relationships among ethnic communities As a minor theme in the implementation of and between them and the state. In possession structural adjustment, a number of agencies (most of such basic information, development assistance notably, the World Bank) have recognized that some 3

components of this strategy, including the elimina- The ideology of “developmentalism” claims broad tion of food subsidies and reductions in public ser- support among staff members of development assis- vices, can hurt some of the poorest and most tance agencies and host governments. According to vulnerable people in the aid-recipient countries. this doctrine, in low-income countries economic These agencies therefore include in their funding growth and modernization must be the main goals of packages provisions intended to maintain minimal public policy, superseding when necessary other val- safety nets for these victims of fiscal retrenchment. ues, including democracy, human rights, and inter- The differential impacts of development assistance ethnic fairness. The temporary sacrifice of these latter on ethnic communities can occur by inadvertence. values may be the price that must be paid for rapid Modernization can threaten the cultures of indigenous economic expansion and social development from societies whose members are ill equipped to cope with which all citizens will eventually benefit. Once self- market economics. Foreign private investment pro- sustained growth has been realized, as in Chile and moted by the International Monetary Fund (IMF) and among the Asian “tigers,” then public policy is better the World Bank has, in several countries, jeopardized equipped to accommodate the demands of aggrieved the livelihoods and threatened the way of life of indige- ethnic communities. nous ethnic communities on the resource frontier. Some interventions may prove to have ethnically Native peoples have resisted, often by force, the donor- neutral effects. A few may be so managed as to pro- supported “transmigration” of Javanese settlers to the duce positive sum outcomes. All too many have distri- outer islands of Indonesia, charging the settlers with butional effects that generate or aggravate ethnic encroachment on their lands and livelihoods. The ex- conflict, as one community is perceived by its rivals to acerbation of ethnic conflict in and its suc- be benefiting at their expense. Agencies may attempt cessor states has been traced by an expert observer to to mitigate these effects: for example, the World Bank inept efforts by the IMF to achieve economic stabiliza- now requires compensatory measures for communi- tion.4 None of these conflict-inducing consequences ties displaced by dams and other large aid-financed was intended by aid donors. Differential impacts can, infrastructure projects.6 The agencies, however, may however, be the product of deliberate manipulation by calculate that negative effects on interethnic relations governments. President Moi in Kenya has attempted can be offset and thus justified by the economic bene- systematically to use aid-provided resources as patron- fits produced by projects or policies. Yet, given the age to members of his regime’s ruling coalition.5 These knowledge that is now available, development assis- abuses have prompted development assistance agen- tance agencies can no longer be excused for blunder- cies to apply a variety of tactics to negate that govern- ing into ethnic quagmires as innocents abroad. ment’s distributional biases. 4

of NGOs. The shrinkage of government and economic marketization, which at present are vigorously pro- moted by the development assistance community, can limit but not eliminate the intermediary role of gov- ernments. THE PRINCIPAL In the face of protests from ethnic communities CTORS AND that their members are hurt by the withdrawal of food A subsidies or the encroachment of foreign corporations COMPONENTS OF on their lands, governments may plead that they are helpless, having been forced by foreign pressures to 3DEVELOPMENT accept these harsh measures, and that blame lies with the development assistance agencies. Such pleas, how- ASSISTANCE ever, are seldom credible to the victims of such poli- cies, who believe that their government, no matter how weak it might be, has the capacity if not the will to block, defer, or at least attenuate such measures. Gov- ernments cannot escape their role as the gatekeepers of development assistance. The state normally serves as agent of the ethnic community or of a ruling coalition of ethnic communi- ties that has captured the state apparatus. These rulers THE MAIN ACTORS distribute the fruits of development assistance dispro- portionately, even exclusively, to their constituents. The principal actors in this encounter between foreign When the dominant ethnic community is a demo- aid and ethnic pluralism are development assistance graphic majority, government can privilege its con- organizations and their agents on the one hand, and stituents by formal majoritarian democratic processes. ethnic communities and their spokespersons on the When the dominant community is a minority, cruder other hand. The critical intermediary is government. methods are employed. Foreign assistance normally reaches a society through Where the benefits of development assistance are— the state as an expression of the latter’s sovereignty. or are widely perceived to be—systematically skewed Development assistance agencies must negotiate with in favor of a particular ethnic community, resentment governments the terms and conditions of the re- and grievances will certainly ensue among those who sources the agencies provide. Once governments ac- believe they have been cheated or left out. Develop- cept foreign assistance, the resultant projects and ment assistance in such cases can exacerbate ethnic policies are administered through their bureaucratic conflict and may precipitate violence, despite the inno- agencies. Many aid-recipient countries depend on for- cence or good intentions of donors. What develop- eign assistance to finance their development budgets, ment assistance agencies can do to preclude or correct to insure their fiscal stability, to balance their external the biased distribution of the resources they provide payments, and to reassure foreign investors, yet gov- or the policies they promote is considered later in this ernments at their discretion can set boundaries for essay. their acceptance of foreign aid or exclude it altogether. The fact that donor agencies are interested in “moving THE COMPOSITION OF DEVELOPMENT money,” in maintaining their programs, and in protect- ASSISTANCE ing the creditworthiness of client states leaves host governments with considerable bargaining strength. Development assistance provides funds, or more pre- When development assistance agencies choose to cisely, supervised lines of credit to implement projects work through foreign or indigenous non-government or policies. organizations (NGOs), the role of the state may be re- Projects may run the gamut of activities from the duced.7 Yet, such arrangements require the acquies- “productive” sectors (agriculture and industry), to cence of governments, because they retain the power physical infrastructure (roads, dams, water-supply to proscribe or circumscribe the scope of the activities systems, airports), to social infrastructure (education, 5

health, family planning, public administration). They large-scale collective violence—the victims of which, as may be very large, “lumpy” activities such as electric- in and Bosnia, may have been targeted on eth- power systems or relatively small, divisible facilities nic grounds. Such assistance is linked to development such as community health services. by the subsequent efforts at rehabilitation to assist vic- Policies are governmentally enforced measures that tims, including , to rebuild their lives and to affect the entire society or particular sectors of the become economically productive and self-sufficient. economy. The most obvious are the macroeconomic Development assistance not only furnishes finan- policies associated with economic stabilization and cial resources but also funds specialists who provide structural adjustment—increasing revenues, reducing training or advisory services, as well as educational public expenditures, devaluing the currency, privatiz- opportunities in donor institutions. The purpose of ing the ownership and management of enterprises, technical assistance is to expand the knowledge, eliminating protectionism, promoting foreign private skills, and capabilities of individuals and institutions. investment, shrinking the role of government in the Since they introduce fresh resources or allocate ex- economy and society. Of growing importance are isting resources in new patterns, all these modalities macropolitical measures relating to democratization of development assistance have potential implications and human rights; these include measures to promote for ethnic conflict. Some interventions may be con- freedom of expression and of political organization, ducive to peaceful coexistence and equity; others may free and fair elections, the , and protection aggravate tensions and precipitate conflict. Some pre- of minorities. sent cruel dilemmas. Food and medical supplies in- Conditionality can be applied by donors to any tended to sustain the victims of civil wars may be flows of assistance. Of particular interest are the con- hijacked by military contingents of their own ethnic ditions attached by assistance providers to policy- community, leading their enemies to interdict all related transactions. Budget support by the World humanitarian assistance. Privatization, intended to Bank may be contingent on tax reform; IMF loans to enhance economic efficiency, may be perceived or in- support the balance of payments may be forthcoming deed have the effect of favoring members of one ethnic only if the host country government agrees to devalue group over others. Majoritarian elections may con- its currency or reduce its budget deficit; consortiums demn an ethnic minority to structural subordination of donors may release development assistance funds and . Resources provided in the project only on condition that the host government allows op- mode may be diverted by governments to favor fellow position parties to function and conducts honest elec- ethnics, while reductions in public expenditures may tions. To enforce conditionality, assistance agencies be administered in ways that spare one ethnic commu- monitor the performance of governments, releasing nity while imposing costs on others. Development as- funds only in periodic increments depending on com- sistance can, however, be designed to minimize such pliance. negative consequences, to mitigate the effects of devel- Humanitarian assistance is intended to provide re- opment projects on vulnerable ethnic communities, lief and sustenance to victims of natural disasters or or to promote positive sum interethnic equity. 6

calculations, the implementation of development as- sistance has assumed a decidedly apolitical cast. The multilateral agencies have been especially deterred by their charters from involvement in the politics of their member countries. Considerable attention has been paid to microeco- THE CULTURE AND nomic factors such as the efficiency and output of indi- INSTITUTIONS OF vidual projects, but the master criterion for success has been macroeconomic growth. For several decades, DEVELOPMENT this approach was warranted by the ascendancy after 4 World War II of the Keynesian paradigm and the evi- ASSISTANCE dence in Europe and Japan during the twenty years after the Marshall Plan that macroeconomic growth seemed, in fact, to lift all boats. When, as noted above, that impressive growth was found to bypass large numbers of people, questions of distribution were in- troduced into the culture of development assistance. Although those protected from harsh distributional consequences may have turned out to be members of minority ethnic communities, this was often acciden- tal. Explicit ethnic concerns tended to be defined as “political,” and thus beyond the proper purview of de- or nearly half a century development assis- velopment assistance. tance has been a significant presence in inter- The end of the Cold War has introduced into the Fnational affairs.8 It has touched nearly all universe of development assistance fresh themes that countries as contributors or recipients and generated challenge its established culture. The culture of for- intense controversy between contributors and recipi- eign aid has begun to make room—albeit cautiously ents and within contributing countries. As develop- and selectively—for political values. Concerns such as ment assistance institutions, bilateral and multilateral, human rights and democratization are now vigorously gained experience in this novel enterprise, they promoted by the United States Agency for Interna- evolved a culture—a set of beliefs and practices, and a tional Development (USAID), the Netherlands, and special vocabulary—that guided their behavior and the Scandinavian countries. These concerns cannot be into which new recruits were inducted and socialized. written off as nonpolitical, and for that reason the A prominent theme in this culture has been its tech- World Bank limits itself to fostering improved “gover- nocratic and economistic bias. As the common goal nance,” but not democratization. Where human rights was believed to be economic development, economic violations by weak states are conspicuous and fla- resources and their efficient utilization were seen as grant, as in Kenya, the development assistance com- the proper concerns of donors, supplemented by the munity is prepared to impose conditions on aid flows, enhancement of skills, transfer of technology, and demanding specific political reforms; where human strengthening of institutions required for economic rights violations and restrictions on democratic development. Formal deference to state sovereignty processes by more powerful states are similarly bla- has been a component of this culture. Many recipient tant, as in and Indonesia, development assis- governments, recently liberated from colonial control, tance agencies have found it expedient to overlook have been especially jealous of the symbols of their them while continuing to provide substantial assis- sovereignty, and their assertion of this newfound inde- tance. When the Indonesian government expelled the pendence has not been challenged by the technicians Dutch foreign aid agency for protesting human rights and economists who have staffed the donor institu- violations, an aid consortium chaired by the World tions and who tend to be uncomfortable with political Bank made up the loss of Dutch assistance. Following matters. While decisions about which countries the logic of developmentalism, economic growth took should be assisted and the volume of assistance have precedence over concern for human rights and justice often been based on political, strategic, or commercial for ethnic minorities. 7

While some political factors have intruded into the selected from university graduates from the left of cen- practice of development assistance, this has been chal- ter of the political spectrum. Whereas the IMF works lenged as inappropriate interference both by leaders only with governments and investors, OXFAM works of developing countries and by some staff members of directly with disadvantaged communities and is sensi- the development institutions. Japan, now the world’s tive to the needs of ethnic communities that are op- largest aid donor, is reluctant to impose political con- pressed or disadvantaged by governments or by the ditions. Donors are aware that ethnic tensions may results of development assistance. Thus OXFAM’s in- undermine their assistance programs, but consistent stitutional personality contrasts sharply with that of attention to ethnic concerns has been slow to emerge. the IMF. Although USAID has been the lead agency in promot- OXFAM exemplifies one tendency in the heteroge- ing democratic development, its interest in ethnic neous ranks of the NGOs. While some NGOs are politics has not been institutionalized.9 USAID has content to promote humanitarian or development ac- declined to act in situations that might have promoted tivities in a strictly nonpolitical mode, others have be- democratic development, fearing that intervention in come concerned with distributional equity among the ethnically sensitive situations might be construed as publics they assist, including ethnic communities. Ini- too “political.”10 tially apolitical, they have been converted to advocacy The allergy to political concerns in general and to for disadvantaged communities by their experiences ethnic realities in particular is more evident in institu- in the field. NGOs in Ecuador, for instance, facilitated tional headquarters than among field personnel. ethnic mobilization against structural adjustment Headquarters may be constrained by the language of measures and attempted to replace some government their institutional charters and remoteness from the services to the poor that had been eliminated.11 Some incidence of conflict, whereas field staff must find NGOs, notably sectarian agencies, may exacerbate ways in their daily operations to cope with the realities ethnic tensions by favoring one community over an- they encounter. Especially when humanitarian opera- other. A well-informed and sympathetic participant- tions encounter , the resourcefulness observer of NGO operations reaches the surprising of field staff is tested to the limit. In the absence of conclusion that NGOs have seldom been effective in agency policies and guidance on the management of mitigating interethnic conflicts: assistance under conditions of ethnic tension and con- flict, field staff are left to their own devices. [T]hrough operational dilemmas encountered in Each of the institutions that supply and manage de- providing aid, NGOs—whether focused on relief, velopment assistance has its own distinctive person- development, human rights, or peace—have, to a greater or lesser extent, inadvertently exacerbated ality, a personality shaped by the ’s main rather than lessened it [conflict] and its conse- constituency, assigned mission, and sources of quences. In some cases the negative consequences staffing. Differences in personality are reflected in have been profound and costly.12 approaches to ethnic conflict. The IMF considers its main constituency to be cen- This observer cites as one of many examples the tral bankers, finance ministries, and the international tragedy of the camps in Goma, , near the investment community. Its main mission is to help Rwandan border. There NGO supplies intended for governments achieve fiscal stability, manage their ex- destitute refugees were distributed through refugee ternal payments accounts, and maintain their credit- “leaders.” These leaders proved to be officers of the worthiness. The IMF’s staff come from the ranks of Hutu militia whose fighters had committed genocidal central bankers, public finance specialists, and fiscal atrocities against unarmed civilians. The militias and monetary economists. It is not surprising that in- were using the camps and the supplies provided by ternal distributional questions, including the status of NGOs for conscription, training, resupply, and suste- ethnic communities, are not among their priorities. nance in preparation for the reinvasion of Rwanda In extreme contrast, OXFAM is a voluntary NGO and the renewal of against the Tutsi regime.13 whose constituents are mainly concerned with social Most government-sponsored development assis- justice and human rights. OXFAM’s mission is to con- tance institutions lie between the IMF and OXFAM tribute to grassroots development, especially among on the cultural spectrum. The World Bank has be- poor and disadvantaged publics, and to help these come the world’s largest and most prestigious devel- groups assert their human rights. Its staff members are opment agency. Although dominated by engineers 8

and economists who emphasize technical and eco- construed as “political.” Yet, under pressure from nomic rationality, its ranks include a minority of so- Japanese public opinion and international human cial scientists who continue to raise distributional rights organizations, JICA did withdraw from the Sar- issues. These concerns are sometimes reflected in the dar Sarovar dam project in , which threatened to World Bank’s decisionmaking and in its loan provi- inflict severe privation on a large indigenous ethnic sions. As its charter forbids it to interfere in the inter- community. nal political affairs of member states, its management Because of the United Nations’s one-country–one- tends to avoid matters such as interethnic relations vote system of governance, most UN-related agencies that could be construed as explicitly political. remain cautious about tackling matters—including USAID has, in the past, been a pioneer in fresh ap- ethnic issues—that might provoke the displeasure of proaches and a bellwether among development assis- member-states. The main exception is UNICEF, tance institutions. Its main constituency is the U.S. whose mandate emphasizes services to disadvantaged Congress. USAID uses its diminishing resources in children and whose constituency includes large num- support of a variety of causes that are dear to mem- bers of private citizens from whom it raises funds and bers of Congress of different political persuasions; who are attracted to its people to-people and humani- these causes include promoting private enterprise, tarian image. In many respects UNICEF’s outlook re- providing humanitarian relief, and, most recently, sembles the distributional orientation of many NGOs. promoting democracy and human rights. Its current Cultures change, usually gradually, as societies are guidelines for “Democratization and Governance Pro- forced to confront fresh problems. So do institutions. grams” refer to ethnic conflicts but provide scant in- The present development assistance culture and the structions for USAID staff on how to proceed. It has institutions that embody it are made up of several begun, albeit hesitantly, to grapple with ethnic reali- components. The largest component is the orthodox, ties, as in its fostering of ethnic in Ethiopia. apolitical, technocratic-economistic growth strategy; The Scandinavian aid programs, reflecting the so- this strategy continues to underpin the culture as a cial democratic values in Scandinavian societies, have whole, which is why social soundness concerns are been prepared to use their resources and their influ- often overlooked. A second, smaller component is the ence to stress social equity and human rights. These distributional concern that claims significant, but mi- social democratic values include considerations of in- nority support. The smallest element, and one strug- terethnic equity as a component of democratic politi- gling for wider recognition, is made up of concern to cal development. advance democratization and human rights values. The attitude of Japan’s International Development To shift the metaphor: The contemporary develop- Agency (JICA) toward “political” matters is an expres- ment assistance culture includes themes that compete sion of Japan’s sensitivity toward foreign, and for attention and support among decisionmakers, especially Asian, perceptions of Japanese ambitions— practitioners, and the publics that finance them. The perceptions colored by Japan’s harsh treatment of greater the influence that distributional and human other Asians during its era of imperialist expansion rights themes come to enjoy, the more likely it is that before and during the Pacific War. Therefore JICA ethnic concerns will gain legitimacy and be factored prefers to avoid taking any measures that could be into intervention strategies. 9

Assuming that donor agencies have gained some sensitivity to the ethnic dimensions of development, or have been compelled to confront them, what should be the goal or goals of those agencies’ inter- ventions? Once violence has erupted, agencies might be satisfied to limit its scale and intensity and to allevi- ate human suffering. But given the opportunity to consider longer-term outcomes, there are three possi- THE GOALS OF ble goals:

INTERVENTION - to avoid conflict and ensure peaceful coexistence; 5 - to achieve equity or rough distributional justice; or - simply to do no harm to any ethnic community. The pursuit of each of these outcomes is fraught with dilemmas and complications. Economic growth theorists believe that restraining growth in the interest of distributional fairness is a misguided, short-term strategy.16 Development assis- tance should, therefore, not be distorted by political objectives, especially because policies conducive to here is an expanding body of knowledge— these ends are surrounded by high levels of uncer- though no general theory—on the origins and tainty and may in any case be beyond the capabilities Tmanifestations of ethnic conflict and on of development assistance. These theorists argue that processes for its regulation and management. It is the best hope for mitigating conflict—class conflict or based on analysis and evaluation of a large body of ex- ethnic conflict—is a vigorously expanding economy perience.14 But the application of this knowledge to that permits the distribution of increments of growth the circumstances of specific conflicts, especially to all competing communities. Because the introduc- when mediated by development assistance institu- tion of ethnic criteria is likely to raise project costs or tions with their special cultures, must be tempered by constrain the implementation of liberalization poli- cautious and informed judgment. Development assis- cies, the net effect of such extraneous considerations tance agencies have not displayed conspicuous hu- is to hobble economic growth and, perversely, thwart mility and prudence when making interventions in the one strategy that its proponents think is most societies they understand imperfectly. Yet humility likely to mitigate and resolve ethnic conflict in the and prudence are precisely what are needed for inter- long run. ventions designed to influence so complex and uncer- Skeptics, including this author, reply that develop- tain a subject as interethnic relations. Once the ment assistance can contribute to sustained economic specific context has been evaluated and efforts have growth only when other conditions are simultane- been made to consult the parties that might be af- ously present. Among them is political stability, which fected, each instance of intervention must be re- is unlikely under conditions of intense ethnic conflict. garded as a hypothesis about the consequences that Nor is there any convincing evidence to support the might ensue.15 Moreover, interventions should be conventional wisdom that economic growth necessar- monitored to account for unanticipated conse- ily diminishes conflict.17 Many ethnic conflicts have quences and to detect opportunities for timely nothing to do with economic considerations, and corrections. competition over the distribution of growth may actu- These cautions should serve as an ever-present re- ally aggravate conflict.18 Political goals are attained minder, not that the reality of ethnic conflict should primarily by political means. be ignored, but that interventions dealing with such The objectives of conflict avoidance and of equity conflicts must be carefully and deliberately measured. are, under some circumstances, contradictory. It may 10

not be possible to pursue both at the same time. tives for extremist leadership and uncompromising Efforts to achieve distributive justice on behalf of low- claims, and tend to abridge the rights of individuals to status, disadvantaged groups can actually prolong freedom of choice.23 In an effort to attenuate ethnic conflict and cost lives.19 Members of more-favored solidarity, they would promote crosscutting organiza- communities charge that is in effect tional memberships and policies that emphasize indi- , depriving them of opportuni- vidual rights rather than ethnic affiliation. ties they have earned by merit and hard work, and Consociationalists and others who argue for legit- benefiting the undeserving.20 Solicitude by NGOs imizing ethnic solidarities as political actors where and other outsiders for depressed communities may ethnic cleavages seem deep and enduring believe they provoke envy from those who are little better off. De- are recognizing political realities and responding to velopment assistance targeted to Indians in Ecuador manifest social preferences.24 They hope to achieve provoked hostile reactions by low-income Mestizos.21 consensual patterns of power sharing where this is Blue-collar whites in the United States have similarly possible, and guarantees for where responded to affirmative action for designated racial this is necessary. The Anglo-American confidence that and ethnic minorities; upper- Hindus in India individuals are the only legitimate claimants to hu- have protested violently against reservations of gov- man rights cannot and should not be imposed dog- ernment jobs and university admissions for members matically on societies where collective solidarities of “backward” , many of whom are by no means are paramount; in such societies the emphasis on in- impoverished. Such measures aimed at equity can dividual as opposed to group rights cannot be a suc- precipitate backlashes that exacerbate conflict. cessful prescription for managing deep-seated ethnic Similarly, the apparently minimalist goal of causing conflicts.25 no harm to any ethnic community may have the unin- The choice of strategies should depend on which tended effect of provoking conflict. Although not approach—recognition or deemphasis—is most com- harmed in an absolute sense, a mobilized ethnic com- patible with local preferences and which would be munity, aware that others are gaining at its apparent most likely to produce social peace and distributive expense, can readily succumb to the malady of invidi- justice. Because the goals of distributive justice and ous comparison, and its consequent sense of relative peaceful coexistence may not be compatible, at least deprivation may produce aggressive reactions.22 in the short run, donors should be prepared to con- These dilemmas illustrate the maxim that good inten- front the likely trade-offs. A powerful case can be tions and good works can yield bad consequences. made for according priority to social peace even at the Should development assistance activities aim to expense of distributive justice, since peaceful coexis- deemphasize, even delegitimate ethnic solidarity in tence is prerequisite to the realization of other values. the hope that other, presumably less violence-prone, By the same token, donors should accept as legitimate collective identities may emerge as more salient existing expressions of ethnic solidarity, instead of at- sources of political alignment? Or should foreign aid tempting to transcend them, as there is little evidence agencies recognize ethnic solidarities as enduring and that strategies intended to break down ethnic solidari- legitimate allegiances, while promoting measures con- ties (especially when those strategies are promoted by ducive to peaceful, consensual coexistence? These outsiders) can be effective except over very long peri- two approaches are mutually exclusive, since they vi- ods of time. President Tito’s campaign to create an sualize contradictory political and societal futures. overarching Yugoslav identity failed completely; the Those who advocate deemphasis argue that ethnic merging of Normans and Saxons into Englishmen re- alignments are inherently unstable, provide incen- quired four centuries. 11

consensus may not be possible to reach, the commu- nities affected will know at least that an effort has been made to solicit their views and take them into account. Dialogue does not necessarily yield agreement or even mutual respect, and spirited differences among partic- ipants may endure. But the search through dialogue OLICIES TO VERT for common interests and mutual accommodation P A can increase the legitimacy of the process and reduce OR MODERATE the likelihood of conflict. 6ETHNIC CONFLICT ALLOCATION FORMULAS Among the important policy choices for development assistance agencies are allocation formulas. These ap- ply both to projects and to policies. The first such formula is the search for common in- terests. The ideal policy or policy set produces positive sum outcomes for all the parties concerned and mu- tual confidence that benefits and costs are equitably shared. USAID’s support for rehabilitation of the large Gal Oya irrigation system in that served both Sinhalese and Tamil farmers successfully incor- here are no standard formulas for managing porated this objective.26 To be avoided are interven- ethnic conflict. Unlike the IMF formula for tions that, whatever the intended macroeconomic Teconomic stabilization, which is believed to benefits or anticipated economic rates of return, will be equally valid for Bolivia, Uganda, and , one be perceived as benefiting one community at the ex- size cannot fit all. Context conditions the effectiveness pense of another. of interventions. To appreciate the ethnic dynamics of The second formula calls for divisibility. Where eth- the society in which they intend to intervene, develop- nic communities occupy territorially separate en- ment assistance agencies must first investigate and claves, the divisibility of projects can contribute to learn. The first step should be to examine the recent mutually tolerable results. Thus, each major ethnic history of ethnic relationships; the next should be to community benefited from the World Bank–spon- consult with representatives of ethnic communities sored project that provided waste-management facili- that might be affected by the programs the agencies ties for every district in Lebanon. Especially where sponsor or support. Development assistance agencies projects reflect local demand rather than the prefer- can rely on government spokesmen only for the gov- ences of governments or donors, where communities ernment’s often incomplete or biased assessment of participate in project design and management and underlying realities and the effects of proposed inter- acquire a sense of ownership, the satisfactions pro- ventions. duced by such activities contribute to an atmosphere Definitions of fairness by governments, by ethnic of tolerance that bodes well for peaceful coexistence. communities, and by intraethnic factions are likely to Large projects that are perceived as damaging to an diverge. Donors must take care that local consulting ethnic community may be redesigned and divided firms or NGOs that they employ to clarify these rela- into several smaller projects that avoid the original tionships do not harbor biased viewpoints. Where damage, but achieve similar economic benefits. For possible, donors should foster dialogue that involves example, several small-scale, locally controlled water- government agencies and ethnic communities, with management systems could substitute for a large, the objective of achieving consensus on a fair appor- government-operated dam project. tionment of benefits and costs. Proposed projects The third formula produces interdependence, or policies and methods of implementation may have where a division of labor between ethnic communities to be adjusted and modified, and delays may occur rewards cooperative rather than competitive behavior. as differences are identified and debated. While An excellent example is an NGO-sponsored project in 12

Tajikistan designed to create economic interdepen- governments and with the relevant ethnic communi- dence among two hostile ethnic communities. One ties to determine which criterion is most likely to be was given wool-producing machinery, the other car- perceived as fair and workable. Some anticipated effi- pet-making equipment. Economic interdependence ciency and growth may have to be sacrificed in the in- fostered incentives for their joint economic success.27 terests of peaceful coexistence or interethnic equity. Common interests, divisibility, and interdepen- Where governments engage in practices that fla- dence are formulas that should be conducive to grantly discriminate on the basis of ethnic member- peaceful coexistence. ship and cannot be trusted to change their practices, Western donors normally favor allocation of re- development assistance agencies may (1) disqualify sources and opportunities flowing from their inter- them for further assistance, (2) channel all resources ventions—scholarships, employment, business loans, through local authorities or NGOs, or (3) impose privatization of state enterprises—according to objec- their own criteria and rigorously monitor perfor- tive criteria such as individual market competition or mance. In Kenya, donors interested in democratiza- individual merit (no preferences, no discrimination). tion and interethnic equity imposed stern measures But where societies are divided along ethnic lines, to prevent deliberate skewing of the distribution of market-merit processes can have unexpected results. the resources they provide. However, to repeat an ear- While they can overcome the gross favoritism, cor- lier caveat, because of their neocolonialist implica- ruption, and ethnic patronage practiced by some gov- tions, such interventions should be undertaken with ernments, they can also yield allocations that are circumspection and only after quiet diplomacy has skewed along ethnic lines, generating grievances proved to be futile. among those who feel disadvantaged, left out, or cheated. Members of ethnic groups that are initially THE PROMOTION OF privileged by superior education or business experi- DEMOCRATIZATION AND HUMAN ence benefit disproportionately from market-merit RIGHTS competition, widening the original gaps and yielding resentment that fuels ethnic conflict. When foreign aid promotes democratization and hu- One remedy is proportionality, where jobs, li- man rights in ethnically divided societies, the first censes, contracts, university admissions, and so forth concern of aid agencies should be to eschew majori- are allocated among ethnic communities according to tarian politics. Majoritarian or winner-take-all elec- relative numbers. Although proportionality may re- tions, while impeccably democratic in form, tend in duce economic efficiency and retard growth, it can fact to exclude minorities from positions of power ensure equitable participation. Governments may at and influence, leaving their fate entirely to the mercy times favor compensatory allocations for members of of majorities. This was the predicament of the disadvantaged communities and expect donors to Catholic minority in Northern Ireland before it comply with this policy. Like proportionality, com- launched mass protests and demonstrations in the pensatory policies are vulnerable to the previously late 1960s. If ethnic minorities are to feel secure they mentioned backlash reaction. Affirmative action for must have some control over their destiny. Where ter- Sinhalese produced the anger among Tamils that ritorial autonomy is feasible, is an precipitated the ; more skillfully option, as in Ethiopia. The national minority becomes managed, compensatory measures in Malaysia con- the majority in its region; the regional government is tributed to rectifying previous patterns of ethnically operated by fellow ethnics, who control the allocation based economic and occupational inequality at tolera- of land and natural resources, credit, and public con- ble social cost.28 tracts; the language of the regional majority becomes Individual competition, proportionality, and com- the language of government and education. At the pensatory preferences are the principal formal criteria same time, representatives of the national minority for allocation. All have implications for ethnic conflict. participate in the affairs of the central government in None is inherently superior on moral grounds. The decisionmaking roles. The complications and risks of effectiveness of each depends on the context in federalism are well known: minority communities which it is applied. Development assistance agencies within federalized regions may demand their own should evaluate the trade-offs in consultation with autonomy or independence; federal autonomy may 13

lead to demands for full independence. , An analogy can be found in the arena of environ- however, may be preferable to prolonged and violent mental affairs. There, too, most development conflict; it may forestall ethnic conflict when peaceful assistance professionals and Third World govern- coexistence within a single polity proves impossible. ments actively opposed the imposition of environ- Where territorial autonomy is infeasible, power- mental conditions on foreign aid or were reluctant to sharing arrangements can enable minorities to con- complicate their calculations, negotiations, and oper- trol some of their own institutions, including schools, ations with still another noneconomic impediment. to use their own language in transactions with gov- They were forced to do so by the requirement that en- ernment, and to hold positions in government in vironmental impact statements be prepared for pro- rough proportion to their numbers.29 Less far- jects that might have environmental implications. reaching than formal power-sharing arrangements, While the results have often been disappointing to en- minority rights provide the means for minorities to vironmentalists, these impact statements have greatly feel secure in a multiethnic polity, to maintain their increased the salience of environmental values in de- cultures and their corporate existence.30 Minority velopment assistance operations. Why not follow this rights exemplify recognition and respect by govern- precedent and introduce ethnic impact statements ment and the majority for the minority’s distinctive for policy initiatives or projects that might affect in- status, thereby diminishing the sources of conflict.31 terethnic relations or relations between ethnic com- Although they are not panaceas for managing con- munities and governments? flict, free elections do contribute to the legitimacy of What might ethnic impact statements contain? governments. Election machinery should make it pos- First, they would present a country or regional back- sible for minorities to be equitably represented. Pro- ground analysis identifying the major ethnic commu- portional representation is one such method. Another nities and describing their demography, economic is to design the electoral system to favor ethnically base, levels of mobilization and solidarity, the recent moderate candidates and make it more difficult for history of relations among these communities and be- extremists to be elected. A number of such arrange- tween them and government, and government poli- ments have been identified and analyzed.32 Even cies and practices affecting these relationships. The where ethnic solidarities are recognized as legitimate background analysis would be followed by an esti- and enduring, multimember districts in which voters mate of the effects of the proposed project or policy from all ethnic backgrounds cast ballots for all the change on the state of ethnic relations in the country. seats provide incentives for candidates to appeal for This would include the effects on relations among support across ethnic lines. This favors ethnic moder- ethnic communities and between these communities ates and penalizes extremists. Party and election rules and the government. If the proposed intervention can sanction political organizations and candidates would be likely to generate grievances, could they be whose campaign appeals are blatantly ethnic and neutralized by measures that the donors and the gov- likely to provoke interethnic hostility. ernment might implement? If not, should the pro- posal be modified or abandoned? ETHNIC CONDITIONALITY Ethnic impact statements would constitute exten- sions, in effect, of social soundness analysis. They Understanding of and concern for ethnic conflict would be not dissimilar to the economic background among the major development assistance institutions statements that donor agencies routinely require to has been hesitant, reluctant, inept, or completely ab- inform their economic policy initiatives and project sent. This despite the fact that ethnic solidarities have interventions. Armed with this information, donors become politicized in many of the countries in which would be better equipped to estimate the effects of these institutions operate and that ethnic conflict, of- proposed interventions on these relationships—on ten violent, has become a global reality likely to affect equity, coexistence, or conflict—and to design inter- the success of their interventions. The tendency to ventions accordingly. ignore this reality or to address it awkwardly or These statements would be prepared by qualified obliquely represents a culture lag that will have to social scientists with the participation of indigenous be overcome. scholars and consultants. Some development assis- tance professionals would resist these statements, just 14

as they initially resisted similar requirements for envi- be dissipated unless that society enjoys internal peace ronmental analysis and for similar reasons: that the and order, which may be jeopardized by ethnic con- statements would increase costs and delay implemen- flict. Measures should therefore be taken to reduce or tation of economically sound projects or policies, di- preempt ethnically based grievances by ensuring the vert attention from economic development, and have equitable division of benefits and costs among ethnic the mischievous effect of fomenting ethnic activism communities and by other measures intended to fos- among previously passive communities. Govern- ter peaceful coexistence. At the same time, donors ments may also regard this interest by donors in their may find it impossible to work with regimes that de- domestic politics as an illegitimate, neocolonialist en- liberately discriminate against or otherwise harm eth- croachment on their sovereignty nic communities, or that flout the rights of minorities. Yet it is no longer possible for donors to disregard Under such circumstances donors may opt for ethnic the reality that dominates public affairs in many of the conditionality. countries in which they operate, that has the potential The process would be similar to methods em- to disrupt these interventions, and that can even rein- ployed for other forms of conditionality. First, donors force or provoke conflict. Even equipped with accu- would engage in discussions with representatives of rate information, development assistance initiatives government, a form of constructive engagement will still encounter uncertainties and unanticipated intended to negotiate agreement on the terms of con- consequences. But reliable information can help re- ditionality. Such agreements may represent compro- duce risks and avoid unnecessary mistakes. mises, followed by the monitoring of compliance and Does ethnic sensitivity by donor agencies imply further dialogue. If donors can persuade themselves ethnic conditionality? What is the rationale for any that there has been progress—albeit perhaps less than form of conditionality, and would ethnic conditional- stipulated in the terms of the original agreement—they ity satisfy that rationale? Conditionality has been im- may continue support, hoping that steady pressure posed for both instrumental and intrinsic reasons: will eventually persuade the government that the con- - Donors may believe that successful implementa- ditions serve its interests as well. Donors try to avoid tion of policies, projects, or humanitarian assis- the disruptive effects of terminating support, prefer- tance requires responsive behavior by host ring to remain engaged rather than to apply sanc- governments. A project cannot be expected to tions. When all else fails, donors can suspend succeed unless, for example, certain technical assistance until the government is willing to renew and administrative procedures are followed and discussions. In that event, the process begins anew. measures are in place to ensure that the project Meanwhile donors may try to continue assistance to can be sustained after external assistance ends. the country through NGOs or local authorities. Such conditions are routinely incorporated into Ethnic conditionality may prove to be necessary to project agreements before projects are launched. ensure effective or morally acceptable uses of devel- Funds intended to achieve fiscal stabilization opment assistance resources in some ethnically may be wasted unless revenues increase and ex- divided countries. As the limitations of developmen- penditures, including subsidies, are curbed. talism become more apparent and foreign assistance These expressions of economic conditionality are increasingly emphasizes such political values as good intended to be instrumental to policy outcomes. governance, human rights, and democratization, - there will be further need to resort to this instrument Donors may believe that certain values and prac- of development assistance. tices that are important to their constituents, that Ethnic impact statements and ethnic conditional- are intrinsic to their conception of a good society, ity reflect the increasing salience of ethnic politics in must be respected. These may include free enter- the affairs of aid-receiving countries and the need for prise, basic human rights, and the rule of law. development assistance to take account of this reality. They become conditions for the donors’ partici- Failure to do so has exacerbated tensions and vio- pation in a development assistance relationship. lence. If it is skillfully administered, development as- In the absence of some progress along these lines, sistance can contribute to preventing, to mitigating, donors may halt their assistance. and to regulating ethnic conflict. Ethnic conditionality is warranted by both sets of cri- teria. Funds intended to promote economic develop- ment and improve quality of life within a society will 15

9. Heather McHugh, “USAID and Ethnic Conflict: An Epiphany?” in Esman and Herring, eds., Develop- NOTES ment Assistance and Ethnic Conflict.

10.Alison Brysk, “‘Indian Market’: The Ethnic Face of Adjustment in Ecuador,” in Esman and Herring, eds., Development Assistance and Ethnic Conflict.

11. Ibid. 1. Conrad Phillip Kottak, “When People Don’t Come First: Some Sociological Lessons from Completed The Experience of NGOs in Con- Putting People 12. Mary Anderson, Projects,” in Michael Cernea, ed., flict Intervention: Problems and Prospects First: Sociological Variables in Rural Development (Cam- bridge, Mass.: The Local Capacities for Peace (Oxford, Oxford University Press, 1991), 325—356. Project, 1995). 2. Michael Lipton, Why Poor People Stay Poor: Ur- 13. Mary Anderson, personal telephone conversation ban Bias in World Development (Cambridge, with the author. Mass.: Harvard University Press, 1977). The Politics 3. Robert Cassen and Associates, “Aid and Poverty,” in 14. See, for instance, M. Crawford Young, Does Aid Work? Report to an Intergovernmental of (Madison, Wis.: University of Task Force (Oxford: Oxford University Press, Wisconsin Press, 1976); Nathan Glazer and Daniel Ethnicity: Theory and Ex- 1986). Patrick Moynihan, eds., perience (Cambridge, Mass.: Harvard University Ethnic Groups in 4. Susan Woodward, “Redrawing Borders in a Period Press, 1975); Donald Horowitz, Conflict of Systemic Transition,” in Milton J. Esman and (Berkeley, Calif.: University of California Ethnic Groups and Shibley Telhami, eds., International Organization Press, 1985); Paul Brass, ed., the State and Ethnic Conflict (Ithaca, N.Y.: Cornell Univer- (Totowa, N.J.: Barnes and Noble, 1985); The Politics of Nationalism and Eth- sity Press, 1995), 198—234. James Kellas, nicity (London: MacMillan Education, 1991); Ted 5. John Cohen, “Foreign Aid and Ethnic Interests in Robert Gurr, Minorities at Risk: A Global View of Kenya,” in Milton J. Esman and Ronald J. Herring, Ethnopolitical Conflicts (Washington, D.C.: United eds., Development Assistance and Ethnic Conflict States Institute of Peace Press, 1993); and Milton J. (forthcoming). Esman, Ethnic Politics (Ithaca, N.Y.: Cornell Univer- sity Press, 1994). 6. Daniel R. Gibson, “The World Bank and Displace- ment: The Challenge of Heterogeneity,” in Esman 15. Dennis Rondinelli, Development Projects as Pol- and Herring, eds., Development Assistance and icy Experiments (New York: Methuen, 1983). Ethnic Conflict. 16.W. W. Rostow, Economic Growth from David 7. David Korten and Rudy Klauss, People-Centered Hume to the Present (Oxford: Oxford University Development: Contributions toward Theory and Press, 1990). Planning Frameworks (West Hartford, Conn.: Ku- marian Press, 1984); and T. J. Carroll and John D. 17. See, for instance, Mancur Olson, “Rapid Growth Montgomery, eds., Supporting Grass Roots Orga- as a Destabilizing Force,” Journal of Economic His- nizations (Cambridge, Mass.: Lincoln Institute for tory 23 (December 1963); and H. W. Arndt, The Land Management, 1987). Rise and Fall of Economic Growth (Melbourne, Australia: Longman Cheshire Press, 1978). 8. Organization for Economic Cooperation and De- velopment (OECD), Development Cooperation: 18. Milton J. Esman, “Economic Performance and Eth- Efforts and Policies of the Members of the Devel- nic Conflict,” in Joseph V. Montville, ed., Conflict opment Assistance Committee, 1994 Report and Peacemaking in Multiethnic Societies (Lex- (Paris: OECD, 1995). ington, Mass.: D. C. Heath, 1990); Steven Majs- torovic, “Politicized Ethnicity and Economic Inequality,” Nationalism and Ethnic Politics 1, no. 1 (spring 1995): 33—53. 16

19. Robert Miller, ed., Aid as Peacemaker: Canada’s 31. Gurr, Minorities at Risk, chap. 10. Development Assistance and Third World Con- flict (Ottawa, Carleton University Press, 1992). 32. Horowitz, “Making Moderation Pay.”

20. Neil Nevette and Charles H. Kennedy, eds., Ethnic Preference and Public Policy in Developing States (Boulder, Colo.: Lynne Rienner, 1986).

21.Brysk, “‘Indian Market.’”

22. Esman, “Economic Performance and Ethnic Con- flict.”

23. Donald Horowitz, “Making Moderation Pay: The of Ethnic Conflict Manage- ment,” in Montville, ed., Conflict and Peacemaking in Multiethnic Societies, 451—475; Thomas Sowell, Preferential Policies in International Perspective (New York: W. Morrow, 1990).

24. See Arend Lijphart, Democracy in Plural Societies: A Comparative Exploration (New Haven, Conn.: Yale University Press, 1977); and Kenneth McRea, ed., Consociational Democracy (Toronto: McClel- land and Stewart, 1974). For a critical assessment, see Alvin Rabushka and Kenneth A. Shepsle, Poli- tics in Plural Societies: A Theory of Democratic Instability (Columbus, Ohio: Charles E. Merrill, 1974).

25. Vernon Van Dyke, “Justice as Fairness: For Groups,” American Political Science Review 69, no. 2 (June 1975): 607—614.

26. Norman Uphoff, “Ethnic Cooperation in Sri Lanka: Through the Keyhole of a USAID Project,” in Es- man and Herring, eds., Development Assistance and Ethnic Conflict.

27. As reported in Anderson, The Experience of NGOs in Conflict Intervention.

28. Esman, Ethnic Politics, 72—74.

29. Arend Lijphart, “The Power-Sharing Approach,” in Montville, ed., Conflict and Peacemaking in Multi- ethnic Societies, 491—509.

30. Timothy D. Sisk, Power Sharing and International Mediation in Ethnic Conflicts (Washington, D.C.: United States Institute of Peace Press, 1996). 17 ABOUT THE AUTHOR

Milton J. Esman, a leading au- thority on ethnic politics and de- velopment assistance, is emeritus John S. Knight Professor of Inter- national Studies at Cornell University. After receiving his doc- torate in politics from Princeton University, he served with the U.S. Army and the U.S. Department of State and taught at George Wash- ington University and the University of Pittsburgh be- fore becoming director of Cornell University’s Center for International Studies, a post he held from 1969 to 1983. Esman has been a consultant for a number of organizations, including the World Bank, USAID, FAO, ILO, UNDR, the Ford Foundation, and the Con- gressional Office of Technology Assessment. He has been a visiting professor at Leiden University in The Netherlands, the Hebrew University of Jerusalem, and the Osmania University in India, and was Senior Adviser to the Prime Minister’s Department, Govern- ment of Malaysia, from 1966 to 1968. His list of publi- cations on topics ranging from ethnic politics to rural development to institution building is long and dis- tinguished; indeed, he was named as one of the most frequently cited political scientists in the June 1989 is- sue of PS. Recent books include Ethnic Politics; Inter- national Organizations and Ethnic Politics (edited with Shibley Telhami); Ethnicity, Pluralism, and the State in the Middle East (edited with Itamar Rabi- novich); and Management Dimensions of Develop- ment. 18 ABOUT THE INSTITUTE

The United States Institute of Peace is an independent, nonpartisan federal institution created by Congress to promote research, education, and training on the peaceful resolution of international conflicts. Established in 1984, the Institute meets its congressional mandate through an array of programs, including research grants, fellowships, professional training programs, conferences and workshops, library services, publications, and other educational activities. The Institute’s board of directors is appointed by the Presi- dent of the United States and confirmed by the Senate.

Chairman of the Board: Chester A. Crocker Vice Chairman: Max M. Kampelman President: Richard H. Solomon Executive Vice President: Harriet Hentges

Board of Directors

Chester A. Crocker (Chairman), Research Professor of Diplomacy, School of Foreign Service, Georgetown University Max M. Kampelman, Esq. (Vice Chairman), Fried, Frank, Harris, Shriver and Jacobson, Washington, D.C. Dennis L. Bark, Senior Fellow, Hoover Institution on War, Revolution and Peace, Stanford University Theodore M. Hesburgh, President Emeritus, University of Notre Dame Seymour Martin Lipset, Hazel Professor of Public Policy, George Mason University Christopher H. Phillips, former U.S. ambassador to Brunei Mary Louise Smith, civic activist; former chairman, Republican National Committee W. Scott Thompson, Professor of International Politics, Fletcher School of Law and Diplomacy, Tufts University Allen Weinstein, President, Center for Democracy, Washington, D.C. Harriet Zimmerman, Vice President, American Israel Public Affairs Committee, Washington, D.C.

Members ex officio Ralph Earle II, Deputy Director, U.S. Arms Control and Disarmament Agency Toby Trister Gati, Assistant Secretary of State for Intelligence and Research Ervin J. Rokke, Lieutenant General, U.S. Air Force; President, National Defense University Walter B. Slocombe, Under Secretary of Defense for Policy Richard H. Solomon, President, United States Institute of Peace (nonvoting) Other titles in the Peaceworks series:

Turkey’s Role in the Middle East: A Conference Report by Patricia Carley (Peaceworks No. 1) Central Asians Take Stock: Reform, Corruption, and Identity by Nancy Lubin (Peaceworks No. 2) From the Managing Chaos conference: Keynote Addresses by Les Aspin and Ted Koppel (Peaceworks No. 3) Sources of Conflict: G.M. Tamás and Samuel Huntington on “Identity and Conflict,” and Robert Kaplan and Jessica Tuchman Mathews on “‘The Coming Anarchy’ and the -State under Siege” (Peaceworks No. 4) NGOs and Conflict Management by Pamela R. Aall (Peaceworks No. 5) Humanitarian Assistance and Conflict in Africa by David R. Smock (Peaceworks No. 6) Self-Determination: Sovereignty, Territorial Integrity, and the Right to Secession by Patricia Carley (Peaceworks No. 7) and the Origins of the Yugoslav Crisis by Vesna Pesic (Peaceworks No. 8) Peace Operations and Common Sense: Replacing Rhetoric with Realism by Denis McLean (Peaceworks No. 9) State and Soldier in Latin America: Redefining the Military’s Role in Argentina, Brazil, and Chile by Wendy Hunter (Peaceworks No. 10)

Peaceworks No. 13. First published March 1997.

The views expressed in this report are those of the author alone. They do not necessarily reflect views of the United States Institute of Peace.

UNITED STATES INSTITUTE OF PEACE 1550 M Street NW, Suite 700 Washington, DC 20005-1708

Phone: 202-457-1700 Fax: 202-429-6063 E-mail: [email protected] Internet: www.usip.org Many other publications from the United States Institute of Peace address issues of direct relevance to development assistance and ethnicity and ethnic conflict. Recent Institute books and reports include:

Autonomy: Flexible Solutions to Intrastate Conflicts, by Ruth Lapidoth (1997) Minorities at Risk: A Global View of Ethnopolitical Conflicts, by Ted Robert Gurr (1993) Power Sharing and International Mediation in Ethnic Conflicts, by Timothy D. Sisk (1996) Zaire: Predicament and Prospects, by J. C. Willame et al. (1997) Humanitarian Assistance and Conflict in Africa, by David R. Smock (1996) NGOs and Conflict Management, by Pamela R. Aall (1996) Sri Lanka: The Invention of Enmity, by David Little (1994) Transitional Justice: How Emerging Democracies Reckon with Former Regimes (3 vols.), edited by Neil J. Kritz (1996) Preventing Violent Conflicts: A Strategy for Preventive Diplomacy, by Michael S. Lund (1996)

UNITED STATES INSTITUTE OF PEACE 1550 M Street NW Washington, DC 20005 Peaceworks No. 13