• Does ethnic diversity lead to conflict?
• If so, why?
• Instrumental Explanations
  • Bates
  • Habyarimana et al

Many conflicts appear ethnic

Is ethnicity inherently linked to conflict?

Is it ethnic diversity more prone to generate conflict?

we rejected essentialist views of ethnicity:

• “ancient hatreds”, intrinsic differences can’t explain conflict
• ethnicity is one among many forms of social identity

we rejected primordialist views of ethnicity:

• ethnic conflict not likely to be result of inherent psychological weight put on descent-based ties
• salient ethnic identities might be result of conflict

Might there be something about ethnicity (as opposed to other forms of identification) that makes conflict more likely…

• for instrumental/strategic reasons?
• for psychological reasons?
• for structural reasons?

modernization of economy and government creates goodies:

• education, jobs, goods and services
• goodies are limited (finite) and exclusive
• people mobilize politically to gain access to these goodies:

Competition between groups over resources \(\to\) conflict (not necessarily violent)

Individuals seek to maximize access to goodies:

• in democratizing states, access requires winning elections (need to win \(\to\) larger coalitions)
• goodies are finite/exclusive (prefer to have minimum winning coalition )
  • leaders w/in groups need many supporters to keep access
  • supporters need leaders in power to gain access

But why would this lead to mobilization around ethnicity (rather than, e.g. class)?

• Ethnic groups are well suited to mobilize over goodies, because they have organizational advantages (Bates 1983; Horowitz 1985, pp 99-105)

• Because ethnic groups more likely to succeed in capturing resources, mobilize around ethnicity for instrumental reasons

Bates: ethnic groups well suited for political competition, because…

• Ethnic groups share preferences over resource allocation:
  • ethnic groups are spatially distributed, as are goodies like development projects
  • ethnic groups more likely to agree on policies around language/religion/culture
• Ethnic groups are better at solving collective action problems:
  • ethnic institutions/networks facilitate coordination
  • shared language reduces organization/transaction costs

On the board

• Why shared preferences help coordination
• Why coordination problems persist, even with shared preferences
• How do we solve collective action problems?
  • selective incentives
  • lower transaction costs
  • norms and repeated play

In sum, Bates argues that mobilizing ethnic groups is strategically more effective, because they can better overcome collective action problems.

Robust finding that places with greater ethnic diversity provide fewer public goods (caveats to come)

• public goods are non-excludable, possible free-riders
• this is the “flip-side” of argument that ethnic groups are good at solving collective action problems

Is poor public goods provision in diverse areas the result strategic advantages of shared ethnicity?

• Habyarimana try to answer this question.

Several explanations for diversity-public goods connection:

\(1.\) Preferences

• commonality of tastes (Bates’s argument)
• other-regarding preferences (taste-based discrimination)

\(2.\) Technology

• efficacy: language, common experience \(\to\) collaboration
• findability: easier to locate and identify in-group members (reward/punish)

\(3.\) Strategy

• social sanctioning: norm of tit-for-tat w/ co-ethnics

Test Design:

• random sample of people residing in high diversity/low public goods areas of Kampala, Uganda
• people played cooperative games with co-ethnics, out-groups
• random assignment to play with co-ethnics/out-groups, knowledge of ethnicity
• different games get at different mechanisms.

Preference: commonality of taste

Surveyed people over most important public goods (security, garbage, drainage) and method of delivery (public vs private contractor)

No evidence that preferences for type or delivery of public good varies across ethnic group.

commonality of taste not relevant in this case.

Preference: taste-based discrimination

dictator game: player is anonymous and must allocate money (ten coins, two coins) between themselves and two other players.

• because player is anonymous, preference for co-ethnic reflects taste-based discrimination
• No evidence of discrimination by ethnicity

Technology: Efficacy

puzzle game: two people play face to face; rewarded for completing a small jig-saw puzzle

• each person has two pieces, cannot see the others’ pieces.
• modest, not significant advantages to co-ethnic teams (48% success vs 38%)

Co-ethnics not particularly better at working together

Technology: Findability/Sanctioning

network game: randomly selected 148 people in neighborhood. Subjects randomly assigned to find a person, report their birthday.

• Rewarded for finding the person quickly.
• Co-ethnics found their person 43% (vs 28%) of the time

Co-ethnics can more easily locate each other \(\to\) easier to punish defectors/reward cooperators

Strategy: Social Sanctioning

dictator game: player is NOT anonymous and must allocate money (ten coins, two coins) between themselves and two other players.

• If players show a greater bias towards members of their own ethnic group when others can recognize their ethnicity, reflects norm that co-ethnics cooperate and sanction defection.
• “Egoists” (selfish players) show strong discrimination toward co-ethnics only when ethnicity is visible.

Habyarimana et al:

• limited support for ethnic groups better at solving collective action problems
• one place, one time; compare this to class?

Caveats about ethnic diversity and public goods.