Lecture 6.1

Ethnicity and Public Goods

Emmanuel Teitelbaum

Diversity and Collective Action

Collective Action and Public Goods

  • Because they are non-excludable, public goods and common pool resources threatened by collective action problems (CAPs)
    • Public goods = socially beneficial
    • Individuals have incentives to free-ride
  • Tragedy of the commons
    • Large-scale illustration of collective action problem
  • Prisoner’s dilemma
    • Small-scale illustration of collective action problem

Diversity and Public Provisioning

  • Macro-level studies find a robust negative correlation between ethnic heterogeneity and public goods provisioning
    • Seminal study: Alesina, Baqir, and Easterly 1999
  • Micro-level studies find that shared identity promotes collective action among “coethnics,” and prevents cooperation among “non-coethnics”
  • What explains this relationship? What links ethnic diversity and public goods provisioning?

The Central Issue

  • Ethnicity turns public goods into club goods
  • Collective action occurring at the level of the ethnic group, not at the level of society as a whole

Economic Theories

1. Selective Incentives

  • Collective action is more likely with the right set of incentives
  • “Selective incentives” are incentives targeted at individuals to make participation advantageous and non-participation disadvantageous (Olson)
    • Inducements–positive incentives
    • Constraints–negative incentives
  • Ethnic groups may be better at devising and applying selective incentives than other groups or institutions
    • Expectations of reciprocity
    • Ability to sanction through informal institutions

2. Size of Groups

  • Coordination problems are even bigger among large groups (Olson)
    • “Action taking groups” tend to be much smaller than “non-action taking groups”
    • Examples: committees in Congress, partners in a small company, executives, lobbyists, unions
    • Versus broad classes of people (working-class; citizens; shareholders in a large company)
  • Why? Distribution of gains versus costs of collective action
  • Ethnic groups are a smaller subset of society
    • Thus a more viable “action-taking” group
    • Think here about public provisioning at the local (e.g. village) level

3. Contract Theory

  • Basic problem: when ‘a’ contracts with ‘b’, how does ‘a’ know that ‘b’ will follow through?
  • In developing countries, governments often lack enforcement capacity
  • Ethnic groups help overcome deficits of government institutions
    • Informational shortcuts (e.g. “I trust her because she is from my community.”)
    • By compensating for reputation and trust deficits through dense social networks (e.g. “We are from the same village and people say that she is trustworthy.”)

The Role of Politics

Explaining Variation

  • Correlation between ethnic diversity and service provisioning is not absolute
  • We see substantial cross-national variation in how much ethnicity matters
  • We also see lots of sub-national variation
  • How do we explain this variation?


Singh, 2015

  • Subnationalism: a form of group identification based on language or regional culture
  • Subnational identity increases support for collective welfare
  • State enacts social policy, promotes human development

Encompassing Ethnic Parties

Source: Thachil and Teitelbaum, 2015

  • Narrow Ethnic Parties
    • Small winning coalitions
    • Politics remains clientelistic
  • Encompassing Ethnic Parties
    • Larger winning coalitions
    • More social spending

Sectarianism (Cammett)

  • Future movement vs. Hezbollah
    • Future movement provides benefits across sectarian lines

    • Whereas Hezbollah favors coethnics

  • Key reason is electoral mobilization
    • Future Movement prioritizes winning elections
    • Hezbollah prioritizes contentious politics (protests, demonstrations, or warfare)

Source: Wikimedia commons


Miguel Study

  • So far we have talked about how structural factors condition effects of ethnicity on service provisioning.
  • Can governments choose to alter incentives if they want to?
  • Tanzania vs. Kenya
    • Which country does a better job of managing ethnic diversity?
    • What is the evidence of this?
  • How do Tanzania and Kenya approach diversity differently?
    • Why is one more successful than the other?
    • Are there lessons for other countries?

Micro-Level Explanations

Which Mechanisms Are Relevant?

Source: Habyarimana et. al., 2007

Four Types of Games

  • Dictator game
    • 500 shilling (2 coins)
      • Determines who is “egoist”
    • 100 shilling (10 coins)
    • Distribute to coethnics and non-coethnics
    • Anonymous and non-anonymous
  • Lockbox game
    • Work together to open lock
  • Puzzle game
    • Work together to solve puzzle
  • Network game
    • Find and receive message from person

Dictator Game (non-anonymous)

Source: Habyarimana et. al., 2007

Dictator Game (anonymous)

Source: Habyarimana et. al., 2007

Basic Findings

  • No evidence of preference mechanisms

  • Weak evidence of technology mechanisms

  • Stronger evidence of strategy selection mechanisms

  • “Egoists” give to coethnics in non-anonymous game