October 22, 2021

Ethnicity and Conflict


  • How do institutions shape which groups are in conflict?
  • Locus of electoral competition
  • Parties

Institutions and Instrumentalism


  • sets of rules that dictate how people interact
  • can be laws/constitutions, rules within organizations
  • can be informal rules that people just know

They are the “rules of the game”:

  • just like in boardgames, card games, changing rules alter the rational strategies of self-interested players.

Institutions and Instrumentalism

Horowitz’s account of ethnic parties and ethnic outbidding:

  • rational voters, rational parties respond to rule of electoral game, leading to extremism.
  • changing electoral rules can make it rational for parties to pursue alternative

Other institutions?

But do other kinds of institutions other than voting rules matter?

A puzzle:

Chewas and Tumbukas

Two ethnic groups divided by Zambia-Malawi border.

Chewas and Tumbukas

In Malawi:

  • Extensive hostility between the groups
  • Express negative stereotypes of each other

In Zambia

  • These groups get along amicably
  • Often express that they are actually united, one people.

Even when comparing villages just across the border from each other

Chewas and Tumbukas

Despite recognizing same cultural differences in both countries…

Chewas and Tumbukas

Why do the same ethnic groups behave differently just across the border?

Political Institutions

Administrative Boundaries

  • If elites (and possibly voters) are rational, they seek to mobilize ethnic categories that generate minimum winning coalitions(Posner 2005)
  • Jurisdictional boundaries (within which laws made, policies set, patronage given) determine ethnic demography (relative size of membership within different ethnic categories)

Administrative Boundaries

Administrative Boundaries

Posner (2004):

Malawi: Much smaller country…

  • Chewas and Tumbukas make up a large share of the population. Large enough to win elections.
  • President mobilized Chewas as Chewas during elections, expanded membership rules to include more people as “Chewa”.

Zambia: Much larger country…

  • Chewas and Tumbukas are small groups, not large enough to matter for national elections.
  • Parties mobilize regional language groups: Chewas and Tumbukas mobilized together as “Easterners”

Another Puzzle:


In 1960s, Language/Region was politically salient ethnic cleavage

Between the 1970s and 1980s, tribal divisions were more important.

In the 1990s, language/region became the more salient divide.

  • Always used FPTP, SMD… can’t be electoral rules
  • Zambia hasn’t changed in size… can’t be administrative boundaries

Party System

Between 1972 and 1990, Zambia was a one-party state. Elections were held within the ruling party.

  • National level office (President, party control of parliament) not decided in competitive elections
  • Only choice of who local MP would be was competitive \(\to\) competition at the local level

Before 1972 and after 1990, Zambia had multi-party competition.

  • Elections do choose national leadership (President and ruling party)
  • Competition takes place at national level

Party System

Changing the focus of competition from national to local changes the relevant ethnic demography (Posner 2005)

With one-party rule, only demography of constituency matters:

  • language does not form minimum winning coalition \(\to\) tribe

With multi-party competition, national demography matters:

  • tribal groups are too small to win, rational to mobilize by language/region

Party System