• Linking strategic, psychological, and structural explanations
• British Census in India
• Census enumeration and conflict: evidence

Instrumental theories of conflict:

• which ethnic groups become operative to mobilize around?

  • to create ethnic parties
  • to appeal to through ethnic out-bidding?

Psychological theories of conflict:

• which ethnic categories become operative (how salient is ethnicity)?
• what is the basis of status comparisons?
• what determines the relative status of groups?

Instrumental theories emphasize rational calculations of ethnic entrepreneurs…

Psychological theories emphasize individual psychological responses…

But these calculations, cognitive processes take place against a structural background.

What features of ethnic boundary (structures) do you think might …

• amplify/intensify instrumental logic of ethnic conflict?
• amplify/intensify psychological logic of ethnic conflict?

Hindu-Muslim divide a continuing source of conflict, violence

• Major political parties divided on the issue
• Major riots in Delhi, 2020.
• Conflict over Citizenship Amendment Act, National Register of Citizens

Hindus

• no unifying label was in widespread use
• texts, rituals, traditions highly varied
• disputes within “Hindus”

Muslims

• often labeled by their geographic origin, not by religion

Syncretism

• Hindus and Muslims shared saints, holy men, shrines, gods, festivals
• religious practices/taboos spread across the divide
• cross-cultural hybrid identities

In this period…

• Muslim rulers broadly tolerant of different religious practices
• Hindu-Muslim violence rare; often not understood as about religion at the time

How did the Hindu-Muslim cleavage emerge?

One part of the story is the British Colonial Census of India

• First Census in 1871
• Recognized only the religious categories of: “Hindu”, “Muslim”, “Sikh”, and “Christian” (later more added).
• Defined rules for applying these labels to colonial subjects

• What kind of boundary change is use of census categories?
• Why might this lead to conflict?

Census and Structure:

British colonial imagination imposed on diverse population:

“‘Hindu’ means a non-Musalman native of India.”

“a quarter of the persons classed as Hindus denied the supremacy of the brahmans and the authority of the Vedas; more than half did not receive the mantras from a recognised Hindu guru, a quarter did not worship the great Hindu gods, and were not served by good Brahmans priests; a third were denied access to temple; a quarter caused pollution by touch, a seventh always buried their dead, while a half did not regard cremation as obligatory and two-fifths ate beef.”

Lieberman and Singh (2017), p. 41

Census and Structure:

Census categories…

• are institutional use
• basis of institutional separation
• produced reports on size, geography, economic/political status of “groups” (groupness)
• created new knowledge about power disparity

Census and Psychology:

• Social Identity Theory: people seek to maintain or promote their group status
• fear of reduced majority/permanent minority status \(\to\) ethnic mobilization

When census showed smaller Hindu majority\(^*\)

• Hindu elites used census data to argue Hindus “a dying race” in their homeland

• Hindu political organizations: reassert Hindu “traditions”, purge hybrid religion, enforce “purity”

• Muslim elites feared status compared to new “Hindu” majority: lobbied to split up Hindu census category

Census and Strategy

Ethnic “entrepreneurs” exploit the census to make political gains

• Muslim elites used Census data to lobby for special protections, institutions (separate electorates)
• BJP (Hindu political party) used 2011 census data to attack rival secular parties for ignoring growth rate of Muslims vs. Hindus

Ethnic parties use census figures/labels to make out-bidding/more extreme ethnic appeals to voters.

Building on these insights:

Does including more ethnic dimensions on the census induce greater ethnic conflict?

Recognition and definition of identity categories by the state…

• structurally: possible to imagine groups, see economic/political disparities, be recognized by the state
• psychologically: facilitates inter-group comparison, efforts to maintain/change relative group status
• strategically: creates opportunities for ethnic entrepreneuers

Data

Census

• tracked down 1333 census questionnaires for 156 countries between 1800 and 2005, nearly all of modern censuses

Data

• “Politically Relevant Ethnic Groups”: expert coding of ethnic groups that are politically “activated” (from Ethnic Power Relations database)
• Ethnic Violence: ranging from riots to armed conflict
• Ethnic Civil War: armed conflicts coded as “Ethnic” (from Cederman et al)

Research Design

Conditioning:

Comparing countries that differ in ethnic census but similar\(^*\) in:

• ethnic fractionalization, GDP per capita, population, colonial ruler, regime type, political exclusion of ethnic groups, oil production, terrain, year, time since independence

\(^*\) using mathematical approximation

Countries with more ethnic cleavages enumerated in the census, compared to (otherwise similar) countries with fewer, have…

• more politically relevant ethnic groups

Countries with more ethnic cleavages enumerated in the census have…

• more ethnic armed conflicts

• Are you convinced that census enumeration of ethnicity causes ethnic conflict? Why or why not?

Placebo:

If prior ethnic conflict \(\to\) enumerating more ethnic groups, then cannot conclude census \(\xrightarrow{causes}\) conflict.

Does census enumeration of ethnic groups predict earlier ethnic conflict?

Takeaways:

• changes in ethnic boundary (structures) can intensify/inhibit instrumental and psychological pathways to ethnic conflict
• Institutional recognition of identities in the census is one example of this; decent empirical evidence

Three perspectives on ethnicity are complementary, not either/or