October 17, 2022
instrumental:
psychological:
Key implication of these arguments is that states with greater ethnic diversity\(^*\) are more prone to conflict, which could then turn into civil war.
We have lots of theoretical explanations for ethnic diversity producing conflict…
… but is there evidence that there is a link?
First major attempt to answer:
Collect data on:
Then compare the incidence of civil war in places with greater/lesser ethnic diversity.
Fearon and Laitin count as civil wars conflicts in which …
In total, they find \(127\) civil wars.
We want to compare ethnic diversity of countries around the world.
Discuss with your neighbors:
what are ways we can or should measure ethnic diversity?
what are challenges or difficulties that would make measuring ethnic diversity difficult?
Think about:
They compare country-years with differing levels of ethnic diversity, “holding constant” economic growth, population size, democracy, terrain features, and other sources of conflict…
they find that…
\(1.\) Ethnic diversity: no relationship - measured as ELF, as size of largest group, presence of majority with large minority
\(2.\) Linguistic/religious discrimination: no relationship
\(3.\) GDP per capita: negative relationship
\(4.\) Population size: positive relationship
\(5.\) Newly Independent: positive relationship
\(6.\) Mountainous terrain: positive relationship
Two ways of responding to Fearon and Laitin’s findings:
Measurement: Does the measurement of ethnic diversity make sense?
Theory: Ethnic diversity as such does not lead to civil war. Particular types of ethnic configurations produce conflict
Where \(s_i\) is fraction of population for group \(i \in \lbrace 1 \ldots n \rbrace\)
Interpret as: probability that any two random individuals belong to different ethnic groups
Measure size \(s\) of ethnic groups \(1 \ldots n\) using
Can you think of any potential problems with this approach?
ELF treats these two situations as identical
ELF measure
No clear relationship between ethnic diversity and civil war
BUT
Overall, well done! Mean of 77.1% (B+), including late penalties.