Weak States and the Roots of Ethnic Identification in Africa
Committee: David Waldner (chair), Daniel W. Gingerich, Lynn M. Sanders, and Sophie Trawalter
Why do some citizens place so much importance on their ethnicity in political, social and economic interactions, while others do not? Why do some Kenyans think of themselves as “Kikuyu,” while others think of themselves as “Kenyan,” when they are objectively members of both groups? In much of sub-Saharan Africa, the absence of shared culture and language mean that national governments are the clearest manifestation of a shared national identity. State service provision, in particular, is a potent reminder of one’s membership in a national group and the relevance of that membership in everyday life.
In my dissertation, I formalize and apply a social-psychological theory of identity formation to argue that state service provision can induce stronger national identities, displacing and weakening ties to subnational ethnic groups. Individuals who receive services from their governments will express lower levels of ethnic identification. Where the state is absent, however, membership in parochial subnational groups remains important—the weakness of many African states, then, can help us understand the strength of ethnic identification on the continent.
To assess this argument, I develop a new tool to overcome social desirability bias and measure the strength of ethnic identification. I use this new measure in a series of survey experiments implemented around central and western Kenya, showing that individuals primed to recall government services express less coethnic favoritism. I also assess the effect of a large-scale government cash transfer program in Kenya’s arid, impoverished north: using a regression discontinuity design, I find that recipients of a small, bi-monthly cash payment from the government exhibit lower levels of ethnic identification than comparable members of their community. Finally, I demonstrate the importance of social identity in shaping political outcomes—exploiting naturally-occurring variation in the salience of national identity, I find that increases in the strength of national identification reduce distributional conflict over public goods provision between rival ethnic groups. These findings suggest that ethnic diversity and ethnicity’s salience are endogenous to characteristics of the state and public policies. They also have implications for policymakers seeking practical solutions to overcome ethnic tensions.
Committee: David Waldner (chair), Daniel W. Gingerich, Lynn M. Sanders, and Sophie Trawalter
Why do some citizens place so much importance on their ethnicity in political, social and economic interactions, while others do not? Why do some Kenyans think of themselves as “Kikuyu,” while others think of themselves as “Kenyan,” when they are objectively members of both groups? In much of sub-Saharan Africa, the absence of shared culture and language mean that national governments are the clearest manifestation of a shared national identity. State service provision, in particular, is a potent reminder of one’s membership in a national group and the relevance of that membership in everyday life.
In my dissertation, I formalize and apply a social-psychological theory of identity formation to argue that state service provision can induce stronger national identities, displacing and weakening ties to subnational ethnic groups. Individuals who receive services from their governments will express lower levels of ethnic identification. Where the state is absent, however, membership in parochial subnational groups remains important—the weakness of many African states, then, can help us understand the strength of ethnic identification on the continent.
To assess this argument, I develop a new tool to overcome social desirability bias and measure the strength of ethnic identification. I use this new measure in a series of survey experiments implemented around central and western Kenya, showing that individuals primed to recall government services express less coethnic favoritism. I also assess the effect of a large-scale government cash transfer program in Kenya’s arid, impoverished north: using a regression discontinuity design, I find that recipients of a small, bi-monthly cash payment from the government exhibit lower levels of ethnic identification than comparable members of their community. Finally, I demonstrate the importance of social identity in shaping political outcomes—exploiting naturally-occurring variation in the salience of national identity, I find that increases in the strength of national identification reduce distributional conflict over public goods provision between rival ethnic groups. These findings suggest that ethnic diversity and ethnicity’s salience are endogenous to characteristics of the state and public policies. They also have implications for policymakers seeking practical solutions to overcome ethnic tensions.