Institutions and Ethnic Politics in Lebanon and Yemen.

Abstract

This dissertation presents three essays with the theme of ethnicity and institutions, utilizing insights and data from Lebanon and Yemen, two Arab societies within which ethnicity (sect, tribe, region) is salient politically, but which use different institutions to channel these cleavages through the political system. The first essay uses a methodological innovation to study illiterate voting rights in Lebanon, which has normative, sectarian, and distributional consequences. It first addresses the difficulties of studying sensitive topics with surveys, in which systematic response bias limits the reliability of self-reported data. I present an augmented version of the list experiment and a new statistical estimator called listit to mitigate incentives for respondents to misrepresent themselves. I show that responses to a direct question on illiterate voting rights produce sectarian answers: community membership drives attitudes, whereas material conditions do not. The opposite obtains when the question is asked indirectly via the list experiment: community membership has no influence on attitudes, which instead are driven strongly by material conditions. The second essay studies institutional preferences in Lebanon. Given the salience of sectarianism in Lebanon, it argues that preferences should vary by community membership. Although religion provides the nominal boundaries between the sectarian communities, the Lebanese are also able to invoke shared religious ideals to imagine a larger community beyond the sect: religion unites as well as divides. I show that religiosity reduces favorable assessments of autocratic institutions in all sects, suggesting that religious individuals conceive of the polity in more inclusive terms than do sectarian individuals. The third essay compares Lebanon and Yemen, arguing that the descent principle makes ethnic constituencies captive audiences to their own elites, reducing the cost of political support. The price of votes depends on the institutionally-influenced intraethnic competitive environment: oligopsony, in which elites compete for their coethnics' votes, or monopsony, in which a single vote-buyer dominates and constituents compete for patronage. I provide evidence that constituents in monopsonized communities (Lebanese Sunnis and Yemeni Shiites) make overt displays of political support for leaders with patronage considerations in mind, a dynamic unseen in the more internally competitive communities in either country.