Playing the Communal Card: Elites, Parties, and Inter-Ethnic Electoral Violence in Kenya and India

This dissertation proposes a theory on the determinants of inter-ethnic electoral violence. Based on the construction of an original dataset on communal clashes that occurred around multiparty elections in Kenya between 1991 and 2013 as well as extensive field research including a total of 175 elite interviews in Kenya and India, I develop an explanation to account for why such violence has accompanied some contests but not others. The argument advanced by the dissertation holds that in ethnically divided societies such as Kenya and India, the instrumentalization of election-related communal conflict affords vital opportunities to power-seeking elites to skew the electoral odds in their favor and ultimately, to capture political office.

Based on this assumption, my work proffers a three-stage model to account for when politicians are likely to organize such violence. First, I argue that inter-ethnic electoral violence stems from elites’ politicization of long-standing grievances held by rival ascriptive groups—such as Kikuyus and Kalenjins in Kenya and Hindus and Muslims in India. Second, I show that whereas driving ethnic wedges is a necessary and sufficient condition for the outbreak of election violence, when antagonistic communities are also voting on opposite sides, then the associated electoral clashes are likely to be severe in scale. Third and finally, in order to explain when and with what frequency elites can be expected to politicize such divides, my work finds that contrary to existing explanations about the level of inter-party competition or the ethnic composition of civic associations, this variance is more accurately explained by the degree to which a country’s party system is institutionalized. More concretely, I show that depending on whether political contestation takes place between purely ethnic parties versus relatively programmatic parties, politicians will come to face very different incentive structures, which will, in turn, affect the likelihood with which they will deploy divisive versus accommodationist electoral strategies. Whereas in the former circumstances—which typify electoral competition in Kenya—I find that tactics of ethnic accommodation have merely been responses to revised institutional arrangements, in the latter situation—which is characteristic of Indian politics—my work uncovers that the outbreak of elite-driven communal violence actually reflected a temporary rupture in broader governing strategies of ethnic accommodation. As such, this dissertation demonstrates that it is only when prevailing modes of political contestation are interrupted that we can expect a) peace to come about in previously conflict-ridden locations or b) violence to break out in hitherto calm sites.