Why do some political leaders in some Indian states impress upon their local officials that communal riots and anti-Muslim pogroms must be prevented at all costs? Why do governments in other states fail to protect their minorities or even incite violence against them? In previous chapters we showed that factors such as declining state capacity or India's changing level of consociational power sharing cannot explain the geographical or historical patterns in the effectiveness of states' response to the threat of communal violence. In this chapter, I argue that we can best explain state-level variation in levels of Hindu-Muslim violence if we understand the electoral incentives facing each state's government. I show that states with higher degrees of party fractionalization, in which minorities are therefore pivotal swing voters, have lower levels of violence than states with lower levels of party competition. This is because minorities in highly competitive party systems can extract promises of greater security from politicians in return for their votes.

The chapter is organized into three parts. First, I develop the theoretical argument about the importance of state-level electoral incentives and outline the conditions under which high levels of multiparty electoral competition will lead to higher levels of state protection for minorities. Second, using data from 1961 to 1995 for 14 major Indian states, I show that greater party fractionalization leads to a statistically significant reduction in states' levels of Hindu-Muslim riots.