Latin America has been plagued by electoral volatility in the last few decades. A recent study found that "more than two-thirds of the Latin American countries expe rienced more electoral volatility than the most volatile European democracy."1 Electoral volatility, which is measured as the net shift in votes (or seats) between parties from one election to another, may undermine democratic governance and economic growth by creating political and economic uncertainty.2 Thus, it causes considerable concern to both democratic theorists and economic planners.3 What is the source of electoral volatility within Latin America? What contribu tion, if any, do ethnic voting patterns in the region make to this volatility? A variety of studies on electoral volatility in other regions have argued that ethnic diversity tends to reduce electoral volatility because in ethnically diverse polities voters will consistently vote for the party that represents their ethnic group. Bartolini and Mair, for example, find that ethnolinguistic diversity has reduced electoral volatility in western Europe, and Birnir shows that ethnic diversity has similarly stabilized elec toral preferences in new democracies around the world.4 Surprisingly, however, stud ies of electoral volatility in Latin America have not examined the impact of ethnic diversity on electoral volatility but have focused instead on economic and institu tional causes.5 The ethnic composition of societies in Latin America has significantly influenced electoral volatility, but not in the way that Bartolini and Mair and other studies of electoral volatility would predict. Electoral volatility has tended to be higher in areas with proportionally large indigenous populations, largely because the major parties in these countries have failed to address the needs and demands of the indigenous. Indigenous voters have consequently shifted their votes away from these parties and toward unstable new parties. Studies of indigenous political participation in Latin America have paid relatively little attention to the electoral behavior of the indigenous and its impact on party sys tems. Most have focused instead on the emergence of indigenous social movements.6 One important exception is a study by Van Cott, which found an inverse bivariate correlation between the proportion of the population that is indigenous and the degree of party system institutionalization in the region.7 She argues that this corre