UCLA

This dissertation examines the nature of party competition in the African democracies. Political parties are an integral part of contemporary democracy where their choices of favored policies create winners and losers and therefore supporters and opponents. Yet the role of parties vis-?-vis policy choice remains poorly understood in Africa. I scrutinize economic interests, whose policy preferences are collective but also exclusive, that form the bases of political support for parties. I argue that industrial sectors shape the basis of party support.

I examine case studies of African democracies that have experienced partisan alternation in power to learn how parties have strategically transformed the sizes and the dimensionality of their support bases and how they vary over time and across countries. My first case is Ghana where the two major parties have managed to develop stable, multiethnic support bases. While each party has its stronghold, economically and ethnically defined, a candidate cannot win the presidency without appealing to the unattached voters. The example of Zambia shows how parties can and do adapt the “shapes” of their support bases by shifting the dimension of political competition from ethnic cleavages to policy issues and by narrowing the range of a targeted support base. While a largest voting bloc on any dimension is often sufficiently large to set the basis of a winning coalition, if politicians are successful in reducing its size to a smallest winning, they gain most benefits possible. But if they overshoot, they can lose everything. In my third empirical chapter, I examine how economic interests based on agricultural subsectors account for seemingly ethnic coalitions in Kenya. The case study highlights voting behavior of cross-pressured voters whose ethnicity and economic stakes pull them in different directions. The vote decisions by the co-ethnics of the third place presidential candidates reveal that the sector-based voting provides a powerful explanation of the political coalitions even in ethnically divided countries. The main argument is that African parties are “normal,” in that they do not exclusively trade in clientelistic favors for ethnic kin, but also offer policy promises to attract broader support.