Big men and ballots The effects of traditional leaders on elections and distributive politics in Zambia

This dissertation examines an inconsistency in the literature on African politics. Most scholars accept that African politics is "patrimonial"; politicians stay in power by building relationships with local big men, such as traditional chiefs, who can mobilize support for them. However, the vast majority of governments in Africa are now elected, and when voters choose their government in the secrecy of the ballot box, it is not clear that traditional chiefs can influence how they vote. An "institutionalist" perspective would suggest that chiefs' political views are irrelevant once the secret ballot has been instituted.;The first chapter develops a new explanation---the "cooperative interaction" theory---for why the political views of traditional chiefs continue to carry weight, even when governments are chosen in secret elections. In circumstances where elected politicians need to collaborate with local leaders, such as traditional chiefs, in order to implement local development programming, the views of these leaders affect the performance of politicians. Elected representatives will be able to collaborate more effectively with traditional leaders in situations where they have good pre-existing relationships with them. Other things being equal, voters prefer a political candidate with a stronger relationship with their chief, as they can expect to receive more local programming in these circumstances.;The main empirical component of the dissertation assesses the power of the patrimonial, institutionalist and "cooperative interaction" perspectives in explaining the influence of traditional chiefs in Zambia. The empirical analysis compares the influence of traditional chiefs within and across different Zambian communities. The second chapter justifies the decision to focus on chiefs in Zambia by comparing the powers of traditional chiefs in this country to those elsewhere on the continent. The chapter also provides historical background on Zambia's 286 traditional chiefs.;The third chapter considers Zambians' perceptions about the political influence of traditional leaders. Interviews that I conducted with politicians, chiefs and voters in Zambia demonstrate that voters and politicians care about chiefs' views of political actors, but that most Zambians do not think chiefs can dictate how their subjects vote. Instead, many respondents emphasized the importance of politicians being perceived to respect their traditional leaders, as this implies that the two leaders will be able to work together productively.;The fourth chapter examines the extent of chiefs' influence over the political behavior of their subjects. The results suggest that chiefs have a limited overall impact within their communities, but they have a large impact on those respondents who perceive greater need for their chief and the elected representatives to cooperate in the governance of their communities.;The fifth chapter examines how relationships between chiefs and members of parliament affect local development activities. An original survey of 110 chiefs demonstrates that the length of the relationship between members of parliament and chiefs increases the amount of local development activities in which they each engage. A larger data set that I compiled on educational investment in each chiefdom in Zambia demonstrates that locations where chiefs and MPs have longer relationships receive more educational programming.;The sixth chapter considers how chiefs' support for politicians affects election results drawing on a variety of data from the 2008 presidential by-election in Zambia. The results suggest that the governing party was able to win public endorsements from traditional leaders by giving them cars, but that these endorsements were worth little in terms of mobilizing electoral support and show that the length of the relationship between chiefs and the governing party's presidential candidate had a positive impact on local electoral support, even controlling for a suite of additional variables.;The final chapter of the dissertation reflects on the utility of the "cooperative interaction" theory in explaining the influence of traditional leaders elsewhere in Africa. The theory is broadly consistent with cross-national trends in the data, and secondary research from Malawi, Ghana and Senegal suggests the theory has some explanatory power in these cases. (Abstract shortened by UMI.)...