Measuring Campaign Effects in African Contexts: A Panel Survey on the 2011 General Elections in Uganda

Michigan State University, East Lansing MI

Investigators

Abstract

Since 1990, there have been 133 presidential and 162 legislative elections in forty-five of Sub-Saharan Africa's forty-eight countries. Despite the frequency of these events, the periods that immediately precede them--electoral campaigns--have received relatively little attention from political scientists. And even fewer studies have examined how individual citizens respond to campaigns in Africa. Democracy, of course, requires an engaged and informed citizenry, and the possibility for robust, public discussion of alternate policies and representatives. In the new, often struggling, and certainly imperfect democracies of Africa, it is especially important, therefore, to have the fullest possible understanding of the factors that impact societies' performance on these fronts. As essential institutions in democracies, election campaigns cannot be overlooked. Democracy promoters have focused extensively on ensuring that campaigns in developing democracies, such as those in Africa, facilitate open competition and are followed by cleanly conducted and peaceful elections. These efforts could be further enhanced by data on individual-level campaign effects, which this project will provide. This project represents a distinctive attempt to identify and measure a broad range of campaign effects in African settings. Extant literature on the subject treats electoral campaigns in Africa as primarily mobilizational. In Africa, party loyalty is often held to be determined by ethnicity or patronage relationships between elites and their poorer clients. Therefore, parties' campaign-time challenge is to encourage those loyalists to turn out at the polls. This project seeks not only to measure these mobilizational effects--and to identify campaign activities that might be most important here--but also to measure two other, largely overlooked, potential campaign effects in Africa. First, the project hypothesizes that campaigns might have important persuasive effects, resulting in individuals switching or developing new loyalties. Second, campaigns might have informational effects; through campaign communications, citizens might gain important knowledge about candidates' and parties' political preferences, and about political institutions and democratic practices more generally. In sum, the project seeks to measure campaign effects on individual citizens. The project's primary data-collection strategy involves a three-wave panel survey. Participants will be interviewed immediately prior to the launch of the campaign preceding Uganda's 2011 presidential and parliamentary elections. These same individuals will be interviewed again during the campaign, and then once more after the elections. Interviews will be conducted with approximately 1200 randomly selected citizens, from all of Uganda's regions and largest ethnic groups. Participants will answer questions about their exposure to campaign events and communications, including advertising; expectations for voting (in terms of whether they will turn out and, if so, which candidates they prefer); political attitudes on a number of issues; and level of political knowledge. Re-interviewing the same individuals over time means that the project will be able to measure the extent to which individuals become more politically engaged and knowledgeable over the course of election campaigning. It will also allow the researcher to examine whether citizens enter the campaign period with pre-established and rigid political loyalties, which are unaffected by candidates' and parties' campaign activities, or whether some subset of the population can be "won over" during the campaign. In sum, the findings of this project will facilitate further research in broad areas such as political party development, political communication and mass media, and electoral decision-making in developing democracies.