Campaigns, the media and insurgent success : the Reform party and the 1993 Canadian election

It is well recognized that the 1993 election campaign catapulted the Reform party into the national political scene, but our understanding of how this was possible is quite limited. Drawing on the work in cognitive psychology on attitude change, the work on the news media coverage of elections, and the political science work on election campaigns, this thesis locates the impetus for Reform's success in the dynamic flow of information about the party that was available in television news broadcasts and voters' likelihood of being persuaded by that information. This link is developed by an analysis that makes use of a content analysis of the 1993 campaign, the 1993 Canadian Election Study, and a merged analysis of the election and news data. The Reform party began the campaign as a minor component of the news coverage of the election, but the news media coverage changed dramatically. Reform was provided with more news access than its support indicated it deserved and that coverage focused on what became a major theme of the election; the welfare state and the role of government. Coverage of Reform underwent a further change as it both decreased and focused on cultural issues during the last two weeks of the campaign. Using a two-mediator model of attitude change, the analysis shows that people who were predisposed to agree with Reform's anti-welfare state message and who were likely to be aware of the news information, changed both their perceptions of the party and increased their support for the party. Further support for the impact of the media is derived from the analysis of voter response to the second change in news coverage. The analysis suggests that campaigns do matter, but that the size of the impact is dependent upon the underlying uncertainty associated with the parties and candidates, and on the degree to which the information flow of the campaign changes. The information flow contributes to both learning and priming among people who receive and accept new information. While voters respond reasonably to new information, the outcome will depend on what information voters are given and what information actually reaches the habitually unaware segments of the population.Arts, Faculty ofPolitical Science, Department ofGraduat