Agenda setting and public opinion formation: Media content and opinion polls on divorce referenda in Ireland

This dissertation traces the development of the theories of public opinion and agenda setting. It then applies both theories in understanding the role that newspaper coverage plays in the formation of public opinion. The case study is the formation of public opinion on the topic of constitutional change on the issue of divorce in Ireland. The 1937 Irish Constitution banned divorce. In 1986 the first divorce referendum in Ireland was held in an attempt to change the constitution and was defeated. In 1995 the second referendum was held and was passed by a very narrow majority. Newspaper stories about divorce were content analyzed on a number of variables--the prominence of divorce stories, the number of such stories, and the valence of those stories vis-a-vis constitutional change. Newspapers were sampled for a three year period preceding each of the divorce referenda. The dissertation looks at 27 opinion polls dealing with the issue of constitutional change on divorce. It then attempts to relate changes in opinion trends in those polls with newspaper content from four national newspapers--two dailies and two Sundays--using time lagged regression analysis. The analysis is applied to the pro-divorce trend, the anti-divorce trend and the don\u27t know/no opinion trend. Major findings include a lack of difference between content in media campaigns, high negative correlation between media support for divorce and public opinion against divorce, evidence of media content being highly correlated with specific poll outcomes, and a definite pattern of media influence on poll outcomes. The analysis suggests that the media influence the public agenda, and that the public influence the media agenda. The data also suggest that there are other social/institutional forces at work in the process of public opinion formation which are deserving of further research. It appears that media content plays a greater role in those constitutional ballot questions in which the support and opposition sides are close in size, and have little influence where there is a great deal of difference in size between the opposing groups. Of special interest in this analysis is the don\u27t know/no opinion grouping, given that the second referendum was decided by a majority of only 50.03 percent