| This dissertation is a theoretical and empirical test of a series of propositions relating to the influence of focusing events on the federal institutional agenda. The theoretical foundation for the dissertation is provided by the social science literature in agenda setting in government and in the news media. The theory consists of three major sets of propositions and several hypotheses relating to the propositions that are tested in the subsequent empirical case studies. These propositions relate to the influence of focusing events on the news agenda, the influence of focusing events on the institutional or governmental agenda, and the influence of focusing events on the composition of the policy community.;The empirical element consists of of an assessment of the propositions in four policy domains: earthquakes, hurricanes, oil spills, and nuclear power plant accidents. The results of the study suggest that while focusing events do, in general, influence the agenda as hypothesized and as might be expected intuitively, differences between the policy domains and the characteristics of the events themselves have a substantial influence on agenda-setting and political outcomes in the wake of focusing events. Differences in event characteristics include the nature of the event, including whether the event was natural or humanly-caused and whether the disaster or accident is considered "dreadful" in risk terms. Differences in the composition of the policy community include the number and type of groups that participate in policy making, the intensity of that participation, and the substance of the issues raised by participants in the debate. The result is that the focal power of an event requires that there be an existing policy community that can use the event as an example of why policy change is needed. Where there is intense conflict over the interpretation of the meaning of an event, however, the focal power of an event may be tempered.;Based on these results, comparisons between the domains are made, and suggestions for generalizing the results beyond the four domains are offered. The prospects for extending the study to other levels of government and other domains are noted. |
