Economic Voting and Issue Ownership. An Integrative Approach

Sammanfattning: The aim of this dissertation is to improve our understanding of how economic changes influence popular support for political parties. Previous studies have generally found that there is a tendency for the popular support of incumbent governments to diminish when the economic development deteriorates. The extent of the electoral punishment has, however, been found to vary strongly between different studies. Several institutional and contextual explanations for this variation have been proposed in earlier studies. The main argument of this dissertation is that, when we analyse how economic changes impact government support, we should simultaneously consider changes in economic evaluations and changes in the public agenda and issue ownership. This is the integrated model of economic voting and issue ownership. In an economic crisis with rising unemployment, for example, a government is likely to be electorally punished for deteriorating economic evaluations. But, if the incumbent still owns the issue of unemployment, it is also likely to be rewarded electorally when that issue rises on the agenda as a result of the crisis. In such a case, the latter effect might counteract, and possibly cancel out, the former. A set of requisites is specified and examined to test whether the integrated model is important to our understanding of economic influences on government support. This study focuses empirically on the case of Sweden, and its special focus is the issue of unemployment. The main data sources are the Swedish National Election Studies, the SOM institute at the University of Gothenburg and the Swedish National Institute for Economic Research. The results indicate that issue ownership is generally more volatile than previous studies have usually assumed. However, if issue ownership is fairly strong, bad performance such as rising unemployment levels does not immediately threaten the incumbent’s issue ownership. At the individual level, it is also shown that the effects of economic evaluations and the issue sailence of unemployment do counteract and reinforce each other in accordance with the pattern predicted by the integrated model. For the period of 1988 to 2002, the integrated model seems highly relevant for understanding how economic changes have influenced government popularity and election outcomes in Sweden. When the incumbent owns the issue of unemployment, electoral punishment for a deteriorating economy is substantially mitigated. For the 2006 election, however, the integrated model does not appear able to improve our understanding of how the economy influences election outcomes. This dissertation demonstrates that future studies trying to explain differences in the extent of economic voting have good reason to take issue ownership into account.