Do economic differences or institutions explain variance in the economic vote among countries

Despite all the methodological problems in defining the proper variables and methods to estimate it, an association between economic perceptions and vote choice or vote outcomes in the expected direction seems to be a regular feature of voting behavior. Nonetheless, there is important instability among elections and countries that tends to be attributed to methodological flaws or to contextual differences. Three basic research questions are addressed in this paper: the weakness and instability in the effects of the economic retrospective vote (ERV) results from contextual differences among countries or from measurement errors and bias in the models used? Do differences in the general economic contexts of countries explain variation in the effects of the ERV? If so, is it variance in the economic vote among countries or are elections better explained by differences in general economic context or by differences in institutional arrangements? The data used in this paper come from 31 post electoral surveys taken in 29 countries, as part of Module 1 of the Comparative Study of Electoral Systems. Multilevel models were used since they are geared towards the statistical analysis of data that have a hierarchical or clustered structure. Relatively small effects of economic perceptions on vote choice are found, and no consistent explanations of variance among countries associated with economic or institutional contextual variables could be identified. These findings do not seem to be the consequence of serious measurement errors, but of insufficiently specified variables and models. The number of country-elections at hand and the variance in institutional arrangements are too small.