Voters, Information Heterogeneity, and the Dynamics of Aggregate Economic Expectations

Theory: It is argued here that previous studies analyzing the formation of voters' economic expectations overlook important subtleties found in a differentiated (heterogeneous) electorate. Since different segments of the electorate possess varying information capabilities, it is asserted that they will form their prospective economic evaluations differently from one another. Hypotheses: The impact of expectational electoral cycles, news coverage of the economy reported by voters, retrospective economic evaluations, and personal financial expectations are hypothesized to have a differential impact on their economic expectations based on varying levels of information (education) among subsets of the electorate. Methods: Time series multiple regression analysis is conducted for the monthly period between January 1978 and December 1990. The empirical analysis involves testing for the properties of the statistical time series (i.e., integration and cointegration) as well as specification of an error correction mechanism (ECM) that captures long-run statistical relationships in appropriate circumstances. Results: These findings uncover electoral heterogeneity (based on information/education differences) with respect to voters' prospective economic evaluations. Specifically, the information capabilities of voters is inversely related to their reliance on past/retrospective sources of information when forming both egocentric and sociotropic economic expectations. The main implication of this research is that the assumption of electoral homogeneity contained in previous longitudinal analyses of aggregate economic opinion formation provides misleading results that either overlook intra-electoral variations, introduce aggregation bias, or possibly both.