Estimating voter transitions by ecological regression

Abstract

This article focuses on the development of an ecological regression approach for voter transition estimation, avoiding the arbitrary assumptions in Goodman's classical model of ecological regression (Goodman, 1959). In doing this, we further develop previous attempts made at the ridge regression approach in ecological regression (made, for example, by Miller (1972), Brown and Payne (1975)). To illustrate the application of the proposed estimation technique, we include a comparative study based on a set of actual election data, where we analyse whether our own estimates of transition probabilities are congruent with patterns obtained from other models. We find that the ecological regression model using the modified estimator results in estimates which are superior to those obtained from survey data.