On estimating distance-determined voting functions

Abstract

The distance from one's home to a proposed project site is often the important factor determining an individual's opinion whether such a project should be built. This article first discusses why regression is inappropriate to analyze such preference functions at the individual level. It then examines alternative estimation techniques, concentrating on the utility and limitations of probit. Probit is particularly useful for spatial analysis because an indifference distance, i.e. the distance from a project at which an individual is indifferent to whether it is built, can be estimated and used to map such preferences.