Challenges to the Standard Euclidean Spatial Model

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Spatial models of political competition over multiple issues typically assume that agents’ preferences are represented by utility functions that are decreasing in the Euclidean distance to the agent’s ideal point in a multidimensional policy space. I describe theoretical and empirical results that challenge the assumption that quasiconcave, differentiable or separable utility functions, and in particular linear, quadratic or exponential Euclidean functions, adequately represent multidimensional preferences, and I propose solutions to address each of these challenges.

This working paper is meant to be published as a chapter in the volume “Advances in Political Economy”, edited by G. Caballero, D. Kselman and N. Schofield. I thank Scott Tyson for suggestions. Comments to ammend errors or to provide updates to the working paper are welcome even after the publication of the volume.