Proximity Models

The Downsian Proximity Model

To establish a baseline, we describe in some detail the classic Downsian proximity model. As in all spatial models, both voters and candidates are represented by points in a multidimensional space that reflect their opinions on issues; each issue corresponds to one dimension of the space. In the simplest case, there is only one issue (say, economic policy) and, hence, one dimension. At this level, the spatial model is no more than a formalization of the familiar left-right or liberal-to-conservative scale, represented by a line. By convention, numerical values on this scale increase from left to right. Other things being equal, it thus seems plausible that, say, a liberal voter on economic policy will prefer a candidate who is liberal on that issue to either a middle-of-the-road or conservative candidate and will prefer a moderate over a conservative. This suggests that the voter's evaluation, or utility, may be based on proximity in the spatial model, with highest utilities reserved for candidates located close by on the policy scale. This is the basic idea in the Downsian proximity model.