Content area
Full Text
A Unified Theory of Voting: Directional and Proximity Spatial Models. By Samuel Merrill III and Bernard Grofman. New York: Cambridge University Press, 1999. 213p. $54.95 cloth, $19.95 paper.
Keith T. Poole, University of Houston
A Unified Theory of Voting is a "must read" for anyone wishing to stay current with the controversies in modern spatial theory. The book is somewhat uneven in exposition, but Merrill and Grofman do an excellent job of developing an elegant synthesis of several important spatial models and explaining how each model works. Part 1 is concerned with spatial models of voter behavior, and Part 2 covers spatial models of candidate/party behavior and strategy.
In the classic Downsian spatial model, citizens and politicians are assumed to be points in a multidimensional policy space. When a citizen votes, she chooses the politician closest to her in the policy space. This model has come under strong attack during the past 13 years from the directional theory of voting developed by George Rabinowitz. In contrast to the Downsian model, the directional model states that a voter chooses the politician who will change policy in her direction. She bases her voting decision upon direction, not distance. For example, in a simple one-dimensional world the citizen could be just to the right of zero (the "neutral point") with two candidates, one just a hair to the left of zero and another to...