Recently the spatial theory of elections has been challenged by critics who dispute the realism of its assumptions and the empirical veracity of its predictions. This paper critically evaluates two theoretical alternatives to the spatial model and subjects these to a series of strategic tests using data from six West European party systems. It is argued that a model of representational policy leadership, which combines insights from the spatial and the theories of voting, best accounts for the observed patterns of voting. Thus, voters prefer parties that offer clear and intense political alternatives, but they turn away from parties that deviate too radically from voters' own stated policy positions. In terms of party strategies, it is demonstrated that spatial theorems, contrary to claims made for the directional theory, hold for all three models of voting. The policy leadership model, however, does explain why parties in multiparty systems tend to be more dispersed than predicted in spatial theory.
Stored only when both papers are in the corpus — 229 citations exist upstream.