A set of models of voting behaviour in which a voter's strength of preference for a candidate depends on whether the voter and the candidate take the same side on a policy issue/dimension. These contrast with the theory of spatial competition (or proximity model) in which voters prefer the candidate closest to them in the policy space irrespective of whether they lie on different sides of a neutral (or status‐quo) point. In its simplest form, voters prefer candidates on the same side to those on the opposite side of a single dimension. If the single dimension is the left–right dimension, right‐wing voters would prefer all the right‐wing candidates to the left‐wing candidates, even if they are closer to a left‐wing candidate. This model can be extended to allow for different dimensions of varying importance or salience. Rabinowitz and Macdonald (1989, American Political Science Review) elaborate further by incorporating ‘intensity’. Both voters and candidates can vary in the extent to which they take a particular stance. This is modelled as distance from the neutral point. The utility function (strength of preference for a candidate) is the product of the voter and candidate intensities, accounting for direction. This implies that right‐wing voters prefer the most right‐wing candidate, no matter how right‐wing they are themselves. The model therefore predicts that candidates have an incentive to be the most extreme candidate on their side of any particular dimension. Since this is an unattractive feature of the model, mechanisms have been proposed to constrain movement, such as a ‘region of acceptability’ outside of which candidates are penalized.