This study employs a decision model to study the effects on popular support of candidate chances and candidate moderation in the American presidential nomination process. The decision model draws upon psychological and rational choice theories of decision making to accommodate both the complexity of choice citizens face in the typical nomination campaign, and the need to keep decision costs in bounds. We use the model to simulate voter decision making under repeated trials of nomination campaigns. The simulation trials vary the chances candidates have of winning their party's nomination, and the ideological moderation of competing nomination candidates. Our results show that increasing a candidate's chances for the nomination can have substantial payoff in support, but that the consequences of moderating a candidate's ideological stand are more muted. Our research speaks to broader guest ions of decision making when information costs are high but cognitive effort remains low, and to the resolution of intraparty conflict in a dynamic and volatile political environment.

The literature on new political parties is vast but often neglects strategic aspects that are important in the formation period. In this article I develop a simple incomplete information game consisting of the interactions between an established party and a potential new party. The game shows that higher ‘costs’ for forming a party decrease the likelihood of new actors appearing on the electoral scene. Apart from this intuitive result the paper also shows that important demands from potential new parties are less likely to be taken seriously by the established party than more modest demands. Consequently, established parties often change their policies only marginally and are likely to refuse important modifications to their platform.

This paper uses newly available Canadian data to address long‐standing debates on the rationality of the political economy of party support. We find that the subjective economic variable driving governing party support is sociotropic, not egocentric, and, pace recent American and British studies, prospections do not dominate retrospections. Rather, models using national economic evaluations encompass rivals employing national expectations, personal expectations and perceived trends in personal expectations. Egocentric considerations are not irrelevant; rather, their effects on party support are indirect. We argue that these findings are consistent with an image of voters whose party‐support decisions are governed by a ‘rough‐and‐ready’ rationality appropriate to the information available to them and the politico‐economic systems of contemporary Western democracies.

The purpose of this paper is to test a rational actor model of political participation against rival models to see which provides the best theoretical account of participation. This is done using data from the first panel survey of Labour party members in Britain. The tests use encompassing methodology, an approach to model building which arises out of recent developments in applied econometrics. The rational choice model is tested against a social‐psychological model, and a ‘general incentives’ model which has been utilized in earlier research to explain party activism. The results show that while the rational choice model has important explanatory power, it provides an incomplete account of participation. A ‘reduced form’ model is developed which encompasses the other models, and provides the best explanatory account of party activism.

Why do sudden changes occur in electoral turnout from one election to the next? This article proposes two general hypotheses as explanations for declining turnout. The first hypothesis draws on rational choice theory, and is based on the assumption that elections are devices for the selection of governments. It states that abstention is rational when government alternatives are unclear. The second hypothesis implies that non‐voting is a result of social‐psychological mechanisms, and suggests that voters abstain in situations of disagreement with their own party concerning some important issue. The hypothesis predicts that strong identifiers under cross‐pressure are more inclined to stay home than weak identifiers and non identifiers. In using the 1993 parliamentary election in Norway as a test case, the analysis suggests that the application of a rational choice approach is fruitful for studying electoral turnout. Because of the absence of clear‐cut government alternatives voter involvement was low in the election and people did not distinguish much between the various political platforms offered to them. The second hypothesis of disagreement with own party received only limited support. The differences due to cross‐pressures are not very big, and the prediction that strong identifiers under cross pressure are more inclined to stay home was not supported. Altogether, our case study suggests that very few costs are involved in changing partisanship and that party identification reflects a strong cognitive rather than affective aspect.