Lo studio del comportamento elettorale si è considerevolmente modificato dai tempi della supremazia del Michigan model negli anni Sessanta (Campbell et al. 1960 e 1966) e della congiunzione di quest'approccio socio-psicologico con quello della scelta razionale di Downs (1957).

«Reasoning voter» is a term coined by Popkin to describe a situation in which voters do reason about parties, candidates and issues in order to come to terms with a low-cost decision. The acquisition of perfect information and its processing are too expensive so that reasoning voters use information and calculation shortcuts to approach rationality in their choice.

In a field where research agenda focuses heavily on the American two-party system and its presidential election, this paper is about preference formation and party choices of reasoning voters in a multiparty system. More specifically, the Author discusses a model of voting decision which is built on the assumption that party preference profiles of voters are the crucial link between the factors influencing the reasoning about parties and the final voting decision in a multiparty system.

The first section deals with the concept of party preference in multiparty system, its theoretical status and the different devices to collect data on party preference profiles (paired comparisons, rankings, ratings). The second section discusses the factors influencing the preference formation. From those factors, issue proximities and retrospective evaluations are selected as the factors most proximate to the process of preference formation in the assumed funnel of causality. The Author also takes into account future expectations too, but they are more difficult to integrate into models for multiparty systems, since the future government does depend not only on election results but also on future coalition building. The third and last section focuses on voting behavior. In the analysis of the decision problem itself the concept of habit, the expressive consideration together with the instrumental one and the distinction between voters as consumers and voters as investors are considered.

The final question is about the voter paradox, that is partecipation in an election with an infinitely small impact on the outcome, appears as paradoxical as ever. The Author argues that, since continental European democracies have developed parliamentary systems in which governments are usually formed by coalitions, it is very difficult for voters to anticipate future governments, and this constitutes a further element against the decisiveness of voting. Thus instrumental rationality is more downgraded in multiparty systems than in two-party systems. But multiparty systems on the other hand facilitate political orientation by providing ideological signals.