Suchergebnisse

This article estimates for member states of the EU the effect of low levels of turnout on parties' share of the vote in national elections. It does so by comparing the distribution of party choices in national elections for all those who participate in those elections on the one hand, and for the much more restricted group of those who participate in European Parliament elections on the other. As European elections register lower turnout than other nation-wide elections, this comparison provides an extreme, but empirically observed case of low turnout. Turnout effects prove comparatively small, and are non-negligible in only few cases. Turnout effects are slightly different for different kinds of parties: right-wing parties benefit slightly from them (on average) and left-wing parties are (on average) somewhat hurt. Although significant, these differences are exceedingly small, and explain no more than a few percent of variation in turnout effects. No significant effects are found from other party characteristics (such as their size, government status, position on European integration, or interactions of these with government approval or time since the last national election). The analyses are based on data from European election studies in 1989, 1994, 1999 and 2004. [Copyright 2006 Elsevier Ltd.]

Abstract: This article focuses on competition for votes between parties, as it existed in Western Europe in the period of the direct election to the European Parliament in 1989. Following earlier research by Van der Eijk and Niemöller, an instrument is introduced to measure the probability of party choice of EC citizens which establishes the likelihood of respondents to vote for any of the nationally relevant options/parties. A number of substantive conclusions about political parties'competitive performance result from this research. First, a single mechanism seems to structure electoral competition in all EC member‐countries. Second, the competitive performance of political parties is not affected by their governmental status, their ideological position, and the degree of politicisation of the electoral environment. And third, parties'competitive performance is strongly affected by the degree of uniqueness of their electoral potential, their mobilising capacities, their ideological extremity and their sheer size.

Abstract. This short note discusses a few general data‐related problems which occur in analyses of cyclical processes, and offers a brief overview of the advantages and disadvantages of the various approaches and techniques that can be employed. The conclusion is that one should be flexible in approach; it is always possible to model data in more than one way. A prudent strategy will always include the specification and testing of various models.

In the analyses reported in this report weights were used as follow: if a variable occurred in only one wave, the weight for that specific wave was employed. If a variable was measured in multiple waves, either the 'full' weight for all waves was used or a multiple-wave-weight that covered the specific waves of relevance. For the purpose of replication, the British Election Study and most of the Public Standards survey data can be freely accessed and downloaded from the UK Data Service data archives (https://discover.ukdataservice.ac.uk/).