The Clarity of Policy Alternatives, Left–Right and the European Parliament Vote in 2004

Abstract

The importance of the left–right divide for party choice is well established, both for legislative and European Parliament (EP) elections. However, the conditions under which left–right self‐placement becomes more or less important in explaining the vote in both legislative and EP elections are clearly understudied. The article uses the 2004 EP elections as a laboratory to understand if there are indeed systematic differences between political systems’ characteristics that might explain variation in terms of the strength of the relationship between left–right self‐placement and the vote. Using the survey data from the European Election Study 2004 (twenty‐one EU member states), the paper has two goals. First, to examine whether citizens’ left–right self‐placement has a different impact on the vote in different types of democratic regime, defined in terms of the contrast between consolidating and long‐established democracies. Secondly, to examine whether this contrast resists the introduction of controls for three other factors hypothesized to make a difference in the extent to which left–right orientations have a greater influence on the vote: the permissiveness of electoral system; the clarity of policy alternatives provided by the party system; and the particular type of party alignments along both the left–right and anti‐/pro‐integration scales that tend to characterize each country. Our findings corroborate that (the 2004 EP) elections do seem to be about choosing parties in terms of left–right orientations to a considerable extent. Furthermore, we found that the usefulness of left–right orientations as cues to the vote seems to be contingent upon a major contextual factor: greater levels of clarity of the policy alternatives provided by the party system render citizens’ left–right self‐placement more consequential for their EP vote. Finally, we found that left–right orientations may not be equally useful in consolidating and in the remaining established democracies.

Key Words:

• Left–right
• electoral behaviour
• European elections

Acknowledgement

A previous version of this article was originally presented at the American Political Science Annual Meeting (APSA), 1–4 September 2005, Washington, DC: DIVISION 36‐12 (Co‐sponsored by DIVISION 15‐21). The authors would like to thank Gary Marks (chair and discussant), as well as the other participants in that session (Hermann Schmitt, Michael Marsh, Mark Franklin, Radoslaw Markwoski, Bernard Wessels, Jacques Thomassen and Richard Gunther) for their very insightful suggestions and criticisms that helped to improve the present revised version of the paper. More recently, an earlier version of the paper was presented at the EES Spring Meeting 2006 on The European Parliament Election of 2004, organized by the Institute of Social Sciences of the University of Lisbon (ICS), Lisbon, 12–13 May 2006. The authors would like to thank to all the participants in that session for their very insightful suggestions and criticisms that helped to improve the present revised version of the paper. This version was published as Freire, Lobo, and Magalhães (2007). Special thanks go to José Pereira and Edalina Sanches for valuable help in data management and analysis. Of course, all the problems that remain are the authors’ exclusive responsibility.

Notes

1. The data utilized in this publication were originally collected by the 2004 European Election Study research group. The group consisted of Stefano Bartolini (EUI Florence, Italy), Cees van der Eijk (now University of Nottingham, UK), Mark Franklin (Trinity College, Hartford, Connecticut, USA), Dieter Fuchs (University of Stuttgart, GFR), Michael Marsh (Trinity College, Dublin, Ireland), Hermann Schmitt (University of Mannheim, GFR), and GáborToka (Central European University, Budapest, Hungary). This study was made possible by various grants. Neither the original collectors of the data nor their sponsors bear any responsibility for the analyses or interpretations published here. The data are still under embargo, except to the research directors referred to above, and the national research directors in each country. The authors of the present paper are the national research directors of EES 2004 in Portugal.

2. The only cases considered in the Van der Eijk, Schmitt, and Binder (2005) article were Germany, the Netherlands, Norway and Sweden.

3. Those who were still studying were recoded using the year of birth variable. Also France, Poland and Slovakia had a different coding for the education variable and these were harmonized.

4. See Lijphart (1994, 1999).

5. The interpolated median is computed in the following way. First, the variables are defined: M is the standard median of the responses; nl is the number of responses strictly less than M; ne is the number of responses equal to M; and ng is the number of responses strictly greater than M. Secondly, the interpolated median IM is then computed:

6. Traditionally, the arithmetic mean and the standard deviation (or some transformation of the latter) are used to describe such a frequency distribution in terms of central tendency and dispersion. In the case, of finite ordered rating scales these measures can be demonstrated to be biased (by extreme values of the distribution). With respect to the central tendency, see Huber and Powell (1994) and Herrera, Herrera, and Smith (1992), who recommended the interpolated median.

7. Similar measures of ideological distances at the party system level were used in Berglund et al. (2005), Knutsen and Kumlin (2005) and Freire (2006a, 2006b). When using voters’ perceptions of parties’ location in the left–right divide, the major differences vis‐à‐vis the present article is that they used the ‘mean’ for parties’ locations and we used the ‘interpolated median’ value. We believe that the latter value is more accurate because, first, it is less sensitive to extreme values of the distribution and, secondly, it has got a more substantive meaning. See van der Eijk and Franklin (2004) for a different methodology.

8. The question (asked both for respondents’ location and for parties’ location) is stated in the following way: ‘Some say European unification should be pushed further. Others say it already has gone too far. What is your opinion? Please indicate your views using a 10‐point‐scale. On this scale, 1 means unification “has already gone too far” and 10 means it “should be pushed further”. What number on this scale best describes your position?’

9. Religiosity was coded in a five‐point scale of mass attendance, from 1 (never) to 5 (several times a week). Social class is a (subjective) self‐placement five‐point scale, from 1 (working‐class) to 5 (upper class). Education is a continuous variable capturing respondents’ ages when they stopped studying (or current, if still studying).