The Kretschmann Effect: Personalisation and the March 2016 Länder Elections

Abstract

This paper examines the influences of candidate perceptions on Germany’s spring 2016 Länder election results. It takes a comparative approach, using a modified Michigan model on the data collected simultaneously in three Länder (Baden-Württemberg, Rhineland-Palatinate, and Saxony-Anhalt). It explains why the Green party was successful in Baden-Württemberg but not in the other Länder, the impact of the major candidates, and what distinguishes the influences of the current prime ministers running for these elections. Whereas Winfried Kretschmann’s (Green party, Baden-Württemberg) high impact on the election results was driven mainly by a warmth dimension (sympathy), Malu Dreyer (SPD, Rhineland-Palatinate) was viewed as being competent. Both candidates were assets to their parties and co-responsible for the results. In comparison, in Saxony-Anhalt, none of the candidates were as important to the outcomes of the electoral success.

ACKNOWLEDGEMENTS

We would like to thank the two anonymous reviewers as well as the editors, Kristi Winters, Julia Chalupa and Karl-Heinz Naßmacher, for their comments, which helped to improve this paper significantly.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

ABOUT THE AUTHORS

Johannes N. Blumenberg studied political science and sociology at the universities of Oldenburg and Heidelberg. From 2010 to 2013 he worked as research associate at the Mannheim Centre for European Social Research (MZES), conducting research on Länder elections and referendums. Afterwards, he joined the chair of Empirical Political Research at the Johannes Gutenberg-University Mainz. Since 2016, he has acted as manager of the GESIS doctoral programme at GESIS – Leibniz-Institute for Social Sciences. His research focuses on electoral behaviour, parties and party systems, and political psychology (especially processes of identification and stereotyping).

Manuela S. Blumenberg is currently working at GESIS in the GLES project (German Longitudinal Election Study). She studied social science at the Carl von Ossietzky University of Oldenburg and the Södertörn högskola (Huddinge, Sweden) and received her doctorate with the title ‘Ausgabenstrukturen demokratischer Parteien im internationalen Vergleich’ in 2012, supervised by Prof. Dr. Karl-Heinz Naßmacher and Prof. Dr. Wolfgang Rudzio. Her research focuses on electoral behaviour, electoral and party systems, and comparative party financing.

Notes

1 The process itself thereby is a cognitive one which paves the way to base party identification within the social identity approach (e.g. Green, Palmquist, and Schickler 2002).

2 Campbell et al distinguished between ‘record and experience’, ‘qualification and abilities’, and ‘personal qualities’ (Campbell et al. 1990, 54–55).

3 The dimensions warmth and competence are usually used within the literature which discusses the stereotype content model (Fiske et al. 2002). However, as Frank Asbrock (2008, 48) was able to show, political science also uses those dimensions regularly, even though the labels differ.

4 We therefore also recalculated our analysis using thermometer scales for all candidates (including Green party, Left party and so on) instead of our dimensionality items. However, our results are very robust to those changes.

5 As an alternative to this proceeding we could also have used the competence on solving the most important problem(mip)/the most important problems, as has been done before by other authors. However, since our measurement inherits more variance in answers, which has a positive effect on the estimation (lowers multicollinearity), we decided to stay with this measurement. However, again, we also tested using mip-scales instead of our measurement, which did not significantly affect the results of the analysis.

6 As of 13 July 2016 Infratest dimap listed on its website in the so-called ‘Sonntagsfrage (bundesweit)’ that the vote share rose from 15 to 21 per cent at the federal level between 10 and 24 March 2011 (Infratest dimap 2016a). Other non-university based research institutes offer similar predictions. This is also true for the Länder elections.

7 The Green Party has not been part of the Landtag since 1998.

8 As of 13 July 2016 Wahlrecht.de listed on its website the latest forecast of the Landtagswahlumfragen before the elections, ranging from 9 per cent in Rhineland-Palatinate, to 11 per cent in Baden-Württemberg and 18 per cent in Saxony-Anhalt.

9 However, the question used by Infratest dimap here is oversimplified and the result should not be taken too seriously. Not all who said they voted for the AfD because they were disappointed were protest voters, and vice versa. See also Schwarzbözl and Fatke (2016).

10 Pedersen-Indices: Rhineland-Palatinate 14.4; Baden-Württemberg 23.4; Saxony-Anhalt 24.1; Pedersen-Indices without the AfD: Rhineland-Palatinate 8.1; Baden-Württemberg 15.8; Saxony-Anhalt 11.9. Please see Pedersen (1979) for calculation and interpretation.

11 Whereas age does neither indicate qualification nor individual characteristics of the candidates, the effects of candidates’ ages on turnout of young voters have been proven elsewhere (e.g. Pomante and Schraufnagel 2015). Additionally—especially in the run-up to the 2011 election in Baden-Württemberg—it has been part of the discussion whether Kretschmann was not already too old to become prime minister. For this reason we included age as information for the readers.

12 However, we have tested their effects by using alternative measurements – namely thermometer scales – which supports our argumentation.

13 Since party identification with the Green party predicted the outcomes perfectly, this variable has been excluded in Saxony-Anhalt. We selected this approach instead of dealing with the separation (see for example Rainey 2016), since dropping it should have raised the estimated effect rather than lowering it. Thus, in the worst case, the candidate effects have been underestimated.