![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
This paper looks at German regional elections and evaluates the patterns of voting that can be observed in aggregate results and individual behaviour. In the first step, I summarize the characteristic features of political competition in the German multi-level system. I use elections held between 2013 and 2017 to illustrate contextual and election-specific factors that contribute to the observable patterns of voting. In a second step, I evaluate the possible cyclicality of second-order voting using individual data. Based on 139 state election surveys covering 1978 to 2016, I show that the decision-making of German voters changes systematically over the federal election cycle. More specifically, the effect of party identification on vote choice depends on the timing of the election, adding another dimension to the already multi-faceted relationship between Land elections and federal politics.
Acknowledgements
An earlier version of this article was presented at the workshop ‘New Avenues in Regional Political Science’, held at the Hanse-Wissenschaftskolleg, Institute for Advanced Study in Delmenhorst on 14–16 April 2016. I thank all participants of the workshop, the editors of this special issue and three anonymous referees for their helpful comments and suggestions.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1 These instances were among the main topics of recent attempts to reform German joint federalism (see Stecker, Citation2016).
2 In all states, there is a local tier and with the exception of the city states of Berlin and Hamburg, all states are divided into counties.
3 These states are Baden-Württemberg, Bavaria, Hesse and North Rhine-Westphalia.
4 Initially, 11 states were formed in West Germany and five in the Soviet occupation zone, whereas the latter were replaced by 14 districts in 1952. There were also changes in the Federal Republic: in 1952 three states merged to form Baden-Württemberg and in 1957 the Saarland – which was a French protectorate till then – became part of (West) Germany again. Both changes were the result of referenda. Until reunification, West Berlin provided for a specific case from a legal point of view but was de-facto a state of West Germany as well (Plischke, Citation1965).
5 For an evaluation of the consequences of the electoral system, see Raabe et al. (Citation2014).
6 This also applies to the specific ways votes are cast and translated into seats in case of the electoral systems of Baden-Württemberg (Trefs, Citation2003; Däubler, Citation2017) and Bavaria (Rudolph and Däubler, Citation2016: 751f.).
7 Most states decided to switch to legislative terms of five years in in the 1990s and 2000s. Only Saarland (in 1955) and North Rhine-Westphalia (in 1970) switched before.
8 The Hamburg ‘STATT party’ (1993) or ‘Arbeit für Bremen und Bremerhaven’ [work for Bremen and Bremerhaven] (1995) are examples for state-specific parties that enjoyed some, though not persistent, success. The right-wing extremist/populist parties that entered German state parliaments were the AfD, the Republicans (REP), the German People's Union (DVU) the National Democratic Party (NPD) and the Party for the Promotion of the Rule of Law (PRO).
9 See Detterbeck and Jeffery (Citation2009) for a discussion of party organizations in Germany and the role and autonomy of regional party branches.
10 The SPD’s result in Baden-Württemberg was the worst in any state election in West Germany (Gabriel and Kornelius, Citation2016: 503).
11 In the past several minority governments were formed at the regional level. However, most of them were rather short-lived.
12 Coalitions with the PDS/The Left were rejected by many Social Democrats for a long time, particularly in the West German states. Cooperation at the federal level still seems to be unlikely because of the party’s positions on foreign policy (Faas, Citation2015: 245).
13 There were earlier instances where minority coalitions formed by the SPD and the Greens were supported by the PDS.
14 Established measures like the 'party vote share change', 'government congruence' or 'electoral strength for non-statewide parties' (see Dandoy and Schakel, Citation2013; Schakel and Jeffery, Citation2013) appear to be less helpful in this context.
15 A third or other party here refers to a party that is not represented in the (national) legislature.
16 There are also alternative approaches to evaluate the differences between second-order and first-order elections with regard to voting behaviour: One possibility is to compare voting behaviour, say in a regional election with voting behaviour in a national election, using panel data (see, e.g. Giebler and Wagner Citation2015). However, such information is rarely available and if so, it usually covers only a small number of elections. A related approach is to make use of surveys that include questions on past voting behaviour or voting intentions for several elections. Theoretically, such data are panel-like; however, it is questionable to what extent respondents are able and willing to recall the information respectively to make such abstract predictions about their own behaviour.
17 For the variable that captures the voters’ participation in the election, the possible answers differ slightly across surveys. In many cases, respondents could indicate how (un)likely a participation is. In this case, all respondents except the ones who indicated that they will definitely participate (or cast an absentee vote) were coded as abstaining.
18 There are different ways to capture time-variance. I opt for including a cubic polynomial time trend, not only because it is easy to implement and interpret, but also because it allows for the uncovering of different curves. This is helpful as we still have to identify the most appropriate description of the cyclicality of second-order voting.
19 The full results of the underlying regression model can be found in the Appendix.
20 Ganghof et al. (Citation2012) evaluate the legislative activity of the red-green minority coalition that was in power in North Rhine-Westphalia from July 2010 until June 2012. Their study indicates how minority governments can facilitate policy change in the German case.