District candidacies for the German Bundestag: what spurs or hampers intra-party competition?

ABSTRACT

The article analyses district candidate selections as an arena for intra-party competition. To investigate what spurs or hampers contested races, we focus on Germany. Relying on data from the Bundestag election in 2017, the first part of our paper looks at factors that statistically stimulate rivalry at the nomination event. Subsequently, we explore the ‘secret garden’ of preselection by drawing on interviews to examine whether intra-party competition might be muted prior to the selection conferences. A key determinant for competitive nominations is a vacant candidacy rather than the chance of winning the district or the selection mode. Moreover, the involvement of several regional party units lowers competition. Interestingly, there would be more contested nominations if it were up to the nominating selectorates. However, things look somewhat different regarding the local party leadership: In part, ‘coronations’ are caused by early withdrawals of aspirants due to the leadership’s influence as steering agent.

KEYWORDS:

• candidate selection
• intra-party competition
• single-member districts
• party leadership
• steering agent

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 This holds true for investigations beyond the well-studied US case whose characteristics (competition among two parties that conduct open primaries) provide a somewhat different context. An insightful exception for the German case provides Reiser for the 2009 Bundestag election (Reiser, 2013).

2 While 299 MPs are elected in SMDs (first ballot), at least a further 299 seats are filled by candidates from one of sixteen state-level party lists (second ballot). Importantly, seats in the Bundestag are allocated according to the second ballot by proportional representation. The generally higher number of party-list MPs is the result of surplus seats (which arise in the case that a party wins more direct seats than their nationwide party-list share actually allows) and so-called compensation seats (which are balancing these surplus seats for the other parties).

3 More relevant is the case that rivalry only happens at the preselection stage. While it is possible that aspirants come forward for the first time at the nomination conference, this only plays a role in exceptional situations in Germany.

4 The research included all current Bundestag parties. Reports documenting the research design and data collection are available online: https://www.iparl.de/de/methodik-buka2017.html

5 At each conference, at least three interviews were conducted.

6 Some of the variables tested have also been used in Reiser’s (2013) bivariate analysis. While Reiser (2013, p. 144) points out that her findings are to be examined in multivariate investigations, we should be aware of some dissimilarities between her study and ours that make direct comparisons difficult. In particular, she deals with fewer parties (CDU, CSU and SPD in all of Germany, as well as the Left Party in eastern Germany). However, she also draws on a bigger sample than we do. Our small sample size actually is a caveat suggesting that further research is needed to validate the results.

7 Note that for the odds ratios smaller than 1, it is easier to interpret their reciprocal in terms of the likelihood of the opposite event to occur. 3.5 and 2.8. are reciprocals of the odds ratios of 0.282 and 0.352 and, thus, refer to the likelihood of non-competitive nominations.

8 Predicting probabilities is all the more advisable since interpreting odds ratios and assessing the predictors’ effect sizes are an intricate matter in logistic regressions (see Best & Wolf, 2010, p. 832).

9 Since extra-parliamentary parties have no ‘incumbents’, the item was not presented to their respondents.

10 However, the figures differ only slightly if we look at the whole group of members of the selection bodies (instead of the averaged share among the nomination conferences).

11 Such effects are not an unknown phenomenon in the political realm. Kingdon (1967), for instance, found that those candidates for the US Congress who won their electoral races develop complementary beliefs about their voters, while losers rationalise their defeat by saying that voters obviously vote according to party label but not according to their person.

12 To quantify those causes is difficult as a clear-cut classification into disjoint categories is not always possible. Broadly speaking, however, all three groups of reasons spread evenly across the investigated selections.

13 We see both first categories of reasons that lead to withdrawals as part of the selection process even if no other actors are directly involved. The aspirants’ resources and candidacy requirements provide supply-side explanations for the recruitment process (see Norris & Lovenduski, 1995, p. 389).

14 To be fair, the leadership in exceptional cases has to look imploringly for at least one aspirant running within organizationally weak parties (see Kannenberg & Hellmann, 2019).

15 Since our study is constrained by the sample size for our multivariate analysis, research drawing on larger samples could bring to light more nuanced findings. In particular, it could include party variables or further selection modes (like open primaries) as independent variables.

16 In their seminal article, Wuttke et al. (2019) found such effects as to participation in intra-party referendums.

17 Regarding the 2017 election, three telling examples are the selections of Jens Katzek (SPD), Sylvia Pantel (CDU, MP) and Ralph Lorenz (Liberals) who were nominated uncontested by a majority of 53, 68 and 66 per cent of votes.

Additional information

Funding

Data collection for this research was funded by the Stiftung Wissenschaft und Demokratie.

Notes on contributors

Danny Schindler

Danny Schindler is a postdoctoral researcher at the Institute for Parliamentary Research, Berlin.

Anastasia Pyschny

Anastasia Pyschny is a research associate at the Institute for Parliamentary Research, Berlin and a doctoral candidate at the Martin-Luther-University Halle-Wittenberg.

Malte Cordes

Malte Cordes is a research associate at the Institute for Parliamentary Research, Berlin and at the Martin-Luther-University Halle-Wittenberg.