Electoral Systems and Party Systems in Germany on the Local Level

ABSTRACT

The question of how electoral systems should be designed has already been broadly discussed. One of the most promising electoral system types are mixed-member proportional (MMP) systems. However, the positive evaluation of MMP systems strongly depends on the most prominent case applying such a system, Germany. In order to better understand the performance of MMP systems, we focus on this case and research to what extend German party systems are affected by electoral system designs. For this purpose, an examination of the local level is most fruitful. This level provides important variation in electoral systems in terms of two important factors – the general system type (MMP or proportional representation) and the prevalence or absence of a legal threshold. Our results show that (effective) thresholds affect the proportionality of election results whereas the general system type does not. The fragmentation of German local party systems is largely independent of both variables.

Acknowledgements

Former versions of this article have been presented at the Annual Meeting of the DVPW Standing Group “Handlungs- und Entscheidungstheorie” and at the 2020 conference of the German Studies Association. We thank all panelists for valuable comments, in particular Susumu Shikano and Dominic Nyhuis. Further, we are grateful to two anonymous Reviewers for very helpful suggestions.

Disclosure Statement

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

Notes

1 MMP electoral systems use two tiers: one in which candidates are elected by plurality (or majority) rule in districts with a magnitude of one, and another in which voters vote for party lists according to a PR rule. Notably, both tiers are not independent of each other, but the PR tier (partially) compensates disproportionalities of the plurality tier.

2 Disproportionality was operationalized with help of the least squares index (LSI, see Gallagher 1991) which bases on differences between vote shares v_p and seat shares s_p of parties p: LSI = $\sqrt{\sum _p{\left(s_p-v_p\right)}^2/2}$. Party system fragmentation was measured by the effective number of parties (ENP, see Laakso and Taagepera 1979): ENP = $1/\left(\sum _p{s}_p^2\right)$.

3 Certainly, there is more variation in detail that is not shown in Table 2. Examples are differing voting ages and election periods. Such details, however, are not of interest for this study’s purpose.

4 Reiser, Rademacher, and Jaeck (2008) found similar patterns for slightly older data. According to their study, ILLs ran in 73 per cent of municipal elections and reached an overall vote share of roughly 35 per cent (Reiser, Rademacher, and Jaeck 2008, 123).

5 We provide some illustrative examples in Appendix 3.

6 Results base on the most recent elections as to April 2019.

7 We are grateful for one comment of our Reviewers that this might by because of the special role of Bezirke and/or the size of the two cities. With our data, we cannot confirm or reject this assumption, but independent of what causes the difference between Hamburg/Berlin and Saarland, it clearly seems to overlay the expected threshold effect.

8 For the analyses shown in Tables 4–6, we again computed the D values of the single states. The edge cases or special cases mentioned above do not have any noticeably high influence here either.

9 If we broaden our view to a more comprehensive set of evaluation criteria, however, some negative side effects of MMP systems compared to PR systems need to be addressed, such as lower levels of comprehensibility and risks to generate oversized parliaments (cf. Weinmann and Grotz 2020, 2021).

Additional information

Notes on contributors

Eric Linhart

Eric Linhart is a Professor of Political Science at the Chemnitz University of Technology (Germany). His research interests include elections, electoral systems, political parties and coalitions. His publications have appeared, amongst others, in the European Political Science Review, the Journal of Elections, Public Opinion & Parties, Party Politics and Political Studies.

Kristin Eichhorn

Kristin Eichhorn is a PhD candidate at the Chemnitz University of Technology (Germany). Her main research interests are elections, autocracies and political systems in Eastern Europe. Her work has been published in, amongst others, Political Studies and Politikon.