Pre-electoral coalitions, familiarity, and delays in government formation

Abstract

During the past decade, many parliamentary democracies have experienced bargaining delays when forming governments. The previous literature has attributed protracted government formation processes to a high degree of preference uncertainty among the political parties and a high level of bargaining complexity. The article draws on such theories, but also adds a third theoretical mechanism, commitment problems, and highlights two explanatory variables that have not received much attention so far. The first is pre-electoral coalitions, which are declarations by parties stating that they intend to collaborate with each other after the election. The second is familiarity, which is the mutual trust between parties that comes from having worked together in the past. By combining a large-N study of government formation processes in 17 West European parliamentary democracies (1945–2019) with an in-depth case study of the prolonged Swedish government formation process in 2018–2019, it is shown that pre-electoral coalitions that fail to win a majority can sometimes delay, not speed up, government formation. In addition, a lack of familiarity may sometimes lead to a breakdown of negotiations and drawn-out government formation processes.

Keywords:

• Government formation
• bargaining delays
• parliamentary democracies
• Sweden

Acknowledgements

The authors would like to thank the anonymous reviewers, as well as the comments and suggestions made by the participants at the ‘Coalition Dynamics: Advances in the Study of the Coalition Life-Cycle’ conference. In addition, we wish to thank Riksbankens jubileumsfond for funding the original project. Jan Teorell wishes to acknowledge sabbatical support from the Wenner-Gren Foundation. Bäck and Hellström would like to thank Vetenskapsrådet (2020-01396) for financial support.

Disclosure statement

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

Notes

1 This article draws on evidence compiled for a Swedish-language book (Teorell et al. 2020) but also on new and updated evidence that we have compiled for the article.

2 The following (West European) countries are included in this study: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden and the United Kingdom. The data cover government formation opportunities in these countries over the period 1945 (or the beginning of the present democratic regime) to 2019.

3 Diermeier and Van Roozendaal (1998) include a measure of ‘identifiability’—the availability of pre-electoral government options—in their empirical models, but only as a control variable. Similarly, in a more recent study, Ecker and Meyer (2020) include a control for pre-election coalitions. They find that such coalitions speed up the government formation process significantly.

4 More precisely, Martin and Stevenson (2010) specifies the weight as two times the product for the two parties’ individual portfolio shares unless M=K, when the weight is that party’s squared portfolio share. This ensures that the weight for all party pairs in a given coalition sum to 1.

5 In this case, our approach differs slightly from Martin and Stevenson (2010, 509–510), who instead count the ‘per centage of days …that the two parties have participated in the same cabinet up until that point,’ discounted by an unspecified discount rate ‘that is sufficiently high to ensure that periods of cabinet partnership occurring more than approximately eight years before … are almost completely discounted.’ We try to mimic this approach by using the month as the time unit and a 5 per cent discount rate. This ensures that after 8 years, that is, 96 months, only 0.95⁹⁶=.007≈0 of a unit shock remains.

6 In the online appendix we show the results from alternative specifications of depreciation rates and ‘weights’ for the familiarity measurement.

7 As former communist parties were increasingly seen as potential coalition partners after the collapse the Soviet Union and the end of the Cold war, we do not count former communist parties that ceased to be communist in the 1990s, or other moderate left parties, as extreme.

8 Bargaining weights for individual parties are measured as the share of cabinets where a party is necessary to produce a majority-winning cabinet of all ‘potential coalitions’ or ‘potential governments’. The latter term refers to the combination of parties, or an individual party, that hypothetically could form a government in some bargaining situation. The normalised Banzhaf’s index goes from 0 to 1, where 1 indicates that there is a single-party majority winner after a parliamentary election.

9 The negative binomial model is preferred over the Poisson model since there are clear signs of over-dispersion in our data.

10 As shown in the online appendix, using Cox proportional hazards models gives results that are nearly analogous to those reported in the article.

11 Since we include country-fixed effects, we are not able to estimate the effects of political institutions such as semi-presidentialism that do not vary or vary very little over time.

12 When we analyse sub-samples of post-election and inter-election government formations or the conditional effect of post-election situations on our main independent variables, the results are fairly similar to the ones presented here.

13 The predictions are based on Model 4 to increase power (as the sample size is somewhat larger), but as seen in Table 2, Model 2 would produce similar predictions.

14 The predictions from the different models in Table 2 are very similar.

15 A government does not automatically resign after an election, but unless the prime minister steps down voluntarily, a compulsory vote of no confidence is held once the new parliament, the Riksdag, is summoned. If the prime minister fails in this vote, the Speaker nominates a new prime minister after consulting all Riksdag parties. The nominee must pass an investiture vote with a negative decision rule. After four unsuccessful investiture votes, parliament is dissolved and new elections are called; this has never happened so far (see, for example, Wockelberg 2015).

Additional information

Notes on contributors

Hanna Bäck

Hanna Bäck is a Professor of Political Science at Lund University. Her research mainly focuses on political parties, legislators, and governments in parliamentary democracies. Her articles have for example appeared in British Journal of Political Science, Comparative Political Studies, and European Journal of Political Research. [[email protected]]

Johan Hellström

Johan Hellström is a Docent of Political Science at Umeå University. His research focuses on coalition politics, political parties and democratic representation in Parliamentary democracies. He has published books and articles on mainly comparative European politics and is responsible for the (European) Representative Democracy Data Archive—a data research infrastructure for coalition research [[email protected]]

Johannes Lindvall

Johannes Lindvall is the August Röhss Professor of Political Science at the University of Gothenburg. His most recent book is Inward Conquest: The Political Origins of Modern Public Services (Cambridge University Press 2021, with Ben Ansell). He is also the author of Reform Capacity (Oxford University Press 2017), Mass Unemployment and the State (Oxford University Press 2010), and numerous articles in scholarly journals. [[email protected]]

Jan Teorell

Jan Teorell is Lars Johan Hierta Professorial Chair in Political Science at Stockholm University. His research interests include comparative politics, comparative democratisation, state making and party politics. His work has appeared in journals such as the American Political Science Review, Comparative Political Studies, and European Journal of Political Research. [[email protected]]