Can't buy me votes? Campaign spending and the outcome of direct democratic votes

Abstract

Recent American studies that investigate the influence of money on the outcome of direct democratic votes find campaign spending effective for both supporters and challengers. Taking advantage of Switzerland’s far-reaching experience with direct legislation, this article tests the role of money in a ‘harder’ context, where government plays an active role in direct democratic processes. It assesses the impact of campaign spending on 323 federal votes, and on a subset of 60 votes, offering finer-grained cantonal-level data. On the one hand, the results highlight the crucial role of the partisan coalition supporting the government for the outcome of votes; on the other hand, campaign spending is effective for both government and challenger camps, yet more so for the latter. Furthermore, the effects of government camp spending vary depending on the expected vote outcome and the difference in spending between both camps.

Keywords:

• Campaign spending
• direct democracy
• Switzerland
• government coalition
• initiatives
• referendums

Acknowledgements

Earlier versions of this article were presented at the Annual Conferences of the Swiss Political Science Association in St Gallen in January 2017 and in Zürich in February 2019. Many thanks to the participants, as well as to Simon Hug, Adrien Petitpas and Eva Fernandez Guzman Grassi.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 The study was based on simple correlations between the outcome (yes or no) and a weak measure of campaign spending (number of newspaper ads in two regional dailies and one free paper during the last week prior to the polls), and this for only 40 direct democratic votes from 1977 to 1981.

2 The final vote in the National Council, the lower Chamber of the Swiss parliament, could serve as an alternative measure of government support among the elite (Sciarini and Treschel 1996; Trechsel and Sciarini 1998. That measure is strongly correlated with the size of the partisan coalition supporting the government according to vote recommendation, especially for referendums (Pearson’s R = 0.69, N = 180), yet the latter is a more valid measure of the vote outcome according to previous research (Kriesi 2009b).

3 The initial dataset included 328 direct democratic votes, but we must exclude five votes because no newspaper analysis was undertaken. The data on direct democratic votes stems from Swissvotes (2020).

4 For example, during the referendum campaign on the extension and renewal of the agreement on the free movement of persons with the EU in 2009, total spending on newspaper ads and street posters amounted to 11 m CHF, which means 2.20 CHF per registered voter; this is less than the 2008 campaign on the introduction of same-sex marriage in California (84.4 m CHF and 4.88 CHF per registered voter), but far more than the 2016 Brexit campaign (40.3 m CHF, but ‘only’ 0.87 CHF per registered voter), and also more than the 2000 campaign on the euro in Denmark (4.7 m CHF, 1.76 CHF per registered voter).

5 Two tabloid newspapers (Blick for the German-speaking region and Le Matin for the French-speaking region) and two quality newspapers in each linguistic region (Tages Anzeiger, Neue Zürcher Zeitung, La Tribune de Genève, Le Temps – formerly Le Journal de Genève). We thank Hanspeter Kriesi, Thomas Milic, Céline Colombo, Alessandro Nai and the Année politique suisse for making the data available to us.

6 Art. 10 of the Federal Act on Radio and Television.

7 These surveys also show that the booklet presenting the ballot measure and the voting recommendation of the Swiss government and parliament, which is sent to all voters, is also used by an overwhelming majority of them (70%) as a source of information (Sciarini and Tresch 2014). This underlines the importance of the role played by the Swiss government’s cues for opinion formation.

8 Thanks to the organization Research and Studies of Advertising Media (WEMF) for data.

9 We thank Lucas Leemann and Oliver Strijbis for sharing this data with us. The recent trend towards using online tools (online ads, social media) as additional campaign means should not affect the validity of newspaper ads as an indicator of campaign spending, since these tools are still weakly developed in Switzerland.

10 There are only two exceptions in our data, i.e. two initiatives for which the government recommended a Yes vote. We coded them accordingly.

11 Note that the influence of interest groups on Swiss direct democratic votes is difficult to measure, because they typically align with the political parties of their respective camps (Kriesi 2006). In many cases, interest groups’ spending in support of (or against) a given proposal remains unnoticed, since they take place indirectly, with interest groups subsidizing the spending of their allied parties.

12 The very fact that an optional referendum is used demonstrates that consensus-building efforts made during the pre-parliamentary and parliamentary phases were not completely successful.

13 Estimation based on the correlation between the number of ads and the estimation of the campaign money (MediaFocus data).

14 As Table A3 in the online appendix shows, the results are almost the same if we estimate the models in a frequentist framework.

15 Figure A2 (based on Table A4) also provides a more detailed view of the effect of spending imbalance.

16 To give an idea of the rather exceptional character of such a configuration, in our data the situation where the challenger camp spends more than 25,000 CHF (mean + one standard deviation) and outspends the government camp by more than 10,000 CHF corresponds to only 8% of empirical cases.

Additional information

Funding

This work was supported by the Swiss National Science Foundation (grant 100018_159370).

Notes on contributors

Julien M. Jaquet

Julien M. Jaquet is a doctoral candidate in the Department of Political Science and International Relations, University of Geneva. His main research interests include political decision-making processes, legislative studies, political participation and data science. His work has appeared in European Journal of Political Research and Electoral Studies. [[email protected]].

Pascal Sciarini

Pascal Sciarini is Professor of Swiss and Comparative Politics at the University of Geneva. His main research topics are decision-making processes, direct democracy, Europeanization and political behaviour (participation and vote choice in elections and direct democratic votes). His work has appeared in Comparative Political Studies, Electoral Studies, European Journal of Political Research, European Union Politics, Journal of European Public Policy, Journal of Politics and West European Politics, among others. [[email protected]].

Roy Gava

Roy Gava is Assistant Professor in Business and Politics at the School of Economics and Political Science, University of St Gallen. His research interests include policy-making, interest groups, financial regulation and data science. His work has appeared in Political Science Research and Methods, Interest Groups & Advocacy, International Review of Administrative Sciences, and Journal of Legislative Studies, among others. [[email protected]].