![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
While negative campaigning has received increased attention, scholars have mostly focused on its effects. Studies looking at the determinants of negative campaigning remain sparse. Our article contributes to literature by developing a two-level model that takes into account the strategic choices of political actors and their characteristics as well as the context in which the negative strategy takes place. We apply our model to a rich data set of newspaper ads regarding direct democratic votes held in Switzerland. Our results show that negative campaigning, as measured by personal attacks, is more likely if political actors defend the status quo or are lagging behind in the polls, if the ad stems from a populist right party or entails no explicit endorsement, or if the ballot day draws near. Popular initiatives and more intense campaigns also generate a higher share of negative campaigning. Overall, then, a number of causal factors identified in (U.S.) elections also matter in Swiss direct democracy, which suggests that the reasons that make political actors willing to “go negative” are of broad relevance.
Notes
1 Political ads on TV and radio are forbidden in Switzerland. Therefore, political parties and interest groups extensively rely on newspaper ads in their campaign activities.
2 For instance, it has been shown that Conservative candidates are more likely to go negative than Democrats (Theilmann and Wilhite Citation1998; Lau and Pomper Citation2001), that candidates perceived as having attractive personal attributes are less willing to go negative (Harrington and Hess Citation1996), as are women (Kahn Citation1993; Herrnson and Lucas Citation2006) and candidates that belong to a cultural or racial minority (Krebs and Holian Citation2007).
3 This reasoning holds, of course, especially for zero-sum games where two competing sides are opposed (e.g., for majoritarian elections in two-party systems or, as in our case, for direct democratic ballots where voters have to choose between two competing alternatives). For a discussion on the strategic choices of political actors in non–zero-sum games (e.g., in multiparty systems), see Walter and van der Brug (Citation2013); Hansen and Pedersen (Citation2008); Elmelund-Præstekær (Citation2010); Walter and Nai (Citation2015).
4 See Brooks and Murov (Citation2012: 384–388) on how legislative changes (such as the 2002 Bipartisan Campaign Reform Act or the 2010 Supreme Court decision about “Citizens United v. Federal Election Commission”) paved the way for a strong increase in independently financed political ads, and their effects.
5 The 2012 U.S. Presidential election is a stunning example of those dynamics. As journalist A. Leary points out, “with others doing the work for him, [the candidate] can avoid the perception of dirty campaigning” (Leary Citation2012).
6 In addition, studies on election campaigns have shown that it is the party that relies more strongly on negative campaigning (Hardmeier Citation2003; Engeli and Tonka Citation2010).
7 National Lampoon's Animal House, by John Landis (1978).
8 The popular vote occurs either automatically for any change of the Swiss Constitution (“compulsory referendum”) or as a result of the “optional referendum,” provided a committee is able to collect 50,000 signatures against a federal law adopted by the Swiss parliament.
9 Our collection covers all the ads published in six major Swiss newspapers during the last 4 weeks before a ballot. Two newspapers for each linguistic region were selected: Tribune de Genève, Le Temps (French), Neue-Zürcher Zeitung, Tages-Anzeiger (German), Regione, and Giornale del Popolo (Italian). Those newspapers are fairly representative of Swiss press, both in terms of coverage and ideological orientations.
10 Other examples are (our translation): “Red Fundamentalists Want to Shut Down Our Power Plants” (Two popular initiatives on nuclear energy withdrawal voted on 5/18/2003; ad published in Tribune de Genève, April 24, 2003, p. 43); “That's a Scandal! The Federal Council Uses Citizens’ Money for Propaganda!” (Mandatory referendum on CH-EU bilateral agreements extension voted on 9/25/2005; ad published in Tribune de Genève, September 13, 2005, p. 7).
11 The number of days elapsed between the ad and the ballot day would provide a finer-grained measure. Such a measure is, however, problematic for political campaigns with very few ads, as it leads to excessive variability across time.
12 We are grateful to Gfs.bern for providing the data.
13 We also tested two alternative variables, one simply measuring the percentage of support for a given camp in the ballots (support in the polls) and the other subtracting the score of the other camp from the score in the polls for a given camp (net advantage in the polls). The results are similar to those presented here (see Table A2 in the appendix).
14 Note also that for some ballots there was only one pre-ballot survey (about 6 weeks before the ballot), whereas for others there were two pre-ballots surveys (6 and 4 weeks before the ballot, respectively). In some (rare) cases there were even three pre-ballot surveys (6, 4, and 2 weeks before the ballot). For ballots with more than one pre-ballot survey, our variable is based on the average results from the two (or three) surveys. We ran a series of additional robustness tests that showed that all variables have virtually identical effects for the whole set of projects and for the subsample of 24 projects. This suggests that what happens for the subsample of 24 projects (i.e., all models that include the variable measuring the competitive positioning of actors) should logically hold also for the whole sample of 75 projects.
15 A closer look at the data shows that the pro status quo camp is the one that most frequently loses in pre-ballot surveys.
16 See for instance the study conducted by the Pew Research Center's Project for Excellence in Journalism published in August 2012 on negativity during the 2012 US presidential campaign (available at www.journalism.org) or the particularly hostile tone between the two candidates at the 2012 French presidential election during the TV debate (Vandoorne et al. Citation2012).