Skip to content

Going negative in Germany – but why and with which effects?

Written by Jürgen Maier.

Most scholars agree that in many countries there is a high amount of negative messages in political elites’ campaign communication. Moreover, some experts even find a steady increase of negative campaigning. But why do candidates opt (more and more) for attacks instead of sending positive campaign messages? And do such negative campaign strategies really work?

Although there are a vast number of studies on negative campaigning, both questions have not been fully answered yet. On the one hand, there is a serious lack of research on negativity outside the United States. Because society, politics, and the media in the U.S. are very different from European countries, it might be inappropriate to simply transfer our knowledge about negativity from one culture to another without prior verification. On the other hand, research on negative campaigning is basically research on campaign advertising. No matter if we look at the United States or at other countries, there is remarkably little evidence on the use and the impact of attacks in other campaign messages.

Three recent studies (de Nooy & Maier 2015; Maier & Jansen 2015a, 2015b) expand our knowledge in several ways. First, the studies refer to the question why German candidates use negative messages and what their impact is. Second, all studies focus on televised debates. Televised debates are not just another campaign message. Televised debates are special. In general, televised debates can be considered as the most important single campaign event. For instance, 17.6 million citizens watched the 2013 German debate between Angela Merkel and her challenger, Peer Steinbrück (market share: 50.7%). No other political message had an outreach anywhere close to these numbers.

Although Germany is not considered as a country with a high level of mudslinging (for instance, political advertising rarely contain attacks), attacking the political opponent seems remarkably attractive to candidates participating in televised debates. In average, one quarter of all candidate statements in televised debates held on the national or the state level are negative (by comparison, the share of attacks in U.S. debates is 31 percent; see Benoit 2014, p. 38). But for some candidates attacking the political opponent seems to be more appealing than for others. This raises the question what factors drive candidates to go negative?

Based on all televised debates ever broadcasted in Germany – 46 debates in total between 1997 and 2015 – our results indicate that only a handful of variables are responsible for negativity. First, the most important factor is if a candidate is a member of the government. Candidates who are part of the government are less likely to use attacks than candidates who are not. The difference is about 13 percentage points. Second, candidates tend to be more aggressive if there is more at stake. Hence, the average level of attacks is almost 9 percentage points higher in national than in state elections. Third, the strategic context matters. Candidates who are behind in the polls attack more than candidates who are ahead in the polls. For instance, a deficit of ten percentage points in the polls accounts for an increase of negativity of about two percentage points. In sum, negativity will be maximal in televised debates if it is about a national election, if none of the candidates is a member of the government (this is a rare situation in European elections but a constellation occurring in the U.S.), and if the debate might be the last chance for one of the candidates to catch up with the frontrunner. As factors like the timing of the debate, the presence of an audience, or the number of contenders are not able to explain candidates’ different use of attacks, it is important to note that there is no mechanism to prevent or to foster conflicts in a debate.

But what is the impact of such attacks? Do they pay off for the messenger? Based on survey and real-time response data we do not find much evidence that negative debate messages considerably affect the evaluation of candidates. Most of the estimated effects on candidate support turn out to be statistically insignificant. Among the significant effects, there is neither clear evidence that negativity has an effect intended by the attacker (i.e. increase support for the attacker respectively decrease support for the candidate in target) nor that attacks cause backlash effects (i.e. decrease support for the attacker or increase support for the candidate in target). Both types of effects appear, but there is no distinct pattern related to a particular candidate, a certain political party, or a candidate’s specific political position. The most consistent effect we have found is that attacks tend to polarize the electorate: Negative messages can increase support in the own camp. Simultaneously, the aversion toward the attacker will grow in the camp of the political opponent.

Hence, our findings from German televised debates can be summarized as follows: Undoubtedly, negativity is important for candidates. In particular constellations, attacking the political opponent even seems indispensable for candidates to achieve electoral success. Unfortunately, the effects of negativity are not as helpful as expected. Actually, most of the impact dies away without any consequences. If negative messages do have an influence, it is not yet predictable if the effect will be favorable or unfavorable for the attacker or if the result will merely be a polarization of the electorate.

One reason for this mixed picture is that capturing the use and effects of negativity in televised debates is an extremely ambitious task. This is true because televised debates are rather long and complex compared to other campaign messages. Our results indicate that the impact of attacks is more complicated than the often rather simple dicta of campaign managers suggest. To fully understand the mechanisms underlying attacks, more detailed and methodologically sophisticated analyses are necessary. These analyses have to acknowledge that attacks can differ with respect to, e.g., focus or civility. In addition, as the processing of negative information differs between subjects we need more refined analyses on the recipients’ side. As attacks are relevant for political actors and voters are directly confronted with those messages (at least if they watch a televised debate), it is, of course, it certainly makes sense to increase the efforts to uncover the conditions for the use and the effects of negative campaigning.

Jürgen Maier researches political communciation and sociology at the University of Koblenz-Landau in Germany. This article is part of the Center for British Politics (CPB) research series into negative campaigning. Image credit: Wikipedia Commons.

References 

Benoit, William L. (2014). Political election debates. Informing voters about policy and character. Lanham, MD: Lexington Books.

de Nooy, Wouter & Maier, Jürgen (2015). When do attacks work? Moderated effects on voters’ candidate evaluation in a televised debate. In Alessandro Nai & Annemarie S. Walter (eds.), New perspectives on negative campaigning. Why attack politics matters. Colchester: ECPR Press (in print).

Maier, Jürgen & Jansen, Carolin (2015a). When do candidates attack in election campaigns? Exploring the determinants of negative candidate messages in German televised debates. Party Politics (ahead of print). doi:10.1177/1354068815610966

Maier, Jürgen & Jansen, Carolin (2015b). Zerstört „negative campaigning“ in TV-Duellen das Vertrauen in die Kanzlerkandidaten? Eine Analyse für die Fernsehdebatten im Vorfeld der Bundestagswahlen 2002, 2009 und 2013  [Does negative campaigning in televised debates destroys trust in chancellor candidates? An analysis for the 2002, 2009, and 2013 German televised debates]. Politische Pychologie (in print).

Published inEuropean PoliticsInternational PoliticsNegative Campaigning

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *