ABSTRACT
Mobile instant messaging services (MIMS) are emerging as important digital environments in citizens’ everyday lives. We explore the use of MIMS for talking about politics with unique survey data on samples representative of Internet users in Germany, Italy, and the UK. First, we show that robust percentages of our respondents who use MIMS employ them for posting political messages and discussing politics. Second, we demonstrate that political talk on MIMS is positively associated with users’ tendency to censor themselves politically on social networking sites (SNS) and, to a lesser extent, with ideological extremism. Third, we find that the association between self-censorship on SNS and the likelihood of publishing political contents on MIMS is stronger for individuals living in former East Germany where, due to historical reasons, large segments of the population are reluctant to talk about politics in public. Our findings suggest that MIMS make a distinctive contribution to contemporary repertoires of political talk, with important implications for the quality and inclusiveness of interpersonal political discussion.
Acknowledgements
We thank Yannis Theocharis for his valuable comments on a draft version of this paper which was presented at a meeting of the Social Media and Political Participation Global Network in New York City. In accordance with Italian academic conventions, we specify that Augusto Valeriani wrote the sections titled “Literature Review and Hypotheses”, “Data, Variables, and Models”, and “Conclusion”; Cristian Vaccari wrote the sections titled “Introduction” and “Findings”; both authors collaborated in the design of the study.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Augusto Valeriani is an Assistant Professor in Media Sociology at the University of Bologna. His main research interests are digital media, political communication, and political participation, both in national and international contexts. He tweets as @barbapreta [email: [email protected]].
Cristian Vaccari is a Reader in Politics at Royal Holloway, University of London, and an Associate Professor in Political Science at the University of Bologna. He studies political communication in comparative perspective and is the Principal Investigator of a comparative research project on social media and political participation in Western democracies (http://www.webpoleu.net/). His latest book is Digital Politics in Western Democracies: A Comparative Study (Johns Hopkins University Press, 2013). He tweets as @25lettori [email: [email protected]].
Notes
1 Snapchat is technically a MIMS since it allows users to share videos and chat with their contacts via mobile devices. However, it can be argued that some of its affordances make it similar to a social networking platform: ‘Stories’ can be publicly shared and the ‘Discover’ section contains ‘dedicated channels’ curated by major publishers and news organizations. However, even fragments of public stories can be commented only privately and nobody is allowed to know who is connected to a user. These features make the service closer to MIMS than to SNS, at least for the purposes of the present study.
2 Retrieved May 18, 2016, from http://www.wired.com/2016/04/forget-apple-vs-fbi-whatsapp-just-switched-encryption-billion-people/.
3 Some services can also be accessed via desktop or laptop computers, but only through relatively cumbersome procedures, that is, employing emulators of mobile operating systems or ‒ only for WhatsApp ‒ syncing a mobile device with a computer. Facebook Messenger is an exception, since it is by default synced with the private messaging function of Facebook, but the user interfaces for the mobile and computer versions differ substantially.
4 Retrieved December 8, 2016, from https://www.facebook.com/communitystandards.
5 Retrieved May 22, 2016, from http://wearesocial.com/special-reports/digital-in-2016.
6 See Note 5.
7 NUTS stands for ‘Nomenclature of Territorial Units for Statistics’ and is a geographical classification that subdivides territories of the European Union into regions at three different levels. Retrieved June 2016, from http://ec.europa.eu/eurostat/web/nuts/overview.
8 A total of 9021 invitations were sent in Germany, 9000 in Italy, and 9500 in the UK. In Germany, 6660 recipients did not open the survey link, 247 abandoned the interview, and 364 were screened out or turned out to be over quota. In Italy, 6493 recipients did not open the survey link, 304 abandoned the interview, and 453 were screened out or turned out to be over quota. In the UK, 7163 recipients did not open the survey link, 260 abandoned the interview, and 327 were screened out or turned out to be over quota.
9 We classified respondents residing in Berlin as residents in former East Germany. Given the peculiar history of Berlin, we also ran our models excluding Berlin residents and the results were consistent with those discussed in this article. (Data available upon request.) It might be argued that the internal migration that occurred in reunified Germany since 1989 makes it problematic to employ respondents’ current region of residence to test our hypothesis. However, most of this migration ran from East to West and occurred in the two years after reunification (Hunt, Citation2006). Therefore, our sample likely includes more West German residents formerly living in East Germany than East German residents formerly living in West Germany. To the extent that such distortion exists, it should bias our findings against rather than in favor of H3.
10 We ran the same analyses with data including all respondents, as well as all respondents who used MIMS regardless of whether they used social media. In these models, we recoded as ‘0’ those respondents who were not asked a question about political uses of SNS or MIMS because they had not reported using either or both platforms. The models based on all respondents using MIMS, irrespective of whether they used SNS or not, yielded substantively identical findings to those reported here for all three hypotheses. The models based on all respondents, irrespective of whether or not they used both MIMS and SNS, yielded substantive identical findings for H1 and H3. With respect to H2, the coefficients for ideological extremism were in the same direction and had similar strengths as those reported here, but fell slightly below the conventional significance thresholds (B = 0.092, p = .074) when posting political messages on MIMS is the dependent variable. (Data available upon request.)
11 Our questions referred to the six months preceding the interview, while the British general election was held more than six months before our surveys were in the field.
12 Our battery on political activities on SNS did not feature a question identical to the one measuring political posting on MIMS. Here, we refer to answers to the item ‘ … published political news or comments’ (on SNS). This comparison is acceptable since political news and comments are indeed messages about politics and public affairs.