![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
As membership levels decline, parties are developing new forms of linkage with supporters, many of which rely on Internet technologies. To date, the discussion of these new modes of affiliation has been largely theoretical in nature, with little, if any, systematic empirical analysis undertaken on their appeal and impact on formal membership. This article seeks to fill this gap by examining the presence of three new types of digital affiliation—audience, friends, and digital activists—among the French electorate using original survey data from the 2012 Presidential election. Our findings are important in showing that while the new methods of affiliation are increasing parties’ reach into society, they are not necessarily widening parties’ socioeconomic support base. Furthermore, digital activism is mostly a supplementary channel for members’ input although there are a smaller group of people engaging in these activities that avoid formal membership ties. Such results suggest that digital methods of affiliation might offer an important new resource to parties during campaigns.
Funding
This research was supported by an Economic and Social Research Council Grant, “The Internet, Electoral Politics and Citizen Participation in Global Perspective’ RES-051-27-0299” and funding from the French Agency for Research and the Québec Society and Culture Research Fund “webinpolitics.com” ANR-11-FRQ-003-01. The authors would like to thank colleagues T. Giasson, F. Bastien, G. Blanchard, S. Gadras, M. Lalancette, and S. Wojcik for their contribution to the latter project.
Notes
1. Recent Eurobarometer 80 reports show France at half (7%) of European Union (EU) average of 14% saying they trust parties (Autumn 2013). Data available at http://ec.europa.eu/public_opinion/archives/eb/eb80/eb80_anx_en.pdf.
2. The French National Election Study 2012. The authors thank Nicolas Sauger for making the data available.
3. A proportion of the sample (73.5%) did not have access to the Internet, which meant they could not engage in online activities. This proportion is consistent with, albeit slightly lower than, that reported by other sources. According to the World Bank, 81.4% of the population had accessed the Internet during the previous 12 months in France in 2012 (see http://data.worldbank.org/indicator/IT.NET.USER.P2). The figures from the French National Election Study may be slightly lower given that the sample is restricted to those 18 years old and older and includes French citizens registered to vote.
4. May 6 to May 27, 2012.
5. Full details of the selection process and URLs used are available from the authors upon request.
6. See Appendix A for full question wording.
7. For example, Hollande’s TousHollande, Nicolas Sarkozy’s La France Forte, marinelepen2012.fr, www.bayrou.fr, as well as Jean-Luc Mélenchon’s website, placeaupeuple.fr. These candidates and their parties also had Facebook pages and Twitter accounts.
8. See Appendix B for variable coding.
9. This decision was based on the consideration that our main focus is on how citizens engaged with the online campaign during the 2012 Presidential election. Also, from a methodological perspective inclusion of non-Internet users would mean they were coded as zero for the purposes of analysis and thus treated incorrectly as equivalent to those with access but who did not participate.
10. More than three-quarters of respondents (75.8%) said they were “very interested” in politics and a majority (57.4%) were highly educated, holding a postgraduate degree. There was also a gender bias in that three-fifths (61.9%) were male. In terms of ideological outlook the distribution tended to overrepresent leftist parties as compared with support in the wider French electorate.
11. The view that online political activists are more left-wing in France is in keeping with the extant literature (Gadras & Greffet, Citation2013). It is also possible that the distribution reflects our use of snowball sampling techniques and a greater efficiency by leftist parties in recruiting respondents. The Greens and the Socialists both held party primaries in the preceding year to choose their Presidential nominee, which may have given them an advantage in developing and nurturing communication networks with their supporters. In addition, the research team were aware that Green party officials, including the head of the e-campaign team, were circulating the survey link on Twitter to encourage participation. Also, anecdotal evidence from discussions on radical-right forums and blogs where the survey link was mentioned suggested a stronger reluctance and even suspicion toward participation in the project than was seen among left-wing equivalents. That said, we can see that when the sample is restricted to members only we do see a reversal of partisan dominance in that right-wing parties are actually better represented than their left counterparts. Appendix C presents the full breakdown of vote choice among the overall activist sample, members only, and compares this to the national vote choice of the general electorate. Finally, to try to control for any bias that party recruitment efforts may have introduced to the analysis, we introduce a control for party contact in our multivariate analysis (see ).
12. The objective of this non-parametric technique is to produce estimates that are based on the distribution function of the observed data. This is done through a process of resampling observations (with replacement) from this data. In this way, standard errors and p-values are not computed based on the assumptions of a known distribution (e.g., the multinomial distribution) but on the distribution of the sample itself.
13. One exception may be among the party friends where a negative coefficient for gender is significant, meaning that those engaging in this mode of party support are more likely to be male than the average member. On first view this does appear to run counter to the earlier findings, which revealed that it was audience members (not friends) that tended to be more male. However, this was a feature of audience in comparison to the population as a whole. Here we are comparing against a different reference group—online party members. Nevertheless, the fact that friends now overtake audience as the more male-dominated group of digital affiliates is somewhat surprising and suggests that despite the comparability of indicators we have for the core activities of each of the groups across the two surveys, our differing methods of recruitment (online and offline) may have introduced some socio-structural differences in the composition of the three groups, which reduces their direct comparability.
14. The index of party and campaign engagement online included three campaign actions and one party action: watching a video of a public meeting with a candidate; sending an e-mail to a candidate; downloading campaign material; and donating money to a party online (over the past 12 months). The index of party and campaign engagement offline is similarly composed of three campaign actions and one party action: attending a public meeting of a candidate; delivering leaflets and hanging posters; other campaign actions; and donating money to a party offline (over the past 12 months). The index of civic engagement online includes the following: signing an e-petition; buying products online for political, ethical, or environmental reasons; and volunteering online for a nongovernmental organization (NGO) or a charity (all over the past 12 months). The index of civic engagement offline includes the following: signing a paper petition; buying products (offline) for political, ethical, or environmental reasons; and volunteering (offline) for an NGO or a charity all over the past 12 months).
15. ANOVA results comparing differences of means show that all categories were significantly different (p > 0.05) in the online party and campaign engagement except between audience members and friend non-members; and between friend members and digital activist non-members. In the index of offline party and campaign engagement the only differences that were not significant (p > 0.05) were between audience members and friend members; audience members and digital activist non-members and friend non-members. For online civic engagement, differences were mostly insignificant except for digital activist members against all other categories other than friend members, and between audience non-members and friend members. For offline civic engagement, differences were also mostly non-significant with the exception of friends non-members’ scores, which were significantly lower than those of friend members and digital activist members.
16. Source: Ministere de l’Interieur, http://www.interieur.gouv.fr/Elections/Les-resultats.
17. The results of two right-wing candidates (N. Sarkozy: 27, 2%, and N. Dupont-Aignan: 1.8%) are added to each other.
Additional information
Funding
Notes on contributors
Rachel Gibson
Rachel Gibson is Professor of Politics, University of Manchester.
Fabienne Greffet
Fabienne Greffet is Lecturer in Political Science, University of Lorraine.
Marta Cantijoch
Marta Cantijoch is Lecturer in Politics, University of Manchester.