Version classiqueVersion mobile
OpenEdition Books

Temps et temporalités du Web

Valérie Schafer

Rythmes et tensions

Exploring Real-time Voter Targeting Strategies on Social Media

Lena A. Hübner


This chapter theoretically discusses the role of « real-time » targeting strategies on social media in US politics. By doing so, it intents to clarify the interplay of three perspectives – that of the industry, the party and the users.


Social media, politics, time, voter targeting, US elections

Texte intégral

  • 1 Translation from French by author: « This should have been the triumph of “big data", usin (...)

Political parties all over the world massively invested in web communications, community management, and big data analytics since Obama’s presidential campaigns in 2008 and 2012 (Bor, 2014). During the 2016 presidential campaign, Hillary Clinton continued to pursue this technology and data-driven strategy, building on more than eight years of collected citizen data. Her tech team used the most advanced and sophisticated tools. Most pollsters predicted her victory. As Laurent (2016: § 1) puts it: « Ce devait être le triomphe du « big data », selon l’expression (impropre) consacrée aux méthodes modernes de ciblage des électeurs1 ». Yet, the electoral results say the opposite.

1On the other end, Donald Trump, during the early election campaign, declared that data collection is highly overrated (Bykowicz, 2016). However, he changed his mind with the election approaching, « spending millions of dollars on data and digital services in an effort to land donations and win over votes » (ibid.: § 1) and collaborating with several private firms such as Cambridge Analytica (AFP, 2016). In the aftermath of the election, « commentators tell us that Trump was successful because the Internet has brought about a « post fact » […] era and point to « filter bubbles » as a significant factor in his rise » (Kreiss, 2016: § 1). One thing is certain: the use of big data techniques allows more and more information to be spread at an accelerated pace through numerous online platforms, regardless of whether it’s a hoax, a news story or a campaign ad.

2This chapter focuses on one particular type of accelerated content circulation on social media: political party’s content used for online mobilization campaigns in the US. More precisely, we intend to critically examine the role of real-time micro-targeting strategies in election campaigns. To do so, we start by detailing the research problem. We then take a closer look at the notions of real-time (Weltevrede, Helmond, and Gerlitz, 2014; Fuchs, 2013a) and the acceleration of politics (Rosa, 2005) in western liberal democracies (Macpherson, 1968). In the discussion that follows, we explore the part played by real-time-targeting strategies in the democratic process.

3Before beginning with the text, we would like to clarify our epistemological standpoint. We share the opinion of Daniel Kreiss (2016) that social media and data-mining strategies « did not give us Donald Trump », as this would constitute a commitment to technological determinism. While social media might increase the speed of news and hoaxes, we cannot blame social media for creating « the cynical public that does not understand, or care to, how knowledge producing institutions work » (ibid., § 5). In addition, by accusing social networks like Facebook to be a failure for civic information, one is obliterating other less ideological motivations such as the company’s pure economic incentives (Lewis and Carlson, 2016).

Context and research problem: Going beyond descriptions of real-time targeting strategies

4Farrar-Myers and Vaughn (2015: 2) point out that « we have reached the point where having a social media presence […] is a necessary means to be an effective political communicator. » They also remind us that a sole presence is not enough: in order to successfully communicate online, audiences have to be targeted, user generated content has to be moderated and visibility in an information overload environment has to be assured (ibid.). To do so, costly data-mining techniques supplied by private corporations play a crucial role.

5While real-time consumer-targeting strategies in the cultural domain have been analyzed extensively from a socioeconomic viewpoint (e.g. Zwick and Denegri-Knott, 2009) political communication studies rarely examine the role of corporate online mobilization tools in election campaigns. They rather focus on user engagement, deliberation, voter turnout, content analysis and framing as well as on the technical characteristics of different distribution channels and interfaces (Bennett, 2014).

6On the other hand, literature on online surveillance techniques is burgeoning since Edward Snowden’s revelations (Shorey and Howard, 2016). Yet, these issues are rarely addressed in the context of election campaigns (Kruikemeier et al., 2016) even if campaigning teams and their consulting firms « continually [survey] the electorate and [look] for patterns within their massive databases » (Kreiss and Welch, 2015: 13). While there have been some journalists and bloggers who raised the problem (Issenberg, 2012; Gallagher, 2012), very few researchers have asked the « larger question about how data about voters is being mined and profiled evaluating the individual risks to privacy or the general implications for the democratic system » (Bennett, 2014: 3).

7In addition to the shortcomings regarding the role of corporations and the lack of literature about voter surveillance, the user’s perspective has rarely been considered when talking about real-time targeting. Scholars interested in such strategies often neglect the social contexts, personal needs, motivations and different levels of comprehension and awareness of the targeted, namely the people who interact with the messages constructed as a result of data-driven methods (ibid.).

8Besides, none of the studies mentioned above directly addresses time-related issues. Although authors talk about real-time targeting and the accelerated speed of information delivery, those issues are rarely the main focus of interest. Rather, they are characteristics describing different stages of web communication processes: 1) data can now be collected and stocked in real-time, 2) data-mining techniques then allow institutions to create constantly updated profiles of selected users, 3) using these profiles, they can instantly send tailored messages (or sell data to private communication firms to do so), 4) users can subsequently react to messages in real-time, 5) with their reactions, users leave data that can be collected directly, which brings us back to stage one (Ouellet et al., 2013: 41).

9In order to contribute to a better understanding of some shortcomings described above, the goal of this text is twofold: first, we present a theoretical proposal to include time-related issues in the study of political online mobilization. Second, we clarify the interplay of three perspectives – that of the industry, the party and the users – to better comprehend the role of real-time micro-targeting strategies in the democratic process.

Theoretical framework: Digital politics and time

10To theoretically address the role of real-time targeting in political mobilization strategies, we first look at the definition of « real-time ». We then integrate this definition into a broader discussion about acceleration of politics in western liberal democracies.

Does « real-time » really mean « real » time?

11The notion of « real-time » is not exclusive to the academic world. As Weltevrede, Helmond and Gerlitz (2014: 126) point out, it « has especially been promoted by the industry to label the « post-web 2.0 era ». In this context, the notion was used to describe live activities on social networks (e.g. live tweet walls). Such features are described as producing « more authentic » contents since they are constantly updated (ibid.: 127). If we take a closer look, these activities remain constructed and can thus not be considered to be exactly « real » (ibid.). We agree with Weltevrede, Helmond and Gerlitz (ibid.: 128) that « the real in real-time is […] not a concern of immediacy […] but a question of speed and the organization of content in relation to time ».

12In fact, these authors (ibid.: 137) show that different « real-time cultures » coexist depending on the platform studied. These cultures mainly differ in regard to their pace (ibid.: 138). In turn, pacing depends on two modes of content organization and is therefore considered manifold: first, it is related to the algorithmic logics of a device defining the speed of information collection and ranking (back-end processes) (ibid.: 138; 141). Second, it depends on features for user activity that influence the circulation of contents such as posting, liking or sharing (front-end processes) (ibid.). They conclude that the construction of « real-time » is a fabrication of past, present and future related to both: algorithmic logic and user engagement (ibid.: 142). Time-related issues can therefore be approached from a user experience perspective and from a computational data-driven perspective (ibid.: 128). Consequently, the authors suggest the term « realtimeness » rather than « real-time » to better describe the « continuous movement of new content, its request and display in devices, as well as the engagement by users through web activities and the filtering of content based on freshness and relevance » (ibid.: 143).

13After having discussed the « real » part of « real-time », let us turn to the notion of « time ». Fuchs (2013a: 94) argues that time is not only a societal norm; but also an indicator for structure and its changes:

Time is the development of the existence of being from one condition to the next. This means that time is the duration of processes. […] It is irreversible, which means that when a form of being has reached a certain condition at a point of time, its structure cannot be exactly reversed.

14Simultaneously, time has its own specific economy: it is a resource organizing the economic development of capitalist societies (ibid.: 100).

15Following this definition, Fuchs (ibid.: 112) argues that with micro-targeting cycles « time has become an irrelevant factor in capitalism ». These cycles are always continuing processes: we spend time online, we leave tracks, the data is collected, analyzed, sold and used to target, hence making us spend even more time online. The existence of being is trapped in a cycle of commodified immediacy that tends to be generalized for all online activities (ibid.). In other words, time spent online is not only a social, cultural or political activity but always also a « productive time that generates economic value » (ibid.: 93). In fact, political communications scholars too often only look at the political activity. By doing so, they neglect the small but significant nuance of being always also productive. For Fuchs, contributing to political online platforms turns citizen’s speech into a double commodity: citizens become « commodities themselves » and « their consciousness becomes, while online, permanently exposed to commodity logic in the form of advertisement » (ibid., 2013b: 32). The boundaries between leisure time and labor time tend to blur, that is why Fuchs talks about social media users as prosumers (producer and consumer at once) (2013a: 108). This being said, as much as users of political platforms may not be reduced to their political actions online, neither should they be reduced to citizen-consumers. It is all about being always also something else. In this way, the permanent exposure to functions triggering an immediate reaction abolishes long-term perspectives and leads to a « political acceleration logic » pursuing the goal of taking more and more decisions in « ever-shorter time », thus eliminating deliberation (ibid.). Accordingly, the political domain is not external to time-related economic issues. In order to better understand this argument, the following section further explores political acceleration processes in western societies.

Too slow or too fast? Acceleration of politics in western liberal democracies

16In order to discuss the « realtimeness » of political online mobilization, we have to examine the broader societal acceleration processes in which this phenomenon is taking place. To do so, we refer to Rosa (2013). For this author (ibid: 81-82), temporal norms are dominant standards in our societies defying modern concepts such as reflexivity and autonomy. They consequently play a crucial role in the transformations of the cultural, economic, and political sector. While Rosa explores the three domains to elaborate a broader critical theory of late modern temporality, this text focuses on his reflection about temporality in the political sphere.

17Democracy needs time, Rosa (ibid.: 79-80) reminds us. Deliberation, the forming of opinions, the organization of debates, and the development of political programs – all these activities are time-consuming. This is particularly true in western pluralistic and internationally interconnected democracies where more and more groups of people have to be considered in the decision-making process, while the social, political and economic consequences become more and more complex (ibid.: 80). These relatively slow activities face rapid technological innovation and economic transactions as well as an accelerated tempo of cultural life (ibid.). Rosa (ibid.: 79) describes this phenomenon as a « desynchronization » of two societal processes. On the one hand, there is the democratic-deliberative opinion formation process, which is accelerating at a slow pace (ibid.: 80). On the other hand, we find the economic, technological, and cultural developments accelerating at a much higher pace (ibid.).

18The resulting tensions of this « desynchronization » transform politics. Trying to catch up with the societal domain, political actors attempt to accelerate democracy. With the help of radio, television and Internet polls (and we could add data mining strategies here), political actors are able to make more or less accurate inferences about public opinion in a matter of seconds (ibid.). Nonetheless, the outcomes are not resulting from reflexive deliberation where arguments are formulated, tried out, weighed and elaborated (ibid.). Rather, they reflect irrational feelings: opinions are linked to aesthetics and resist the power of prevailing arguments (ibid.). Rosa talks about an « aesthetic turn » in western politics, where appearances of politicians are more important than complex concepts and coherent political programs (ibid.: 81). This is why parties win elections by spinning events rather than by using valid arguments (ibid.). It seems as if words and arguments have become too slow for late modern societies, Rosa (ibid.) concludes. While westerners still hold on to the thought that politics would dictate the tempo of social and cultural development, both processes couldn’t be less synchronized (ibid.: 102). Neoliberal politics accelerate societal domains while abandoning instruments of political organization by deregulation or privatization (ibid.: 104). Social revitalization is thus more of an obstacle than a goal for modern politics (ibid.).

19Rosa’s theory can help us situate the role of time in online mobilization campaigns in its broader context under one condition: we have to clarify the term of democracy. Rosa (2013) describes democracy as a deliberative time-consuming process of opinion formation. Thus, we presume that he embraces the ideal of deliberative democracy. Yet, he never clearly defines it. This is problematic when studying digital politics. As we argued elsewhere (Hübner, 2016b), the lack of a clear definition results in very different interpretations of political online mobilization (e.g. more dialogue with citizens, opinion regulation, normalization theory). While authors who compare digital politics to ideal forms of democracy (e.g. participatory democracy) tend to be disappointed (Blondiaux and Fourniau, 2011), others focus on empirical descriptions of existing systems, which they characterize as representative and/or liberal democracies (Bennett, 2014). Still others address economic aspects of participation, concluding that corporate players determine democracy (Fuchs, 2013). However, they do not consider the users’ perspective.

20These discrepancies are not surprising. As Macpherson (1968: 2) puts it, « democracy has become an ambiguous thing with different meanings – even apparently opposite meanings ». Defining what we call « western liberal democracy » helps us to bridge the gap between moralistic ideals of democracy and it’s relation to the economic sphere. Macpherson (ibid.: 1) reminds us that democracy used to have a negative connotation « from the earliest historical times down to about a hundred years ago ». Societies used to reject democracy for its original meaning: the « rule by the common people » (ibid.: 5). For a democratic franchise to be accepted, a liberal state needed to be established: « Before democracy came in the Western world, there came the society and the politics of choice, the society and politics of competition, the society and politics of the market » (ibid.: 6). Market society pressured for democratic processes, because those who had no right to vote had no weight on the political market (ibid.: 9). In a liberal state promoting equal opportunities, there were no grounds to refuse this request (ibid.). Macpherson (ibid.: 11) concludes that market society « liberalized democracy while democratizing liberalism ». Considering the above, economics and democracy are closely intertwined. It is thus all the more surprising that the role of corporate players in digital politics is often neglected.

Discussion: Studying real-time targeting from different perspectives

21This theoretical background allows us to discuss the role of « realtimeness » in democratic election campaigns from the industry’s, the party’s, and the users’ perspective.

Creating a fantasy of immediacy and authenticity – the industry’s perspective

22Our interest in temporalities linked to political micro-targeting strategies is inspired by the fact that this phenomenon was largely ignored in political communications research. This may be linked to the fact that the term « real-time » has been promoted by the industry before political institutions appropriated it (Weltevrede, Helmond and Gerlitz, 2014: 126). Having a closer look at the digital services offered by corporations helps understand firstly how politicians use the latter for their purposes and secondly how users interact with political parties’ contents made available through corporate social media.

23As we have seen with Weltevrede, Helmond and Gerlitz (ibid.), the industry created a fantasy about immediacy and authenticity to promote a « post-web 2.0 era ». From open access to more information ( « web 1.0 »), followed by the possibility of editing ( « web 2.0 »), we reached constantly updated, always rectifiable and immediately deliverable information being considered to be even « more authentic » ( « post-web 2.0 »). Scholars constantly deconstruct the fantasy of authenticity and immediacy, showing that the web does not live up to its promises. Yet, they somehow discarded a closer look at real-time itself as a « question of speed and the organization of content in relation to time » (ibid.). Weltevrede, Helmond and Gerlitz’s (ibid.) concept invites us to define a contradictory notion, which has been employed for describing various stages of web communications. Following their distinction between back-end and front-end « realtimeness », we can distinguish two types of services provided for political communicators.

24For clarity’s sake, let us consider the example of « Facebook Live », introduced in early 2016. Since then, Facebook encourages politicians to use this feature for announcements and Q&As. Facebook’s media page devoted to « best practices […] for politicians and governments » describes how to create and share political experiences in real-time, thus giving the viewers a chance to react directly during broadcasts (Facebook, 2016a). For instance, President Trump chose this feature instead of a press conference to announce the nomination of Neil Gorusch to fill in US Supreme Court (Reardon and Sherr, 2017). The feature’s description is referring to the fantasy of authenticity: « Live is the best way to interact with viewers in real time. Field their burning questions, hear what’s on their mind and check out their live reactions to gauge how your broadcast is going » (Facebook, 2016b). Users can comment and sent emoticons to share their feelings. The « realtimeness » described here relates to « front-end processes ». However, we need to also consider back-end processes: users do not only participate in the organization of content circulation, but they also leave traces to feed algorithms. Social media business models depend on collecting more data in ever-shorter time. This is why social media platforms tend to diversify the sources for different back-end and front-end paces. They constitute possibilities « to cater to the interests of their multiple cooperating partners » developing new data and customer services (Weltevrede, Helmond and Gerlitz, 2014: 145). In fact, social media corporations try to conquer a new market: digital politics. Both, democrats and republicans, purchased promoted « trends » on Twitter, which can cost a party up to $ 100,000 per day (Kreiss and Welch, 2015: 24). From this perspective, we can indeed talk about a tendency towards a commodification of political user activities (Fuchs, 2013a; 2013b).

Accelerating political engagement – the party’s perspective

25If we turn to the party’s perspective, the « question of speed and the organization of content in relation to time » become more complex (Weltevrede, Helmond and Gerlitz, 2014: 126). Parties use platforms and services offered by corporations. Admittedly, their uses are bound by the technical constraints of the platforms’ design. Nevertheless, as with any process of appropriation, there is room for adaption, hijacking and alternative practices at the margins of these boundaries. Such practices are more often associated to the citizens’ social media use. Yet, in the end, political communicators are also social media users – only do they belong to another user group with larger financial, cultural and technical resources.

26In regard to front-end uses, political communicators have to adapt to a quickly changing digital environment: after Facebook and Twitter, politicians now have accounts on Instagram, Snapchat and Pinterest (Kreiss and Welch, 2015). Not to mention the rapid changes of these platforms themselves. Besides, political actors also re-appropriated the utopist discourse of the industry by promoting their social media presence as more authentic, less institutionalized and transparent politics (Katz, Barris and Jain, 2013; Small, et al., 2014). Even if scholars showed that this is more of a myth than a step towards grassroots decisions, parties try to use this discourse to innovate and modernize themselves in an atmosphere of political distrust and voter volatility (Theviot, 2015).

27The « realtimness » related to back-end processes can thus be interpreted as an example for what Rosa (2013) calls « desynchronization » between the political and the technological sector. Trying to catch up, they make use of back-end processes such as data mining. Let us explain their political role in the light of Obama’s reelection campaign. As mentioned above, a sole presence on social media platforms is not enough. The political messages need to be heard (Farrar-Myers and Vaughn, 2015). In regard to sophisticated political data mining strategies, this campaign ushered in a new era with the creation of Narwhal, a program that merges data from multiple digital, organizational and financial sources on one common platform (Nickerson and Roger, 2014). It was henceforth possible to collect and update information about voters in real-time by extracting data available through partisan platforms, Facebook and Twitter accounts directly linked to an Amazon cloud (Kreiss and Welch, 2015). Correlation of these different data sources allowed the creation of political profiles for voters, profiles that were used to individually target undecided citizens in swing states (ibid.). Those members of the electorate are generally difficult, if not impossible, to reach (ibid.).

28Following Rosa (2013), we can conclude that these strategies are less « democratic » as perceived since deliberation processes are skipped. With individualized targeting, it may no longer be necessary to adapt messages to the interests of user groups, for example undecided and unemployed young male midwesterners. Big data techniques work the other way around: they adapt the users’ interests to the message. Contents promoted with these techniques are choosen because they affect the individual target directly. For example, a user complaining about jobloss in Detroit on his personal Facebook account could be targeted by a post about Trump’s proposal of bringing jobs back from China or Mexico. Even if this person might not be politically interested, the content is likely to capture his or her attention. As Zwick and Denegri-Knott (2009: 241) explain for the cultural domain, « the actual purpose of the database is no longer limited to supplying epistemological power over consumers but to manufacture consumers ontologically. » Solliciting citizens individually on specific topics thus avoids public debates. Interestingly, legislators have proposed bills to restrict commercial tracking and targeting, « however, their campaigns routinely use these technologies » (Kreiss and Welch, 2015: 22).

29In the light of what we described above, these strategies paint a somewhat anti-democratic picture of politics: adapting technological and managerial solutions to complex problems by surveying the population instead of engaging in time-consuming reflexive deliberative opinion formation processes. However, we also have to consider the possibility that « voter surveillance is not necessarily anti-democratic » (Bennett, 2014: 16) – at least not from a liberal democratic, North American viewpoint. It is the party’s duty to encourage citizens to vote, therefore they « attempt to discover preferences and patterns as a more benign, efficient and legitimate way to reach voters and connect with them about public policy » (ibid.: 17). This side of the story has not yet been rigorously analyzed (ibid.). To do so, we need to turn to the citizen’s perspective.

Good intentions after all – the voters perspective

30As we have argued before, citizens are not always conscious of complex mobilization strategies. One can never be quite sure which data is collected by whom and for what purposes (Katz, Barris and Jain, 2013). But, despite their limited comprehension of mobilization strategies and micro-targeting, users are not completely enslaved by technology, politics or time. First of all, users do not have to engage with the content they are confronted with (Mabi and Theviot, 2014). Within the limits of their comprehension, they can more or less control contents associated to their profile by blocking, deleting or resetting privacy options (Van Dijk, 2013). As I argued before, citizens do not feel like being subjected to commodification, nor do they feel like « working » when participating online (Hübner, 2016a). Kruikemaier et al. (2016) even show that Facebook users resist personalized content. They are less likely to share comment or like « sponsored » posts, on the condition that they are aware of the post’s sponsoring character.

31In fact, recent US surveillance scandals have raised awareness about political and commercial profiling (Stoycheff, 2016). Some studies even show that political profiling might degenerate the « willingness to engage in political activities online » because citizens see their autonomy reduced (Mitrou et al., 2014: 13). Narwhal shows that profiling strategies focus on adapting the individual’s interests to the political or organizational message. There are risks related to this kind of tactics: individuals might loose sight of the big picture; societal debates are avoided. This points in the same direction as Rosa’s argument about the politics of « irrational feelings ». The strategies appear to be triggering immediate action without considering the consequences (ibid.). But, are citizens really clicking before considering their action?

32Exposing political views online might sometimes seem more advantageous to users than censoring themselves (Hübner, 2016a). It is possible that these practices have disruptive power: sometimes « hashtags become vehicles for supporters, and opponents, to creatively repurpose campaign content » (Kreiss and Welch, 2015: 26). Here, one might think about the Trump-hair-memes delegitimizing his political abilities. Users might feel relief by communicating their opinion and by doing so they might engage with like-minded groups to discuss their point of view (Katz, Barris and Jain, 2013). This can lead to « ego-gratifying experiences » that overshadow risks (ibid.). Even if there is evidence for more and more self-censorship (e.g. Stoycheff, 2016), while weighing the pros and the cons about making a political statement online, the need for acceptance in a peer group such as a new work environment, might overshadow the underlying risk of being profiled. Others might be looking for new communication channels, feeling that media is not a valid option (ibid.). This claim is often made by supporters of the alt right movement. Still others actually feel a moral duty to participate in democracy or the need to rectify false argumentations about public policies made by their friends and family, online as well as offline (ibid.). We have to keep in mind that all these positive views make social media efforts a « worthwhile endeavor » for both politicians and citizens (Katz, Barris and Jain, 2013: 150-151).


33The discussion of time related issues in regard to voter targeting from three different perspectives shows that future research needs to address the tensions between what the industry offers (1), how politicians reappropriate these offers (2), how users react to the resulting strategies (3) and how their reactions, in reverse, might transform society (4) or how these transformations are prevented by political or economic actors (5). Defying these tensions might help us overcome the different ways of interpreting digital politics (Hübner, 2016a, 2016b). This is not only important during election campaigns. Overcoming polarized interpretations about institutional social media uses has become particularly important in the recent « fake news debate ». Trump reapproprates corporate social media platforms, mostly Twitter, to conduct what he presents as diplomacy. For example, he tweeted not to trust « fake news » after the press criticised him for not shaking hands with German chancellor Angela Merkel after their meeting in March 2017. When analyzing reactions to this particular issue, scholars need to focus on the interplay of three levels: Trump’s strategy (level 1), which is bound by the technical constraints of the platforms’ design (level 2), and the user’s reactions (level 3). In this context, it is particularly interesting to consider what is preventing users from acting politically online. Politics? Media distrust? Or, is it the funcitoning of the platform, whose algorithms create filter bubbles allowing Trump to only preach to the already convinced?

34Finally, we want to first remind our readers that we mainly focused on the US context. Indeed, the degree to which such strategies are used depends on the culture and the budget of the party in question (building voter databases like Narwhal is extremly expensive) as well as on the country’s political system (Gibson et Ward, 2011). In Europe similar data collecting tools « would generally be regarded as illegal under personal data protection law to process data on political opinions and affiliations on people other than those explicitly signed up as members or who have regular contact » (Bennett, 2014: 6). In other countries, such as Canada, voter management softwares similar to the ones used in the US have been adopted (ibid.). In this country, voter lists are provided legally to the partys by the Canada Election Act (ibid.). However, very little is known about how they have been adapted (ibid.: 2). Second, the theoretical considerations described above need to be introduced in a broader analytical and methodological framework, which allows to operationalize a concrete proposal through case studies.



Agence France Presse (AFP), « Les mégadonnées méprisées par Trump l’ont peut-être porté au pouvoir », in La Presse, 3 December 2016.

Bennett Collin, « Voter surveillance, micro-targeting and democratic politics: knowing how people vote before they do », in Available at SSRN, 2014, p. 1-23.

Bor Stephanie E., « Using Social Network Sites to Improve Communication Between Political Campaigns and Citizens in the 2012 Election », in American Behavioral Scientist, n° 58, vol. 9, 2014, p. 1195-1213.

Bykowicz Julie, « Trump once called data « overrated ». Now he spends millions on it. », in Maclean’s, 6 October 2016.

Facebook, « Tips for Politicians and Campaigners. », 2016a.

Facebook, « Go Live on Facebook. », 2016b.

Farrar-Mayers Victoria A., Vaughn Justin S., Controlling the message. New media in American Political Campaigns, New York, New York University Press, 2015.

Fuchs Christian, Culture and Economy in the Age of Social Media, New York, Routledge, 2013a, p. 424

Fuchs Christian, « Social Media and Capitalism », in Producing the Internet. Critical Perspectives of Social Media, Olsson Tobias (dir.), Götheburg: Nordicom, 2013b, p. 25-44.

Gallagher Sean, « Built to win: Deep inside Obama’s campaign tech. How Obama’s tech team built a force multiplier with Amazon and a narwhal », in Ars Technica, November 14, 2012.

Gibson Rachel, Ward Steve L., « Renouveler le parti? Les stratégies de campagnes et d’organisation en ligne des partis britanniques », in Les partis politiques sur le Web, Greffet Fabienne (dir.), Paris, Presses de Sciences, 2011, p. 109-123

Hübner Lena A., « Opinion Regulation or Civic Dialogue? Seeking New Theoretical Frameworks for the Study of Digital Politics », in Journal for Communication Studies ESSACHESS, n° 9, vol. 2/18, 2016b, in press.

Hübner Lena A., « Où en est le débat sur le potentiel démocratique des TNIC? Regards croisés sur la participation citoyenne en ligne », in Essachess, Journal for Communication Studies, vol. 9, n° 2 (18), 2016, p. 43-51.

Issenberg Sasha, « Obama’s White Whale. How the campaign’s top-secret project Narwhal could change this race and many to come », in Slate, February 15, 2012.

Kreiss Daniel, « Social media did not give us Donald Trump and it is not weakening democracy », in US Election Analysis 2016: Media Voters and the Campaign. Early reflections from leading academics, Lilleker Darren, Jackson Daniel, Thorsen Einar and Veneti Anastasia (dir.), Bournemouth, Center for the Study of Journalism, Culture and Community and Centre for Politics and Media Research, 2016, p. 75.

Kreiss Daniel, Welch Creighton, « Strategic Communication in a Networked Age », in Controlling The Message? New Media in American Political Campaigns, Farrar-Myers Victoria A. and Vaughn Justin S. (dir.), New York, New York University Press, 2015, p. 13-31.

Kruikemaier Sanne, Sexgin Minem and Boerman Sophie C., « Political Microtargeting: Relationship Between Personalized Advertisement on Facebook and Voters’ Responses », in Cyberpsychology, Behavior and Social Networking, n° 19, vol. 6, 2016, p. 367-372.

Laurent Samuel, « L’élection de Trump et les trois échec du « big data » électoral », in Le Monde, 11 Novembre 2016.

Lewis Seth C., Carlson Matt, « The dissolution of news: selective exposure, filter bubbles, and the boundaries of journalism », in US Election Analysis 2016: Media Voters and the Campaign. Early reflections from leading academics, Lilleker Darren, Jackson Daniel, Thorsen Einar and Veneti Anastasia (dir.), Bournemouth, Center for the Study of Journalism, Culture and Community and Centre for Politics and Media Research, 2016, p. 78

Mabi Clément, Theviot Anaïs, « Présentation du dossier. S’engager sur Internet. Mobilisation et pratiques politiques », in Politiques de communication, n° 2, 2014, vol. 3, p. 5-24.

Macpherson Crawford B., The real world of democracy (4th edition), Toronto, Canadian Boradcasting Corporation, 1968 (1964).

Mitrou Liliane, Kandias Miltiadis, Stavrou Vasilis and Gritzalis Dimitris, « Social media profiling: a panopticon or omniopticon tool », Conference held at the 6th Biannual Surveillance and Society Conference, Barcelone, Espagne, avril 2014, p. 1-15.

Nickerson David W., Rogers Todd, « Political Campaigns and Big Data », in Journal of Economic Perspectives, 2014, n° 28, vol. 2, p. 51-74.

Ouellet Maxime, Mondoux André, Ménard Marc, Bonenfant Maude and Richert Fabien, « “Big Data", Gouvernance et Surveillance », research report produced by Centre de recherche sur la communication, l’information et la société (CRICIS), Université du Québec à Montréal, Montréal, 2013.

Rosa Hartmut, Beschleunigung und Entfremdung. Entwurf einer kritischen Theorie spätmoderner Zeitlichkeit, Berlin, Suhrkamp/Insel, 2013.

Reardon Margarithe, Sherr Ian, « Trump names Suprime Court nominee on Facebook Live », in C-net, 31 janvier 2017, s.p.

Shorey Samantha, Howard Philip N., « Automation, Big Data, and Politics: A Research Review », in International Journal of Communication, 2016, vol. 10, p. 5023-5055.

Small Tamara. A., Jansen Harold, Bastien Frédéric, Giasson Thierry, and Koop Royce, « Online Political Activity in Canada: The Hype and the Facts », in Canadian Parliamentary Review, Winter Edition, 2014, p. 9-16.

Stoycheff Elisabeth, « Under Surveillance: Examining Facebook’s Spiral of Silence Effects in the Wake of NSAInternet Monitoring », in Journalism & Mass Communication Quarterly, 2016, n° 93, vol. 2, p. 1-16.

Theviot Anaïs, « Dépoussiérer les partis politiques français sur Internet? », in Savoir/Agir, 2015, n° 2, vol. 32, p. 37-43.

Van Dijck Josée, « Social Media Platforms as Producers », in Olsson Tobias (dir.), Producing the Internet. Critical Perspectives of Social Media, Götheburg, Nordicom, 2013, p. 45-62.

Weltevrede Esther, Helmond Anne and Gerlitz Carolin, « The Politics of Realtime: A Device Perspective on Social Media Platforms and Search Engines », in Theory, Culture and Society, 2014, n° 31, vol. 6, p. 125-150.

Zwick Detlev, Denegri-Knott, Janice, « Manufacturing customers », in Journal of Consumer Culture, 2009, n° 9, vol. 2, p. 221-247.


1 Translation from French by author: « This should have been the triumph of “big data", using the (inappropriate) expression for modern voter-targeting methods. »


Université du Québec à Montréal (UQAM). Lena A. Hübner est doctorante en communication à l’université du Québec à Montréal (UQAM). Après son B.A. en Études francophones (concentration médias) à l’Université Albert-Ludwig à Freiburg (2012) et des expériences de travail en relations publiques, elle choisit une carrière universitaire. Depuis son mémoire de maîtrise (UQAM, 2014), elle étudie la communication politique sur les réseaux socionumériques. Elle est notamment l’auteure de l’article « Opinion Regulation or Civic Dialogue ? Seeking New Analytical Frameworks to Study Digital Politics » publié dans ESSACHESS – Journal For Communication Studies (décembre 2016). Elle est également coordonnatrice des activités scientifiques du centre de recherche interuniversitaire CRICIS (Centre de recherche interuniversitaire sur la communication, l’information et la société).

© Presses universitaires de Paris Nanterre, 2018

Conditions d’utilisation :