the various voter surveillance practices recently observed in the United States, assesses the extent to which they have been adopted in other democratic countries, and discusses the broad implications for privacy and democracy. Four broad trends are discussed: the move from voter management databases to integrated voter management platforms; the shift from mass-messaging to micro-targeting employing personal data from commercial data brokerage firms; the analysis of social media and the social graph; and the decentralization of data to local campaigns through mobile applications. The de-alignment of the electorate in most Western societies has placed pressures on parties to target voters outside their traditional bases, and to find new, cheaper, and potentially more intrusive, ways to influence their political behavior. This paper builds on previous research to consider the theoretical tensions between concerns for excessive surveillance, and the broad democratic responsibility of parties to mobilize voters and increase political engagement. These issues have been insufficiently studied in the surveillance literature. They are not just confined to the privacy of the individual voter, but relate to broader dynamics in democratic politics.Bennett states
Surveillance has arguably become routine, normal or “everyday” and reaches into every corner of modern life (Bennett et al. 2014). It is, according to David Lyon, “any collection and processing of personal data, whether identifiable or not, for the purposes of influencing or managing those whose data have been garnered” (Lyon 2001: 2). And surveillance is not simply about large organizations using sophisticated technology; it is also something that individuals increasingly engage in. It is good and bad, top-down and bottom-up, and directed to humans, non-humans and spaces. It is a mode of power and central to the new forms of governance within modern and post-modern societies (Haggerty and Ericson 2006).
High-level conceptualizations about the nature and causes of surveillance help associate current practices with broad and profound structural transformations in contemporary societies (Lyon 2007). But macro- level theorizing only takes us so far in understanding the nature of individual and social risks in particular contexts (Nissenbaum 2009). Thus, surveillance has particular, and somewhat different, effects depending on whether we are consumers, employees, immigrants, suspects, students, patients or any number of other actors. Theorizing surveillance on a grand level tends not to expose the more subtle relations, norms and harms associated with the institutional and informational relations that attend the particular roles that we play and negotiate in our everyday lives. As Haggerty and Samatas remind us: “A global community of scholars has produced excellent case studies of the dynamics and normative implications of different surveillance practices, but run into more difficulty when it tries to make generalizations about surveillance tout court, often because the surveillance dynamics and implications of, say, spy satellites, are wildly different from those of DNA testing” (2010: 3).
Just as the literature speaks of consumer surveillance or employee surveillance, and analyzes the different practices and issues that arise in these different contexts as we play these different roles, so we can speak of “voter surveillance.” In our capacities as participants, non-participants or potential participants in the democratic electoral process, personal data is increasingly captured and processed about us for the purposes of regulating the fair and efficient conduct of elections and also to influence our behaviors and decisions (Bennett 2013a, 2013b). The norms, dynamics, and dilemmas are, and should be, different in this voting context.
Very little has been written in the broader academic literature about voter surveillance. There is a certain amount of important journalistic commentary on the contemporary trends in micro-targeting in the United States (Issenberg 2012), and on how these practices have been imported to Canada (Delacourt 2013). Communication scholars have analyzed the new “tech-driven” politics as part of a larger assessment of changing campaign techniques (Howard 2006; Hendricks and Kaid 2011). And a number of political scientists have tried to evaluate whether or not new media campaigns affect voter engagement and behavior (Lees-Marchment, Stromback and Rudd 2009; Small 2010; Lees-Marchment 2011; Davies and Newman 2012). Very little of this commentary, however, engages with the larger question about how data about voters is being mined and profiled, nor evaluates the individual risks to privacy and the general implications for democratic politics.
This paper is intended to begin to fill that gap and inspire further analysis and research. The first section of the paper draws upon previous research to distill some of the most important trends in political campaigning, which has implications for the capture and processing of personally identifiable data. The paper then analyses how these practices are likely to influence the democratic politics of different states depending on different electoral practices and party systems. It then offers a set of broader theoretical reflections about the implications for democratic practice, drawing upon the recent literature on the complex and paradoxical tensions between surveillance and democracy (Haggerty and Samatas 2010).He concludes -
It is widely assumed that surveillance and democracy lie at opposite ends of a normative continuum (Haggerty and Samatas 2010: 1). Despite the insistence from Lyon (2001) and others that it should be framed in neutral terms, surveillance still assumes a place in the popular consciousness as a negative force that compromises those freedoms upon which democratic societies are founded, including privacy, and freedom of speech and association. Surveillance seeks to render individual behaviors and preferences transparent in ways that make them conform to pre-existing categories and norms. It inspires conformity, control, and obedience. It discourages the individualism, autonomy, and creativity that democracy requires and thrives upon. As Paul Schwartz remarks, surveillance has “a negative impact on individual self- determination; it makes it difficult to engage in the necessary thinking out loud and deliberation with others upon which choice-making depends (1999: 1701).
The anti-democratic nature of surveillance is reinforced by the prevalence of Orwellian and Kafaesque metaphor and imagery. Various symbols have been used over the years to equate excessive surveillance with the slippery slope to authoritarian repression. That message is continually reinforced by a network of privacy activists that engage in a symbolic politics to create awareness and expand their networks (Bennett 2008: 106-7). We are currently in the middle of a wide-ranging international debate about the appropriate role for security and intelligence services in the wake of the revelations from National Security Agency whistle-blower Edward Snowden. The bewildering range of surveillance programs initiated without appropriate accountability and oversight by the National Security Agency, and its sister organizations in the “Five-Eyes” countries, are generally challenged because of their fundamentally anti-democratic nature (Greenwald 2014).
If it were discovered that the NSA had backdoor access to the kind of voter management databases described above, then similar denunciations would no doubt occur and be justified. Thus, it is not difficult to find arguments that the practices described above are also, fundamentally undemocratic, or even anti- democratic. These tactics might be criticized for their tendency to treat citizens as unthinking consumers, ready to respond with their votes in the same way that they respond with their money. Micro-targeting divides us into niche markets and avoids the hard work of building consensus and national visions. It arguably creates parties and candidates that do not convey a general ideological framework for governance, but a series of carefully chosen, focus-group analyzed, messages to key segments of the electorate in key marginal districts. This messaging need not be internally consistent, nor framed within a larger set of policy ideas. Thus parties only need to mobilize key voters in key places; and if the votes of others are suppressed, then so be it. In her analysis of these trends in Canada, Delacourt (2013: 328) concludes: “Instead of turning consumers into citizens, it has accomplished the reverse. Canadian politics went shopping for votes, and the voters went shopping.” The science of “winning elections” may have the effect of turning people off the political process.
A critical response to voter surveillance, and the consumerization of the political, would contend that the practices surveyed above discourage engagement and deliberation, in favor of the increasing individualization of political space in which we are assumed to have preferences and tastes that only need to be unearthed using the most sophisticated technology to determine what public policies and goods voters “want”: a tax break here; a subsidy there; an improvement to the local school; a clean-up of the neighborhood lake; and so on. Thus the critique of voter surveillance might sit comfortably within a broader critique of neo-liberal governance and of the shrinking public sphere.
The argument is more complex, however. Political parties have a responsibility to mobilize and educate supporters. In so doing, they attempt to promote higher levels of participation and engagement in the political process. Voter surveillance practices have, in part, emerged as a response to the failures of traditional and crude forms of mass messaging through television. Arguably parties can encourage more people to vote and reinforce voters’ agency, if they know more about their beliefs and preferences. There may be some evidence that the 2008 and 2012 presidential campaigns in the United States, the first to be waged with the full range of new media technology to reach voters of all demographic and socio- economic characteristics did, indeed, have a small, but noticeable impact on participation rates and voter engagement, particularly among the younger “millennial generation” (Hendricks and Kaid 2011).
There will continue to be debate about the extent to which the increase in voter turnout in these elections, and among this age group, is attributable to new media and micro-targeting, but the point remains that voter surveillance is not necessarily anti-democratic. At least, the public interest on the other side of the equation is different. The balance is not between the privacy interest and security, nor between privacy and the profit-motive. Instead, we confront a rather different set of interests that need careful consideration and weighing before condemning or regulating the ways that candidates and parties capture data on citizens and use that information to encourage political engagement and participation. Those issues have not been thoroughly analyzed in democratic theory, nor subjected to rigorous empirical examination in different states with different legal requirements and electoral tradition.
At root the contestation of values is reflected in two broad and rich traditions of democratic theory. The first is a liberal vision, which sees the main test of democracy as a representative system, based on majority rule but with established constitutional protections for minority and individual rights. Privacy has tended to be regarded and justified within a broad liberal paradigm (Bennett and Raab 2006) and plays an important role within liberal democratic theory because it: prevents the total politicizing of life; promotes the freedom of association; shields scholarship and science from unnecessary interference by government; permits the use of a secret ballot; restrains improper police conduct such as compulsory self-incrimination and unreasonable searches and seizures; and it serves also to shield institutions, such as the press, that operate to keep government accountable (Westin 1970: 25). So, under this dimension, privacy is protective of individuals and specific organizations from obtrusive invasions that would detrimentally affect their ability to participate in politics or go about daily life.
A second broad tradition sees the test of democracy less in the protection of rights, and more in the participation of a citizenry to take charge of its own affairs. As the liberal democratic tradition has been strained under increasing levels of partisan de-alignment and voter apathy, so scholars have renewed interest in a more “participatory” forms of democratic practice (Pateman 1970). If one creates a more participatory environment, people will be more prepared for the tasks of self-government. Engagement in social and community institutions raises the stock of “social capital” (Putnam 1993), levels of interpersonal trust, and the ability of individuals to translate the “I” into the “we.” As Pateman argues: “individuals learn to participate by participating” (2012: 15).
There may be, however, a less critical response to voter surveillance, which sees the attempt to discover preferences and patterns as a more benign, efficient and legitimate way to reach voters and connect with them about public policy. The conversation on the doorstep, over the phone, or in the social media environment, can therefore be more in tune with what voters perceive and desire. Thus, voter surveillance, like surveillance more generally, is “Janus-faced” (Lyon 2001). It at least requires us to analyze and judge its complex dynamics according to a different set of criteria than those used when we evaluate the security practices of the state, or the profit-driven consumer monitoring by the private sector.