5. Open and Closed Government Data

IT IS IMPORTANT TO NOTE from the outset that the labels “open” and “closed” are not essential but relational, adhering to particular moments in space and time. When articulated to data, the identity of each, and the binary opposition itself, is contingent on the political climate, the market, the security complex, technological capacities, and the veillant conjuncture. The tendency toward secretion identified earlier should be enough to indicate the provisional nature of any identification of data as “closed.” Likewise, because of the inherently opaque nature of much “open” data (which leaves many questions unanswered, such as for whom was this data collected? To what ends?), “open” data is never simply open or transparent.

Open government data is generally understood as the provision of big and small digital data on the part of government agencies. Alongside a few critical voices (e.g., Morozov 2013), open government data is celebrated in the mainstream for democratizing knowledge distribution and research, invigorating economies, increasing efficiency, ensuring accountability, and operating as a key element in digital democracy or “Democracy 2.0” (e.g., Goldstein and Dyson 2013). Open government data is data shared with no depletion: sharing in the sense not of division but of giving multiple citizens access to the same thing.

By contrast, we can understand closed government data as that data that is withheld from public view, whether in the interests of privacy, diplomacy, or national security. As “close” brings forth etymological associations from the old French clore, “to shut, to cut off from,” we can see how citizens are cut off from the state’s data, even data they have (perhaps unknowingly) shared. In sharing this kind of data, we have in effect given it away. Our “share” can never yield. That is to say, without the interventions of whistleblowers or hackers, closed government data will never be given the opportunity to be put to uses other than those determined appropriate by the state.

In its open formation, government data is deliberately and strategically shared by the collecting agent; in its closed formation, data is deliberately and strategically not shared. With respect to closed data, particularly in the case of state surveillance, citizens share data with a proprietary agent in exchange for the privileges that come with citizenship. We might, that is, consciously or unconsciously, explicitly or implicitly, consider the collection of our GPS data or phone metadata a fair price to pay for the freedoms, benefits, and protections that come with owning a British (or Australian, German, American, etc.) passport. This pragmatic attitude to sharing with respect to closed government data, the transmission of citizens’ activity to a veillant other, is echoed in the experience of digital consumers in general. That is to say, users of social media and search engines are familiar with making trade-offs between services they want and acquiescence to data collection. As well as protocological in a technological sense, then, sharing also needs to be thought as a political, cultural, and industry standard. It “frame[s] the terms and parameters by which elements of a system interact and behave” (McStay 2014, 5).

Sharing is not something we do after possessing data but is the basis on which having any relation with that data can be possible at all. All of which does not necessarily indicate that the data we have shared is digested and absorbed and immediately put to work by any surveillant agent. Rather, to borrow the words of Gus Hunt, the Central Intelligence Agency’s chief technology officer, it indicates that “collect[ing] everything and hang[ing] on to it forever” (see Ingram 2013) relies on the idea that the archive is “structurally speculative” (Andrejevic and Gates 2014). The uses to which collected data will be put and the meanings it will be given are dependent on future algorithms and political concerns. This means that in a networked era, we are always already sharing without any actor in the system necessarily knowing precisely why. The principle of sharing overrides any uncertainty over the uses to which shared data can be put. Such a condition is obviously in the interests of commercial and state surveillance, which, in general, currently has monopolies on accruing economic or security value from big, aggregated archives of data.[1]

Although it might seem as though closed government data is open data’s evil twin, open government data is not excluded from this veillant assemblage. All shared data mobilize a politics of visibility, a demand to align with a political and ethical distribution of the digital sensible. Though less potent and pervasive than their closed counterparts, open government data initiatives also involve veillance because the sharing of data includes a call to watch and act on that data—we are envisioned, watched, imagined as entrepreneurial and auditing viewers or subjects. Within a logic of shareveillance, both closed and open data contribute to the construction of an antipoliticized data subject and public. By looking at the national security dataveillance revealed by Edward Snowden and the open government data initiatives implemented by the U.S. government, I will consider how shareveillant subjectivity is produced in the context of the state.

“Closed” Data, Securitized Veillance

The data collected by the NSA, GCHQ, and other security agencies around the globe is mostly experienced as “closed”: inaccessible to those without security clearance. Before Snowden revealed the programs implemented to collect communications data and metadata—programs such as PRISM, which, since 2007, has permitted the NSA to access data from service providers, and Tempora, which saw GCHQ placing interceptors on fiber-optic cables that handle Internet traffic in and out of the United Kingdom—the programs, too, were closed, secret, opaque. That is not to say that there weren’t all kinds of secretions regarding those practices: details or speculations erupting now and again into the public sphere through reportage, whistleblowing, or popular cultural representation (what Tim Melley [2012] refers to as “the covert sphere”—the representation and cultural counterpart to the covert sector). Rather than focusing on the content of the revelations and whether such news really was new, however, what I am interested in is the conceptual apparatus that was available to those who wanted to resist or challenge this aspect of the shareveillant assemblage.

Though domestic protests were subdued, calls to end the NSA’s activities, as evidenced on the banners held at the march on Washington in October 2013, were expressed as an “end to mass spying.”[2] Exercising people’s imaginations and offending their constitutional rights was the suggestion that their own government had the ability to see them and their actions without their knowledge or explicit consent. While many would agree that this move toward ubiquitous communications surveillance is, indeed, something to resist, the appeal to “privacy” falls rather flat. Privacy is like the light we see from an already dead star. We cling to it even though we live in what our digital conjuncture has essentially rendered a postprivacy paradigm. This doesn’t mean that the concept of privacy is no longer important: it still organizes legal processes, rights-based debate, and common understandings of our own sense of self. In some ways, as Andy McStay (2014, 2) points out, “many social changes since the industrial revolution involve a net increase in privacy, be this less familiarity with our neighbours, more geographically dispersed family arrangements, working away from home, weakening of religious authority . . . greater possibility of children having their own bedrooms, increase in car ownership (versus public transport).” Yet, the risk of still appealing to privacy in an era of ubiquitous dataveillance and closed, securitized data is that it reduces rather than increases political agency precisely because it misunderstands the subjectivity in question and because privacy claims are particularly weak when it comes to collective politics. It cannot, to put it in the terms to which this book adheres, redistribute the digital sensible.

To take the first of these issues, the appeal to privacy in the wake of the Snowden revelations misreads the deindividualizing character of mass covert data mining. The fear expressed on the banners and placards of the protests is that the state sees the crowd as individuals—a mass that is made up of many Is, the privacy each of which has been infringed. The concept of privacy imagines a state violating the rights of a fully self-present liberal citizen. But the way in which data mining works means that the security services are not particularly interested in the actions of individual citizens except inasmuch as those citizens are data subjects: how they contribute to a background pattern on which an evolutionary algorithm can work to recognize minority anomalies. As Clough et al. (2014, 154) write,

in the case of personal data, it is not the details of that data or a single digital trail that is important, but rather the relationship of the emergent attributes of digital trails en masse that allows for both the broadly sweeping and the particularized modes of affective measure and control. Big data doesn’t care about “you” so much as the bits of seemingly random info that bodies generate or that they leave as a data trail.

Nova Spivack (2013), in an article infused with techno-utopianism, puts it slightly differently: “We are noise, not signal.” Spivack problematically invokes this argument to excuse the NSA’s data scraping (echoing the common mantra that if you have done nothing wrong, you have nothing to fear from surveillance). But we can also come to a different conclusion. The configuration of citizens as noise, not signal, points toward our delimited role within the shareveillant assemblage read from the perspective of closed data.

The offense, I suggest, is less the intrusion into private space and more the disavowal of the public as potentially political. The surveillant state imagines its citizens in this configuration as primarily an aggregated data set. It is not that citizens are being spied on that is of most concern in this view but that unless their actions are flagged up as extreme outliers, they are not considered fully formed political agents worthy of anything more than bolstering an algorithm for data analysis. To avoid peddling a fantasy about once fully agential individuals, we can invoke the idea of Michael Hardt and Antonio Negri’s (2000; 2004) “multitude” (that which both emerges from Empire and has the potential to subvert this new global sovereignty). The multitude does not resemble traditional configurations but rather, as Jeremy Gilbert (2008, 164) elucidates, suggests “a field of collectivity which is composed of singularities: unique points of intersection and potential self-invention which cannot be subsumed into any simple totality nor reduced to the status of individuals.” This is important because it points to an idea of political agency beyond individualism toward a distributed or networked agency. (The concept of multitude also, we should note, works from the assumption that resistance can emerge from the very same body that is seemingly invested in the perpetuation of regressive forms of power.) When the collective singularities of the multitude are translated into a data set, a hopeful reconfiguration of the political is neutralized. Rather than being of comfort, the fact that citizens only count in terms of their role as flat data has an effect on the scope of political agency to alter the distribution of the sensible (even if this is only an imagined agency) and the possibilities therein that this implies for effective, counterhegemonic collective action.

To take the second limitation of privacy—that it is a poor foundation on which to build collective action—it is one not tied to the digital/big data turn, but it is nevertheless a critique that has been given a new inflection within that context. Although it is not within the scope of this book to provide an extensive survey of work on privacy, it is worth reminding ourselves that privacy has been subject to critique from the Left for its connections with individualism, the perpetuation of oppression, and property. To call on the right to privacy is to frame the debate in terms of an individual’s right to limit the access other people, the state, or commercial entities might have to her “content” (data, thoughts, feelings, information, communications) at any time. It reinforces a sense of a self that lives in political isolation. Therefore, even when people coalesce around privacy concerns, step into the light of the demos, they do so to insist on their right to step back into the apolitical shadows of individualism, away from the possibility of collective creativity or an identity-in-common. Equally, the feminist Left has challenged privacy for allowing gendered inequality and even violence to exist beyond the purview of society (e.g., MacKinnon 1989). The private sphere potentially offers respite and refuge but might be experienced more like a prison for some oppressed women. Moreover, a notion used most adeptly to protect property, privacy has been an uneasy bedfellow for collectivist leftist politics more interested in the redistribution of wealth.

In short, privacy claims are ill equipped to fundamentally challenge the dataveillance being conducted and its essentially unidirectional sharing of information that contributes to the shareveillant subjectivity I am outlining. But closed data and opaque data practices are only half of the story.

Open Data

The provision of open data is a professed concern and commitment for many liberal democracies today. The Open Government Partnership currently has seventy-five participating countries, not all of which, it should be noted, could be described as liberal democracies (the Philippines and Turkey, for example), all at various stages in implementing open government plans.[3] Though its funding was cut in 2010 from $35 million to just $8 million, and its future under Donald Trump is precarious to say the least, the United States’s open government data portal, Data.gov, is notable in this regard, providing public access to many different data sets produced by government agencies. There are a good number of reasons to applaud transparency measures such as this, especially when compared with closed regimes in which extreme forms of corruption are endemic. And yet, this might be an inflammatory comparison, or at least a false construction of the issue. For within ostensibly “open” liberal democracies, we must ask which forms of openness take precedence in any particular era and what kinds of subjectivities they promote. Regions wishing to make the move toward more open forms of society and state often look to those dispositifs already in operation elsewhere, and thus forms of openness, and the political settlements they compound, travel through what I have elsewhere termed (with plenty of caveats, given the contentious nature of the latter word) “transparency imperialism” (see Birchall 2015).

In sharing its data sets with citizens, the state adds to the interpellation of shareveillant subjects. Althusser’s (1971) vivid and therefore oft-quoted scenario depicting the process of becoming a subject sees a figure that responds to being hailed by a “Hey, you there!” (174). The calling policeman sees the subject, who turns to acknowledge the ideological call. The shareveillant subject is hailed with an added imperative—“Hey, you there! Come closer and watch.” The subject turns not only to be seen but also to become vigilant. The shareveillant subject is surveilled (possibly without his or her knowledge, given all I have said regarding dataveillance) but also has to be seen to be seeing. More accurately, the shareveillant subject is asked to see through: the transparency of the state is the interface that hails us, and we cannot but occupy the position (whether we feel technically capable or not, whether we perform the function or not) of auditor, analyst, witness.

For example, on Data.gov, we are presented with an array of data sets and documents that contain statistics of varying importance in a variety of formats (CSV files, Word documents, PDFs, etc.). Watching and seeing through (and acquiring and refreshing the technological competence required to do so) become forms of immaterial labor. In the process, a characteristic of neoliberal logic is performed: the subject is bequeathed responsibility without power. He or she is given the responsibility to watch without the expertise to know what to look for nor the power to act in a meaningful way on what might be found. Echoing Max Horkheimer’s (1989, 265) conclusion that the cultural industries “constantly profess their adherence to the individual’s ultimate value and his inalienable freedom, but . . . operate in such a way that they tend to foreswear such values by fettering the individual to prescribed attitudes, thoughts, and buying habits,” open data offers the promise of agency and freedom (of information) but shapes the scope and manifestation of those ideals to entrench shareveillance.

The act of looking, of being asked to look, is more complicated still, for even while this call to be vigilant is made, the reach widens to draw in unelected mediators: third-party application developers and data visualizers. Data entrepreneurs step into the ideological call to help fulfill the demand to watch, to see (through) the state. The “datapreneurs” happily perform this function—taking the data sets made available by the government on websites like Data.gov and turning them into user-friendly interfaces. In doing so, those datapreneurs are also, themselves, responding to a hailing: to help operationalize the new “data economy.” This call manifests in different forms. The White House, under Obama, hosted “datapaloozas” on health, safety, and energy—events where datapreneurs could make contact and experiment with using open government data. In Helsinki, Finland, events called “Helsinki Loves Developers” facilitate networking between city officials and application developers, while the city runs competitions (such as “Apps for Finland” and “Datademo”) to reward Web-based or smartphone apps that use public data. The U.K. innovation agency Technology Strategy Board partially funds the Open Data Institute, the goal of which is to “train, nurture and collaborate with individuals around the world to promote innovation through open data.” This is because the provision of open government data is fueled not only by its purported social value but also by its economic value.[4] In fact, in his foreword to “Open Data White Paper: Unleashing the Potential,” presented to the U.K. Parliament in 2012, the Right Honorable Francis Maude mentions “demonstrating the value of open governance to economic growth” before “improved citizen engagement and empowerment” (5) when explaining the priorities of the British chairmanship of the Open Data Partnership, which goes under the heading “Transparency Drives Prosperity.”

The data economy is earmarked to stimulate and fuel economies, with some impressive potential figures being cited. For example, a report by the McKinsey Global Institute from 2013 estimates that the global market in open data, measured in terms of job creation, profit margins, and efficiency savings, is worth up to $5 trillion a year. The World Bank (2014, 20) finishes its own macroeconomic report on the value of open data by stating, “While sources differ in their precise estimates of the economic potential of Open Data, all are agreed that it is potentially very large.” Though these two reports are not limited to open government data, a report from the Open Data Institute (2015) analyzing various reports on the value of open data tells us that “those studies focused on the value of public sector open data alone found that it is worth between 0.4% and 1.5% of an economy’s GDP.”

The datapreneur is the key figure in the success of the (open) data economy, as the actor who must harness the potential of the data to create value from raw data sets. That could include the team behind Zocdoc, an app that uses data from the U.S. Census Bureau and the Centers for Medicare and Medicaid Services to assist patients as they manage the complex system of U.S. health insurance and provision, or the datapreneurs behind Spotcrime, an app that again uses U.S. census data, but this time to produce a searchable map that contains crime statistics for each area. In the United Kingdom, the Findthebest website uses the open census and Department of Education data on Data.gov.uk to produce searchable league tables on primary and secondary schools with information as detailed as average teacher’s salary and the distribution of a school’s budget. For the state and datapreneur alike, data is configured as a resource ripe for mining and commodification.

Where does this leave the shareveillant subject? At once asked to watch the newly transparent state, with all its data organs on display, and to rely on the mediating and translating functions of datapreneurs to do so, this subject is one whose relationship to government is shaped by the market. Neoliberal ideology has long ensured the public acquiescence to and accommodation of the marketization of many aspects of social and political life, from education to health. What is new here is that the market (embodied by third-party datapreneurs) gets to decide the very stakes of the political—and many apps made possible by Data.gov and Data.gov.uk are concerned with real estate, finding the best school or surgeon, checking food safety statistics, transport information, and weather data.[5] They help citizens navigate a variable field of provision rather than evening out that field (by, for example, implicitly encouraging people to avoid underperforming schools rather than ensuring that those schools receive more assistance). I am arguing that the reliance on data mediators or datapreneurs to make the transparency of the state meaningful means that, ultimately, the market determines the distribution of the sensible—what we can know, see, hear, touch, encounter. In terms of sharing, only those data sets that can be made to yield profit (in some form) will be shared in such a format that the data can be received, understood, and rendered actionable.

The shareveillant subject is required to be vigilant and veillant to be a fully engaged citizen. Immediately, however, this impossible vigilance of the open state is acknowledged, and mediators are called on to select and package information. This means that veillance is always watchfulness, not of the fully transparent state, but of selected mediations brought forth. Transparency is obscured by its own impossible glare—only the data that the market has primed us to want (usually data that can help us make apparently “informed” choices in a complex public–private landscape) assume the face of state transparency in the data economy. The risk is that it becomes increasingly difficult to participate in and navigate the state outside of these commodified, shaped, and edited forms of aggregated data. What kind of “control” is in play here?

Gilles Deleuze’s (1992) short but influential essay “Postscript on the Societies of Control” offers a reading of power, governmentality, and political economy in postdisciplinary societies that can help here. In the “Postscript,” Deleuze evokes Foucault’s work on discrete and autonomous units of confinement characteristic of disciplinary societies to establish more contemporaneous dispersed mechanisms of control. Data-driven government transparency, which as a move from administrative to democratic accountability might seem like an unequivocal good, can be problematized through Deleuze’s societies of control to give a clear sense of why techniques of emancipation can be experienced otherwise.

Although Deleuze’s essay predates by several decades the data-driven government transparency on which this book focuses, it can help us to understand this phenomenon in a number of ways. First, just as Deleuze identifies the way in which environments of enclosures (like the prison, the hospital, and the school) are now subject to forms of free-floating control, we can see how, in opening up government, making (some of) its boundaries porous through open data, outsourcing, and responsibilization, data-driven transparency ensures that the business of governance (and citizenship) is without boundaries or end. So while government becomes “smaller” in many ways, to allow the market to do much of the work previously accorded the state, government simultaneously has a ubiquitous presence in the form of raw data or, perhaps more important, digital tools to help navigate the state in its market form. Like the corporation, which has replaced the factory, data-driven government transparency makes government “a spirit, a gas” (4).

Second, through Deleuze’s observation that control mechanisms are inseparable variations, we can see open data and the data economy in relation to other “modulations” (4) in the neoliberal field. As opposed to enclosures, which are molds or castings, modulations are “like a self-deforming cast that will continuously change from one moment to the other” (4). Modulations are “in states of perpetual metastability that operate through challenges, contests” (4). The data economy as an entrepreneurial enterprise, in which open data plays a crucial role, necessarily requires such metastability to become profitable. Moreover, datapaloozas and hackathons are clear examples of the “challenges” and “contests” that drive remuneration and profit. Crucially, the logic of control means that each experience of governmentality is a continuity. Open data is thus one modulation within a “continuous network” (6) that demands perpetual vigilance and innovation.

Third, in a formulation that helps us assess what is at stake in the technological conditions of open data, Deleuze shows that we can be controlled through the conditions of access and confinement. He cites Félix Guattari’s example of an electronic card that can open barriers in a city but that “could just as easily be rejected on a given day or between certain hours; what counts is not the barrier but the computer that tracks each person’s position—licit or illicit—and effects a universal modulation” (7). The emancipatory qualities of open data involve control because of the entrepreneurial metastasis required to convert a previously extraeconomic form into capital, the continuous vigilance of data subjects, submission to market logic as a form that can determine the scope of the political, and forms of dataveillance that track and trace.

It would be an exaggeration to claim that Deleuze is nostalgic about the certainties of disciplinary societies, but it is fair to say that environments of enclosure with their clear borders offer more opportunities for distinctive, oppositional subject positions or the creation of counterpublics. While not using disciplinary societies as her reference point, Jodi Dean (2009, 51) writes, “Whereas the Keynesian welfare state interpellated subjects into specific symbolic identities (such as the worker, the housewife, the student or the citizen), neoliberalism relies on imaginary identities. Not only do the multiplicity and variability of such identities prevent them from serving as loci of political action but their inseparability from the injunctions of consumerism reinforces capitalism’s grip.” In general, the neoliberal subject is “one who strategizes for her- or himself among various social, political and economic options, not one who strives with others to alter or organize these options” (Brown 2005, 43). Even armed with the information provided through open data, the shareveillant subject, specifically, is a precarious position from which to enact or coalesce counterhegemonies because it is reliant on continuing the forms of control to which it is subjected. That is to say, the shareveillant subject has learned to experience and accept agency and subjectivity via the forms of veillance and sharing described in this book.