The contentious politics of data: An introduction.
I n 2019, Davide Beraldo & Stefania Milan published an article titled From data politics to the contentious politics of data signalling a connection to an earlier article by Ruppert, Isin and Bigo on Data politics. Both articles were attempts by social scientists to grapple with the social and political implications of datafication. As Ruppert et al. put it:
…data has become a social and political issue not only because it concerns anyone who is connected to the Internet but also because it reconfigures relationships between states and citizens. (p.1)
Thinking politically about the impact of datafication has become even more urgent in the era of artificial intelligence, machine learning and algorithmic governance. But these ideas are not new, even if the fresh challenges from new technologies are. This kind of sociotechnical thinking about how humans and their technologies are mutually influential has existed since the dawn of the industrial age.
The politics of artifacts
Technologies are often presented as neutral artifacts, as objects designed to make our lives safer and more convenient and our work more efficient and rational. However, 40 years ago, the sociologist of technology Langdon Winner (1980) argued convincingly that technical artifacts are not neutral and have what he described as politics. He drew attention to three features of technological innovations:
- the inherent biases built into technology designs that frame social problems in particular ways
- unequal access to decision-making about how technologies are designed and implemented, and
- that technologies may be more compatible with social systems and values that benefit societal elites.
Winner’s (1980) arguments were not only about digital information technologies. He was interested in the social impact of any technological artifact, from a nuclear power station to a city transport system.
In one infamous example of the use of the built environment as a racist technology, Winner argued that the New York urban planner Robert Moses designed the height of bridges on the Long Island Parkway in New York in such a way as to make it difficult for public transport to use them, thereby effectively excluding Black, coloured and low-income people from accessing the new public parks and beaches commissioned by Moses whose racist views were well known.
That claim has since been contested, but Winner’s main point was that the design of high-impact technologies, especially infrastructure, was controlled by social elites who could, consciously or not, design them in ways that met their interests and not the rights and interests of other, especially historically marginalised, social groups. The disability movement has long understood and contested how the built environment excludes and marginalises them, denying them access to transport, buildings, and other social spaces.
Data politics
These physical separations, barriers and divisions with differential impacts on different kinds of people are mirrored in the sorting and classificatory logic of data-driven systems.
Once again, this is not an entirely new phenomenon. Governments have used data about their subjects for hundreds of years. Early parish records, police and court reports, and workhouse records were used to sort populations into categories for the purposes of social and economic ordering: for example, to manage and control the criminal and vagrant, ensure the legitimacy and property entitlements of children, or ascertain the tax burden of a household.
Between the late 19th and the early 21st century, the volume of data maintained on citizens mushroomed, fuelled by the information needs of burgeoning welfare bureaucracies and the growing capability of information systems to store, manage and manipulate citizen data.
Since the 20th and 21st centuries, the interest of state actors in citizen data has been increasingly interwoven with the interests of corporate actors to extract, capture and use data to create new products, find new markets and sell data services to citizens and governmental actors. As Bigo, Isin and Ruppert (2017) put it
There has never been a state, monarchy, kingdom, empire, government or corporation in history that has had command over such granular, immediate, varied, and detailed data about subjects and objects that concern them. (p.2)
Mainstream responses to the consequences of datafication focus on issues such as surveillance, privacy, anonymity and online incivility. However, the concept of data politics draws attention to broader sociological dimensions, including that contemporary data practices are creating new kinds of power relations and politics at different but interconnected scales, “…data and politics are inseparable. Data is not only shaping our social relations, preferences, and life chances but our very democracies” (Bigo et al. 2017, p2). We can refer to this phenomenon as data power.
The contentious politics of data
Data politics offers an account of how governmental and corporate actors use data to create and extend new kinds of power relations to govern and profit from data subjects. These top-down effects on citizens can be described as the institutional politics of data. On the other hand, the contentious politics of data concerns the bottom-up resistance of citizens and data activists, or “civil society, that is to say the realm of human activity outside the state and the market” (Beraldo & Milan, 2019, p.1). Put simply, we are talking about:
…bottom-up, transformative initiatives interfering with and/or hijacking dominant processes of datafication, contesting existing power relations or reappropriating data practices and infrastructure for purposes distinct from the intended. (Beraldo & Milan, 2019, p.1)
The idea of contentious politics is drawn from social movement theory. It is a framework that foregrounds the agency of citizens to contest and challenge societal elites.
No single definition or conceptualisation of the nature of social movements is accepted by all scholars. Not all social movements are engaged in contentious politics, and many forms of contentious politics are unrelated to social movements. However, the following definition is relevant to our discussion of the contentious politics of data:
Contentious politics involves interactions in which actors make claims bearing on other actors’ interests, leading to coordinated efforts on behalf of shared interests or programs, in which governments are involved as targets, initiators of claims, or third parties. Contentious politics thus brings together three familiar features of social life: contention, collective action, and politics. (Tilly & Tarrow, 2015, p.7)
One point of difference between the definition offered by Tilly and Tarrow and the discussion offered by Beraldo and Milan is that the latter extends the scope of the contentious politics of data to include both individual and collective forms of action. This is perhaps because of ways in which individual actions on the internet can be part of a form of networked individualism, where action is taken by individuals informed by practices shared online by others:
Think for instance of the adoption of encryption or the act of covering the laptop camera with a sticker as mechanisms of self-protection from the interference of states, corporations or malicious actors. (Beraldo & Milan, 2019, p.3)
Another significant idea from the framework of contentious politics is that of repertoires of contention, or “what people know they can do when they want to oppose a public decision they consider unjust or threatening” (Della Porta, 2013, p. 1081).
In traditional forms of action, these repertoires might include a combination of sit-ins, mass rallies, strikes, petitions, letter-writing campaigns, lobbying officials and so on. However, the use of the theatrical metaphor of repertoires is deliberately designed to draw our attention to how successful activists innovate and improvise new forms of action that are closely adapted to the issue at hand, local circumstances and the historical conjuncture. The Luddites breaking of machines and the Hong Kong pro-democracy protestors’ use of lasers to highlight and disorient facial recognition cameras are two examples of repertoires of contention.
Data as stakes and data as repertoires
In the context of the contentious politics of data, data activism can focus on data-as-stakes, where some aspect of datafication–such as predictive policing or the use of facial recognition cameras–is the issue with which activists are trying to contend, or data-as-repertoires, where data activists harness data as a tool to contend with an issue of concern–such as the use of citizen-led data collection of air quality, countermapping to map police violence, or data witnessing to document and challenge injustice and abuse.
Beraldo and Milan argue that their contentious politics of data complements the idea of data politics proposed by Ruppert et al. by emphasising the political agency of citizens as data activists to reactively challenge instances of datafication, or proactively harness data practices for the purposes of social change:
To put it simply, in virtue of their skills and/or awareness, data activists are able to “translate” data and data infrastructure into a site or means of struggle. In this sense, data activism, especially when it becomes a collective matter, constitutes an antidote to the atomism of many contemporary approaches to and narratives of datafication (Beraldo & Milan, 2019, p.5)
Biraldo and Milan go on to offer an analytical framework or typology of data activism mapped against two dimensions represented as continua.
One dimension represents examples of data activism that range from data-oriented activism or data-as-stakes to data-enabled activism or data-as-repertoire. The second continuum maps data activist practices from the individual to the collective, from individual acts of resistance–such as the use of encrypted communications–to larger scale mobilizations and everything in between. These analytical categories are ideal types and in empirical reality there are often overlaps, in the following diagram Biraldo and Milan (2019) include some examples of the use of the typology to classify instances of data activism.
Conclusion
Data politics explores the politics of datafication and its social and political implications. It highlights that data are not neutral but inherently political, and their use in technologies, algorithms and AI can reinforce unequal power relations and societal divisions.
However, the contentious politics of data, where citizens and data activists challenge and resist datafication processes, represents a bottom-up challenge to datafication, using data both as stakes to challenge issues related to datafication and as repertoires to promote social change. See the links below for some case study examples.
