From Surveillance Capitalism to Data Colonialism, recent scholarship has characterised processes of extraction and exchange of personal data as a form of dispossession. Many of these works draw on David Harvey’s highly influential, critical reworking of Marx’s account of primitive accumulation: accumulation by dispossession (AbD). Harvey attempts to explain how contemporary capitalist accumulation takes place through predatory processes of privatisation, financialization, fiscal austerity and manipulation of crises. Crucially, he stresses the continual nature of dispossession under capitalism as opposed to Marx’s supposedly originary version.
In their works on data colonialism, Couldry and Mejias describe the arrival of a new social order in which everyday individual and collective life is converted into a data stream to be annexed to capital. Ongoing dispossession is central to their thesis:
“A continuously trackable life is a dispossessed life, no matter how one looks at it. Recognizing this dispossession is the start of resistance to data colonialism.”
Similarly, for Zuboff in Surveillance Capitalism, data is a new type of fictitious commodity fashioned out of human experience – a “raw” material used to produce tradeable prediction products in a process of continual “digital dispossession.” These texts (amongst others) use Harvey’s formulation as their starting point for theorising how personal data comes to be subsumed to capital. Though this understanding of dispossession may appear an apt theoretical lens for studying how this data is extracted and commodified, it leaves unaddressed many important questions about the targets, objects and operations of data dispossession.
Dispossession works through difference
As has been more fully argued elsewhere, AbD, like primitive accumulation, is often treated in ways which do not account for how and why some people and spaces are more expropriatable than others. In the context of data-driven technologies, our corrective to this could begin with further critical inquiry into how dispossession of data works through regimes of racial difference. As Ruha Benjamin shows, we ought to look not just at the racist impacts of data-driven technologies but at the social inputs which expose racialized bodies to extractive observation and make anti-Black surveillant technologies appear desirable and inevitable in the first place.
Scholars including Brenna Bhandar and Davina Bhandar remind us of the need to consider dispossession as being in a dialectical relationship with (self)possession. The very identity of the (white) bourgeois possessive individual is founded on their capacity, and entitlement, to own property. Its corollary, racialized exclusion from liberal personhood, positions some subjects as less capable of owning and appropriating property, and of self-possession and autonomy. Applying these insights to contemporary data practices may help us to understand why some people are not figured as the self-possessing subjects of data but as its objects; why, for example, people subjected to immigration control do not enjoy equal rights under UK data protection law. Our experiences of personhood, and its relationship to data, are not universal.
Data is not an analogue of other objects
Evgeny Morozov has argued that it is wrong to use “dispossession” to conceptualise the extraction of individual behavioural data:
“The fact that Google, after a search query, knows that I like avocado toast does not mean that I myself have now forgotten that I like it. To posit that this is the same type of “dispossession” as the sort that involves someone coming along and physically removing avocado toast from my plate is simply wrong.”
Taking this observation further, we might add that he will also lack awareness of the scale and scope of the data processing taking place. But while it is true that data’s representational quality makes them virtually inexhaustible and endlessly reproducible, one does not have to be wholly deprived of an object, or aware that it is gone, to experience loss. So what is really lost when these kinds of data – biographical, behavioural and biometric – are expropriated?
Many of the works cited so far proceed by analogy, casting data as a type of raw material akin to so-called natural resources derived from land. As Indigenous scholars have shown, however, objects of dispossession are not commensurate, nor are processes of dispossession universal. In particular, land is important in ways that are difficult to contain and measure, as Byrd et al. explain:
The loss of land is not just a loss of property, territoriality, power, nation, or sovereignty; it is the loss of those philosophies that derive from the relationships the land itself activates, fosters, and nourishes.(p. 15)
Like land, we cannot uncover what is at stake in the dispossession of data by reducing it to, and valuing it as, an object of property. In extrapolating from prior instances of dispossession, we may even risk replicating the very colonial capitalist rationality of commensurability which constitutes land, people and social relations as translatable into value form.
Thinking of data on their own terms, we see they are in many ways unlike other objects of dispossession. Whereas proletarianization processes separate workers from their land, compelling them to sell their labour for wages in order to survive, data are not generally required to subsist. Unlike many other objects of commodification (like knowledge or labour), data are already abstractions. Dispossession of data might be better understood as encompassing (amongst other things) loss of integrity, autonomy, dignity, identity, temporality and relationality.
Though often overlooked in discussions about digital data, colonial processes of dispossession (of land, materials and bodies), exploitation, displacement and cultural defilement are not confined to the past but continue in the present day. To empirically grasp the nature of dispossession of digital data, we should approach it not as entirely separate from other raced, gendered and classed modalities of power but as articulating with them.
Dispossession is recursive
Recent work by Robert Nichols has demonstrated how dispossession involves not just the transfer of property but also, simultaneously, the process of recursively transforming an object into property. Though it is often treated as such, dispossession is more than theft, confiscation or unjust enrichment. As suggested by existing theories of primitive accumulation and AbD, acts of dispossession also shape social relations (particularly property relations). Examining these relations, Nichols offers a wholesale critical reconstruction of dispossession in which he shows that it consists in a recursive transformation of nonproprietary relations into proprietary ones which, at the same time, systematically transfers control and title of this (newly formed) property: “In this way, dispossession merges commodification (or, perhaps more accurately, “propertization”) and theft into one moment” (p.8). Simply put, dispossession is not just the transfer of property but is also, simultaneously, the process of transforming an object into property. This process is also accompanied by a retroactive attribution of ownership in which the dispossessed are figured as the “original owners” only after the act of dispossession: “In effect, the dispossessed come to “have” something they cannot use, except by alienating it to another” (p.8).
By drawing inspiration from Nichols’ theory of dispossession as recursive, we can examine not just the process whereby data might be rendered, through dispossession, as property, but what kind of new property relations it sets in motion. Beyond the mere fact of something being “property” lies a multitude of possible proprietary logics and relations. In a legal doctrinal sense, the exact meaning of property is not necessarily fixed across legal scenarios, and the ideas and claims currently emerging about data as a kind of intangible property remain unsettled. As Fourcade and Klutz have argued, the commodification of data is strengthened via a sleight of hand which makes us feel “cheated” of our data and therefore owed remuneration and more willing to see it as property. We ought, however, to further interrogate this new-found sense of entitlement. In her pathbreaking essay Whiteness as Property, Cheryl Harris argues that whiteness shares with property, “a common premise – a conceptual nucleus – of a right to exclude” (p. 1714). This calls for alertness to the potential for any possessive individualist impulse set in motion by data dispossession to produce new exclusions, and to differentially assign and appropriate value.
Against claims that we should be able to market and “profit” from our personal data, Salomé Viljoen has persuasively rejected the data-as-personal-property paradigm. The power, wealth and informational asymmetries between individuals and monopolising corporations would no doubt be a recipe for gross exploitation and widening inequalities. In addition to this, as we have seen, we should be wary of how such individualising approaches, particularly those which posit personal data as private property, might also enact (and disavow) dispossessory power. By building our understanding of how it operates, and what is at stake, we are better placed to resist the dispossession of data and to build more democratic alternatives.
Catriona Gray is a PhD student at the University of Bath. Her research examines the forces shaping the adoption of AI technologies within the global refugee regime. You can contact her about her work at firstname.lastname@example.org or on Twitter @CatrionaLGray