Data Colonialism and Data Commons
This is a revised version of Felix Stalder’s presentation at the New Alphabet School#3 Coding at Goethe-Institut Max Mueller Bhavan, New Delhi on January 16, 2020.
I want to start with a qualification and a question. What follows is informed by my own provincial European, even Western-European experience. Of course, the social and cultural consequences of global IT are highly differentiated across geographic areas, and within them along economic, gender and racial dimensions. However, there are fundamental technological dynamics that are shared across these differences, not least because technology itself, machinery and industry, is highly globalized. More importantly, it might be precisely the experience of being provincial in a new geography and topology of power that could open-up new alliances.
The question I want to ask is this one: what if technology, in particularly IT, isnot a progressive force, empowering the edges, enhancing our collective knowledge of and mastery over the world, but rather a conservative force, in the sense that of empowering the already powerful, extending and deepening existing forms of oppression and exploitation, by creating new types of opacity, new areas of dispossession, and new types of dependencies. So, rather than moving us forward along the trajectory of democracy, it is leading us backwards towards the darkest periods of history.
This is, of course, not a new question, but one that doesn’t come easily to me. Part of my particular European perspective is the experience of the 1990s, when space expanded at a rate faster than anyone could fill it. It was not just the Berlin Wall that had come down, but also the on-rush of the globalization of civil society, based on modems, self-designed websites, which seemed to herald new freedoms and, in my our particular corners, also triggered a wave of practical and theoretical experimentation in order to make sense of this changing landscape. What made it such an exciting moment was the new experience of relatively informal trans-local collaboration, not the least with organizations like Sarai in Delhi, or the Alternative Law Forum in Bangalore, which were grappling with their own techno-political and economic transformations, at least as comprehensive as that which we experienced in the West. Not all was rosy, but everything seemed possible. The horizon was open and we were falling towards it.
These days are now long over now, we need to ask different questions, interrogating a rapidly closing horizon. If we see technology as a conservative force what, then, is being conserved? I would argue primarily three things.
First, central management’s ability to organize dynamic processes, even though these extend over greater and greater distances, involving ever greater numbers of flexibly organized elements. This applies equally to state as well as corporate actors.
Second, the ability to extract value through these forms of management, even though many traditional sources are being exhausted, by the law of falling profit rates, by the exhaustion of natural resources and by bumping against hard ecological limits to expansion. This is done by bringing ever new resources into the gambit of the system of extraction: “unconventional” fossil fuel deposits, the last pristine lands on the globe, and, what concerns me here in particular, ever more domains of life.
The third element that has been conserved is the ability to create “externalities”, that is, effects that are kept outside of the economic calculation and thus can seemingly be ignored, such as environmental degradation and social upheaval—even though the world is becoming ever more connected and integrated and the negative feedback loops caused by these externalities are becoming stronger and harder to ignore. Yet, the vortex of social media, its lack of differentiation between news and propaganda, the trivial and the profound, induces us to do exactly this: ignore what is increasingly visible, even to the naked eye, and replace it with a set of passionately-held, counterfactual beliefs (be that of religious, conspiratorial or other narrative patterns).
Innovation and Regression
If we accept this view, it raises the question: How has this been preserved on the basis of a new infrastructure that is clearly and deeply transformative? I would like to propose 5 vectors along which this double movement of innovation and regression has been achieved.
First, by scaling up: The new infrastructure enables new economies of scale. It’s possible, quite simply, to keep track of more things, across larger distances at higher speeds than ever before. It’s possible now to reach global scales without standardization, indeed, through “personalization” (ie, ultra-flexible standardization). And scale is very important, in a digital world more than before, because the cost of providing the infrastructure is extremely high, but each additional user or object within this infrastructure produces almost no additional costs.
Second, by increasing flexibility. It makes a huge difference whether information is stored in paper files that are hard to reorganize, or in databases where the recombination of data according to changing needs is easily accomplished. This “database revolution” has been going on for the last 50 years, at least, but it has only recently spread throughout society at large, and flexibility is still increasing as we move through different approaches to information management, from centralized to decentralized, and from structured (the “S” in SQL), to unstructured databases, which are currently in use by many big-data projects that combine very heterogeneous data sources.
In effect, this breaks, as Manuel Castells already noted in the 1990s, the connection between size and rigidity which plagued the standard bureaucratic institutions, to a degree that Max Weber saw them creating an “iron cage”. In consequence, large scale organization can now be so flexible that they are able to route around any local obstacle. Strikes are no longer a powerful weapon in the hands of labor. We have seen this multiple time in recent years in Europe, with strikes in Amazon warehouses in Italy and Germany having no discernible effect on their logistic efficiency. We can detect echoes here of the 1990s dream of a censorship-resistant Internet, though in a very different social form.
Third, by speeding up all processes, making them harder and harder to understand, let alone shape from the outside, without access to the infrastructure on which they run. While there is, and has to be, long-term planning on the infrastructural level, society as whole is put on “real-time”, an ever extending, ever receding present, which allows no time to reflect and organize.
Fourth, by pervasive quantification. Accordingly, in order to be manageable, everything needs to be quantified. Everything that is not quantified becomes an externality, that is, something that can be ignored. This is both a continuation of and a break from processes that reach to back to the 18th century. Remember, when the term statistics was introduced by the German administrator Gottfried Achenwall in 1749, it meant “Science of Data about State”, and it was quickly established as the key way in which the state could perceive and manage the world. So, there is a long history of quantification and management, but now it has becoming so hegemonic that it penetrates all areas of life, including the most intimate dimensions of our private life and our corporeal existence. It is crowding out other ways of being and seeing, erasing other knowledges of the world and of ourselves. Distributed personal and cultural experience is devaluated, centralized ways of knowing elevated.
And, fifth, by moving governance from norms to protocols. This started in the 1970s, with privatization and marketization, where the protocols of the markets were given precedence over the normativity of the law and democratic debate. Like quantification, it has spread into all areas of life, most obviously, through social media, where technical and social protocols determine everything, without any regard for what used to be called social meaning, for example, the difference between fact and fiction, which was always entangled, but now its difference has been erased.
In sum, these mechanisms creates an enormous increase in complexity, opacity and intensity of social processes on all scales, while conserving essential dimensions of previous systems based on formal hierarchies.
The big question is now: Are these only a quantitative shift, a general acceleration and intensification, or is something more fundamental going on? I would argue the latter, that we are not just witnessing an intensification of capitalism, but a new mode of production and governance that can justifiably be called “digital” or “data” colonialism.
(Data) Colonialism
For present purposes, I want to understand colonialism as an integrated system with at least three layers.
Economy: The unilateral conquest of land, its resources and bodies, in order to extract maximum value from them, without regard for the consequences for the land or the bodies seized.
Governance: The administration of people as things. Historically, this has happened at the very moment when there has been a great expansion of individuality and subjectivity within the culture of the colonialists. And, in many ways, we are witnessing this contradictory movement again. Some subjects are reduced, others are elevated in their existential capacities.
Ideology: It claims to bring progress (often in forms of techniques and technologies), science, peace, and civilization to people and cultures, so profoundly unable to help themselves that their opinions are entirely irrelevant and do not even need to be considered. It is a self-validating ideology in which any resistance is interpreted as yet another sign of backwardness, and thus justifies any force to overcome it.
In digital colonialism, the land that is dominated is the internet infrastructure on which much of our individual and collective lives unfold. And the body is not enslaved, but dividualized, that is, turned into limited and changing elements of the body, endlessly divided and then reassembled into any form that can be turned to data value. The focus is not on the ability of the body to do physical work, though that is kept ready at hand (through the gig economy) and placed under the shadow of being replaced by machines and robots. Rather, the focus is on behavioral and corporeal data which comes to stand it for the brain, which is the real target.
The mode of governance is a new form of behaviorism, where our thoughts and opinions are ignored and only behavior measurable by a third party counts. In the process, that third party is taken to gain knowledge superior to first person experience, which is deemed incomplete and unreliable. Hence resistance is considered, once again, to be just as a sign of ignorance.
The ideology is that of connectivity, empowerment and full-spectrum efficiency. There is no process that cannot be improved by digital means.
Within contemporary debates, two perspectives on digital colonialism can be distinguished. In most contexts, both of them are at play, in different measures.
Data Colonialism as an external regime
Data colonialism can be understood as being perpetrated by foreign powers, in which the building of ports, roads and railway lines in the name of the colonial project has been superseded by the building of digital infrastructures, from undersea cable to satellites to platforms (like Facebook) or service providers (like Google). It is no coincidence that they often mirror the older colonial geographies. In India, as maps show clearly, the main entry points for international Internet connectivity are Mumbai und Chennai, two of the oldest colonial ports established by the Portuguese in the early 16th century and then later used by the British. Facebook’s “Basic” Internet—free Internet access controlled by Facebook and limited to its platforms—is perhaps the most glaring example of such a new colonial infrastructure, indeed, its colonial character was so obvious that it was rejected in India in 2016. But similar examples can be found further down and further up the internet stack and are introduced by different means. Very widespread is, for example, a practice known as “zero rating”, that is, the mobile carrier includes certain types of data, for example, the messenger WhatsApp, in their package, making all other connectivity subject to an extra charge, often unavailable to large parts of the population.
In this perspective, the colonial regimes, created by and for the benefit of foreigners, have been transformed and extended by new ones. Google and Facebook and the like, then, act like the 21st century version of the East India Company, complemented and supported by the six eyes alliance of secret services and increasingly challenged for colonial dominance by Chinese companies, such as Huawei, Tencent or Alibaba, also with full state support.
But there is a second perspective, one that interest me more here.
Data Colonialism as an internal regime
Digital colonialism is not just oriented outwards, but increasingly also inwards, as Nick Couldry and Ulisses Mejias have recently argued. Here, the focus is on the application of colonial economic, governmental, and ideological principles, not by foreign powers, but by the country’s own elites within the framework of national politics, often relying on the services of the globally dominant data colonialists.
The economic model that is applied in a wide variety of contexts is the extraction of valuable resources without the consent of the communities, and to the detriment of the communities. This can be done directly when new middlemen and infrastructures extract large amounts of value, driving local producers into ever more precarious existences. While some of this is extracted by foreign companies (Uber extracts 25% of all revenue from taxi services globally, a business that used to be entirely local), the “anti-colonial” answer is often simply to build national champions which operate on the same principles, often in close proximity to the government which thus acquires access to much of this data, as is the case not just in China.
This can be done indirectly, when extraction, profiling and segmentation is being built into any business model, which can then be turned against those communities, through manipulation, further exploitation (e.g. dynamic pricing) and monopoly rents.
Governance proceed along two lines. First, through the classification of people by third parties based on the interests of these parties, without regard to people’s subjectivity. The goal here is not a fixed racial system, though it builds on it, but new forms of hyper-flexible partitioning, depending on the changing political and business strategies of the parties controlling the processes. Second, through the application of rules and decisions without recourse or explanation, undermining due process, the rule of law and the presumption of innocence.This is justified by the argument that the colonialists know what’s best for the locals, rather than the locals themselves. This ranges from secret anti-terrorism procedures, implemented in the name of security, that are neither explained nor can be contested, to secret public-private partnerships whose conditions cannot be disclosed in the name of efficiency, to the algorithmic decision-making that is unquestionable in the name of complexity, yet justified by the well-known claim of Google and the like that they know us better than we know ourselves.
And there is the ideological claim of modernism and efficiency, which are often quite credible, given the growing difficulty of established bureaucratic methods in coping with rising complexity.
The Erosion of Democracy
The most severe problem, as in all colonial regimes, stems from massive power asymmetries. In the case of data, they are harder to see, because they arise not only from the collection of data, but primarily from its aggregation and analysis. If data is the raw material, it is the aggregation and analysis that turns it into commodities and power. And that power is created and exercised to benefit the colonizers at the expense of the colonized to such a degree that they destroy local cultures and environments. By local culture, speaking from my provincial European perspective, I mean, basically, democracy.
That is because democracy depends on the assumption that, first, the world is readable, and thus an informed opinion can be held, and second, that it is subjective personal judgment that counts, not objectified human behavior. In theory, voting expresses how people see the world and what they want of it. In the world of data behaviorism, all that counts is to get people to make their mark at the right place, since the thinking and intentions behind it are irrelevant. So every trick, every form of manipulation, is permissible.
This is, in a way, logical. Democracy was never meant for the colonialized, and in the leading circles of the IT industry, particularly in Silicon Valley, there is an ever more explicit distance to, or outright hostility towards it. Over the last decade, their libertarianism has hardened considerably, resulting in a racialized elitism where only a limited number of people, most of them technically educated men, are seen as capable of governing. And for the rest, they are in no position to ask too many questions about the seemingly free services on which they rely to organize their precarious lives.
The case for data commons
Is there anything that can be done about it? There are numerous ways to think about resisting this development. One would be to sharply limit the reach of the colonial system. To defend and create new areas of life that lie outside this drive to quantify, to create zones of human interaction that are structured according to different ways of being in the world together.
However, data regimes are hegemonic for a reason, not least because they are useful at organizing current social realities. The question is, how can we turn data into a resource that helps us to renew democracy and address the complexities of the current situation that requires new and smart approaches to steer dynamic and complex systems?
At the moment, the basic economic principle is that whoever collects the data, owns the data. The only limitations, weak as they are, center around notions of privacy. We need to turn this around. Those who produce the data, should own it, not those who collect it. Yet, the perspective of privacy, which governs much of how we, particularly in Europe, think about data, doesn’t help here. On the contrary, it’s highly problematic as it ties data closely to individuals by making it personal data.
However, most data is not personal, but relational and transactional. Thus it cannot be connected to a single individual. In practice, it means that it’s free to collect and own by whomever is capable of doing so. In order to challenge this we need to think about different types of ownership. The conventional response to the problem of proprietary data was to make it freely available as “open data”. But this is not a one-size fits all solution, because power arises from aggregation and analysis, which is closed even when performed on “open data”. So, open data can easily feed into data colonialism, although not exclusively.
In many cases, the unit of ownership should be the community. And, in many cases, the better framework is not private property or public goods, but the commons. A resource that is managed collectively, for and by the community, with the possibility of selling access to outsiders, if desirable.
In order to have collective ownership, one has to define the collective, which in many cases will mean localizing data, and restricting its use to certain local domains. There are already attempts to preempt that. For example, the WTO aims to expand free trade under the name of the “free flow of data”. This needs to be resisted, as the free flow of data really means the free extraction of data. And this is being resisted. The government of India, for example, recently refused to sign a declaration on the free flow of data at a meeting of the WTO. The European Union is, very cautiously, adopting frameworks for “data sharing”, at the moment in the energy and the mobility sector. There is a growing awareness that data is an infrastructure for all kinds of known and yet-unknown uses and that, like other infrastructures, access to it should be non-discriminatory, and certain uses should be prohibited.
But we cannot leave this solely to policy makers. We also need to develop a new language, or perhaps even a New Alphabet, if you will, and an aesthetics for the new data driven, collective self-consciousness, so that we can replace the dystopian image presented above. Here there might be a point of shared interest between progressive forces in Europe and developing countries, because in this case, they are fighting a common enemy: digital colonialism, both as a force from the outside, but also increasingly as an economic and governmental regime developed and applied internally.
