Digital Serfs

About: Cédric Durand, Techno-féodalisme. Critique de l’économie numérique, Zones

For several years, numerous books and documentaries have explored the fears sparked by the rise of the digital economy: the leading digital corporations – commonly referred to as the GAFAM, for Google, Apple, Facebook, Amazon and Microsoft – are regularly accused of abusing their dominant position; the algorithms developed by these corporations are often viewed as intrusions into the private lives of their consumers, as was made evident by the 2018 scandal surrounding Facebook and the company Cambridge Analytica, which was accused of having played a central role in the Brexit referendum.

In his latest book, Cédric Durand explores the causes and consequences of the transformations sparked by the rise of the digital economy. The book opens with the story of an arrest in the 1980s, that of the designer of a sci-fi game called “Cyberpunk”, in which corporations have become more powerful than the state, and individuals have become entirely dependent on these corporations – two characteristics that reflect what the author calls the techno-feudal scenario. Hence the author’s question: what if, a few decades later, fiction and reality ended up progressively converging? On the one hand, digital corporations – which started out as simple, highly innovative start-ups – have managed to progressively become giants, thus allowing them to become monopolistic rentiers. On the other hand, by developing increasingly sophisticated algorithms, digital platforms have contributed to guiding or even predicting individuals’ choices, creating a relationship of dependency (both on the part of consumers and on that of producers) that resembles a new form of feudalism. This is Cédric Durand’s central argument.

Digital Transformations and Competitive Dynamism

The IT industry underwent a particularly spectacular rise in the 1990s, a period which saw the emergence of the “new economy” based on the dissemination of new information and communication technologies. According to the author, this course of events was very much in line with the Silicon Valley model, the ideology of which appeared in the wake of the 1960s counter-cultural movement, and aimed to promote a new spirit of new Californian entrepreneurs keen to stimulate individual autonomy through new technologies. This ideology then spread widely throughout certain circles of thought, and ended up convincing a large section of the American political class, thus giving rise to the “Silicon Valley consensus” (which replaced the Washington consensus that had been dominant during the 1980s): it was first supported by a conservative Republican fringe, such as the pro-market Progress and Freedom Foundation think tank, which formed an alliance with Silicon Valley stakeholders who had emerged out of the counter-cultural movement, such as Stewart Brand. In the 1990s, this strategy then inspired the new growth strategy promoted by the Clinton administration.

This new growth strategy – which broke with the previous strategy that had led to post-war industrial growth – relied on a Schumpeter-inspired process of “creative destruction”: replacing less efficient corporations with particularly innovative and high-performance new corporations becomes a source of economic growth. Such a dynamic also requires increased flexibility across all markets (product, labor and capital markets), with the aim of increasing the overall productivity of the factors of production in the context of a strong opening of commercial exchanges at the international level. In this new growth strategy, the state is responsible for promoting an environment that will be favorable to this competitive dynamism, by defining clear and enforceable property rights (which guarantee the owners of merchandises that they will be protected in case of dispute).

In the author’s view, the new capitalism based on this strong entrepreneurial dynamism is nevertheless inclined to numerous paradoxes. Indeed, this new dynamism has not translated into renewed GDP growth, even allowing for the argument that it is supposedly difficult to precisely measure productivity (which takes us back, for example, to the famous “Solow paradox” [1]). On top of this comes the delicate question of the generalization of this model of growth, given how specific the Silicon Valley model appears to be: for example, the Lisbon strategy that was promoted by European countries from the year 2000 – and which aimed to turn the European Union into the most dynamic and competitive economy by 2010 –was not as successful as hoped, since the ratio of European countries’ GDPs compared to global GDP has not stopped decreasing since the 1980s. Even worse, the model promoted by the Lisbon strategy – which aims to foster free competition – has led to a decrease in public funding for research, even though it was in fact the American state that played an essential role in the development of Silicon Valley, as has been shown for example in the work of Mariana Mazzucato (who has underlined the importance of public spending, in particular in the context of Apple’s success). In line with these works, Cédric Durand stresses that “the United States’ position as a leader is anything but the result of spontaneous market forces, including as far as Silicon Valley is concerned” (p. 81). Finally, all of this raises the question of what dynamic this strategy takes over time: digital corporations that were merely particularly innovative start-ups in their early days have had to progressively grow in order to survive. As a result, the strong entrepreneurial dynamism that characterized the 1980s and 1990s initially created a situation of strong competitive dynamism, but led from the 2000s to the constitution of new monopolies that are now characteristic of large swathes of the digital industry.

Going even further, the progressive metamorphosis of small, highly innovative start-ups into digital giants can apparently be explained by the nature of the main assets on which the activity of these firms depend: namely, intangible assets. These assets are based on the totality of information about users that is mass-collected (in the form of Big Data) by digital platforms. Indeed, unlike physical assets, intangible assets are mainly characterized by being infinitely extendible, thus giving rise to increasing returns to scale. These returns to scale allow digital corporations to capture what the author calls a “differential rent from intangibles”, to which can be added other rents: rent connected to holding patents (thus giving strategic relevance to legislation governing intellectual property), rent from natural monopolies (connected to network complementarities, economies of scale and non-recoverable investments) as well as rent from dynamic innovation (connected to the control of information systems).

Big Data: what are the consequences for social relations beyond simple individual behavior?

The collection and processing of mass digital data (also known as Big Data) as intangible assets forming the basis for the activities of the major digital corporation have thus become a strategic issue. Unlike marketing or publicity, which aim to influence consumers’ behavior, the new digital model – as seen in the case of the Amazon website or the Netflix platform – aims to guide consumers’ choices by providing them with numerous suggestions. These suggestions are made possible by powerful algorithms capable of precisely analyzing our every digital coming and going. The issue here is no longer simply to control uncertainty – a central theme, in particular in Keynesian thought – but rather to take advantage of the degree of predictability of human behavior: more than this, “the horizon of capitalism is not to increase the predictability of behaviors, but well and truly to pilot them” (p. 114). As Shoshana Zuboff has argued, the emergence of “surveillance capitalism” raises a host of ethical and political questions, in particular in terms of the protection of consumers’ private lives. This great risk of intrusion into consumers’ private lives however comes into conflict with the expectations of corporations whose growth depends on collecting these very same personal data.

However, Cédric Durand goes beyond the atomistic view put forward by Zuboff: he is interested in exploring the effects of this “sea of data” (p. 124) in its collective dimension. Algorithms affect more than the simple autonomy of individuals: they also seem to modify social relations, by creating a real dependency between digital platforms and their users. The Uber platform provides a good illustration of this dependency: the chauffeured passenger car drivers who use this platform are, it is true, deemed to be independent workers in legal terms, which at first view suggests that they are not involved in any relationship of subordination. And yet, they are granted no autonomy in the choice of key parameters such as their rates (and therefore their wages), or the duration of their rides. In addition, this dependency can also translate into a form of social control that becomes automatized through algorithms. Online platforms often ask their users to rate the various services they may have used: these rating systems aim to objectify trust through a “system of reputation based on the mutual evaluation of the agents involved” (p. 139) – a system which then has long-term effects on the relevant parties. This social control can also take on a much more extreme form such as that of the “social credit” system set up by the Chinese authorities, which uses the principles of surveillance capitalism: setting up a surveillance apparatus with the aim of piloting individual behavior.

Towards a New Form of Feudalism?

According to the author, due to the two fundamental characteristics outlined above, the digital economy can be viewed as a new form of feudalism, with reference to one of the modes of production identified by Marx. In this mode of production, feudal lords, in their capacity as landowners, capture a monopolistic rent connected to the land. Workers are not free, even though they are independent. Finally, the extraction of the surplus is not consubstantial with production. In this sense, feudalism can be distinguished from other modes of production such as slavery and capitalism.

For a start, digital corporations – just like feudal lords – are characterized by their ability to extract a rent thanks to economies of scale and network complementarities. As the author points out, “the reference to feudalism points to the rentier – i.e. non-productive – nature of the system of value capture” (p. 221). In addition, the digital economy is characterized by the way it creates a relationship of dependency, both for consumers and producers, such that it can be difficult to extract oneself from the digital platforms, with the first risking being socially marginalized, and the second risking being excluded from the market. Hence the analogy with feudal organization: “this situation of dependency of subaltern subjects on the digital glebe is essential, since it determines the capacity of the dominant class to capture the economic surplus” (p. 218).

We can thus wonder about the viability of this “techno-feudal” model – a question that the author does not pose explicitly – and whether the analogy with feudalism can help us answer this question. Historians have attempted to analyze the sociopolitical causes of the crisis in feudalism. Among these analyses, the Marxist historian Robert Brenner focused on the extortion of the surplus by the lords, and the unproductive nature of this surplus: on the one hand, the excessive taxes collected by lords on their peasants’ income reduced the latter’s ability to invest in order to guarantee the good quality of the land they were working; on the other hand, this surplus was mainly spent by the lords on military ventures or ostentatious consumption. Is then, by analogy, the crisis of this “techno-feudal” model inevitable? In which case, what of the attitude of regulators as regards the upholding of current monopolies, but also as concerns the protection of personal data?