How labour could resist and defeat imperialism
Michael Roberts is an Economist in the City of London and a prolific blogger.
Cross-posted from Michael Roberts’ blog
Historical Materialism is an international Marxist academic conference held every November in London. This year, a whopping 370 panels were organized with over 1,600 registrants, the largest ever. As you can see in the photo below (the main session on the last day of the conference), young scholars participated in large numbers.
The title of this year’s conference was ‘Planetary Warscapes Ablaze: Imperial Mutations in Infernal Times’, which I took to mean that, under modern imperialism in the 21st century, there is increasing risk of more war and widening military conflicts. So, at HM, there were two main plenaries on the nature of fascism and on how to organise against imperialism.
As usual, every year I comment on the ideas and debates that arose from the many sessions at HM. And as usual, it is impossible to cover anywhere near the huge number of panels over the four days of the conference. I can only discuss a tiny proportion of these.
Naturally, I shall start with the sessions that I participated in. The first was a panel, billed as Technofeudalism: rents or profits? This was organised by myself and American tech workers, AK Norris and Tavo Espinosa, who have been developing insightful new research on open source software and AI. My presentation was a broad brush attempt to demolish the trendy concept of technofeudalism, namely that the tech giants, the ‘Magnificent Seven’ companies, obtain their revenues not from profits, or from the exploitation of human labour, but from rents through their monopoly control of cloud services and intellectual property rights. The most prominent of the technofeudalists is Yanis Varoufakis who has declared that ‘capitalism’ as Marxists analyse it, “is dead” and a form of feudal extraction of rents has replaced it.
There are several Marxist economists who also promote this theory of tech monopoly feudalism, for example, Cedric Durand who argues that capitalism now operates “through processes of predation largely disconnected from productive activities—the rationale of surplus appropriation in the techno-feudal hypothesis.” In my presentation, I tried to refute this theory, arguing that, in Marxist theory at least, rent is just part of surplus value and so ultimately must be a product of the exploitation of human labour. If there is no capitalist production, then there is no rent in modern economies. Indeed, if peasants or serfs do not work the land, there is no rent in feudal economies.
How can the technofeudalists and rentierists argue that there are no workers involved in producing the revenues of the tech giants, but instead only ‘cloud serfs’ (i.e. working for nothing?). Knowledge commodities (software, AI etc) are produced in capitalist production in just the same way as physical commodities; and the ‘mental labour’ of knowledge producers creates value in intangible commodities just as much as manual labour does for the value contained in tangible commodities. Both are material and value producing.
Moreover, as the other panellists showed in their research, the tech companies make money from selling products for profit (Amazon: data storage and cloud services, Apple and Microsoft: phones, computers and software, Meta and Google: advertising). All these services employ human labour (tech workers). So revenues are not the result of rent extraction from capitalist enterprises. And even if you were accept income from IPR and cloud services are rents not profits, they do not constitute more than 10-20% of all income in the US economy.
Source: IMF
Does it matter whether it is called profits or rent? Yes, because the claim that the tech giants mainly get rents like landords suggests that they are a different class to the rest of capitalist producers. So the tech giants become the enemy, like finance capitalists, rather than capital as a whole.
Fellow panellists tech workers AK Norris and Tavo Espinosa brought new insights into the nature of software production in their presentations. In particular, they showed the importance of free open-source software (FOSS). FOSS spans various forms of productive activity, from developing new software to ongoing maintenance work. As a result, FOSS is an important counteracting factor that mitigates the tendential fall in the average rate of profit in tech and in the wider economy, both from increasing the rate of surplus value out of tech labour and by sharply reducing the cost of elements of constant capital. Given that open-source software is free, Big Tech, as the main users of FOSS, can extract surplus profits (not rents) above the average.
FOSS emerged from heavy federal funding in the 1950s–1980s, then expanded through the free software movement of the 1980s–1990s, and on to corporate open-source adoption post-2001. Each phase corresponded to shifts in the rate of profit in the nonfinancial corporate sector, suggesting that FOSS development is neither incidental nor coincidental to capitalist accumulation in the digital age. Rather, it has enabled continued profitability amid intensifying capital concentration.
The Rate of Profit in the US Nonfinancial Corporate Sector and Phases of Software Development, 1950–2025
There were many other sessions where the question of ‘technofeudalism’ came up. Ryuji Sasaki of Rikkyo University, Japan re-presented his argument that he had made at a previous HM that late Marx regarded natural monopolies in the sphere of infrastructures as generating surplus profits that escape the equalization of profit rates and thus become fixed as ‘rent’, a category that cannot be reduced to the mere redistribution of surplus value produced elsewhere. “It is only by identifying rent in this way that we can grasp the parasitic and destructive character of rentier capitalism as an extractive regime that preys upon the common.” Sasaki seemed to be rehashing Ernest Mandel’s 1970s theory of ‘technological rents’, ie surplus profits that get turned into rents through monopolisation, but now adding that rents had become the major counteracting factor to the falling rate of profit (in contrast to the explanation of surplus profits in the tech sector offered by Norris and Espinosa in our session). Sasaki’s blurring of the clear distinction that Marx makes between rent as the result of the monopoly ownership of immovable land or property and profits as the result of capitalist production, I found unnecessary and confusing.
In another session called The Rulers. Corporate Power in the Age of AI and the Cloud, leading proponents of some form of technofeudalism like Celilia Rikap and Jodi Dean took part. While Dean rejected Varoufakis’ view that capitalism is dead, she still argued that: “Profit, productivity gains, and competitive advantages no longer determine accumulation strategies. Instead, rent, destruction, and accumulation combine with extra-economic coercion in a social formation driven by privilege and dependency.” I think this means that capitalism no longer exploits labour for profit, but instead engages in extraction of rents and uses force to rule.
There were many more sessions and papers on developments in modern imperialism than on technofeudalism, not surprising considering the Trumpist era of tariffs, sanctions, immigration and wars globally. The one question that dominated was whether US imperialism was in decline and liable to lose hegemony or not. In a session called The Dollar Empire: World Money and American Power, Steve Maher and Scott Aquanno argued that, although Trump has rattled the American empire, its foundations have proven remarkably resilient. The dollar’s position has been sustained by the ongoing strength of the US economy and the capacity of the Federal Reserve to act as a global central bank. The dollar remains the overwhelmingly dominant reserve currency and still rules in financial markets and in global lending. Indeed, dollar use in private transactions is actually increasing. The dollar’s role is thus institutionally and historically embedded in the global financial system. I think this is broadly correct.
In the same session Sean Kenji Starrs similarly argued that the US is more globally dominant than ever, but because of (not despite) China’s “capitalist rise”. This reason I find confusing. Surely, China’s rise in the first three decades of the 21st century threatens US hegemony and now drives US imperialist strategy with tariffs, bans, sanctions and military spending? Gianmaria Brunazzi and Cristina Re also re-presented their argument that US dominance is fostered by running current account deficits that force peripheral economies to lend and invest in dollar assets. So the massive level of US foreign liabilities are not a sign of weakness but of US imperialist strength. This argument seems back to front to me. But I shall leave that argument for now, as I have been kindly invited by Brunnazzi and Re to debate this argument at an online session organised by the University of Milan later this month. You can read my initial arguments on this thesis, here.
I think I prefer the characterisation of US imperialist hegemony offered by Costas Lapatvitsas in another session where he argued that US imperilaism is no longer totally dominant, but neither it is about to give way to rivals, whether from the BRICS, China or Asia and Europe. Instead, US monetary dominance persists despite the erosion of its productive base, producing a structural impasse, an interregnum. “This resulting interregnum, characterized by sanctions, reserve seizures, and proxy wars, is not a temporary disturbance, but part of a deeper imperial conflict whose resolution cannot be assumed peacefully.”
And of course, there were many sessions on the impact of AI on workers, capital and the environment. I participated in a book launch for Algorithmic Capital, Accumulation, Power and Resistance in the Age of AI, written by Jonathan Martineau and Jonathan Folco. The book argues that AI and algorithms are not just new technologies, but they are shaping the very nature of capitalist productuon and social norms. That requires new paths in organizing and control by workers. In another session, Alexandros Minotakis highlighted that the ‘Taylorism’ of exploiting factory labour in the 19th and 20th century was now being repeated through algorithmic management in the 21st century. Surveillance and old and new forms of coercion persist but increasingly have become the dominant managerial practices within the capitalist production process. It was no surprise then that Annie McClanahan won the 2026 Isaac and Tamara Deutscher Memorial Prize for her book, “Beneath the Wage: Tips, Tasks, and Gigs in the Age of Service Work.”
I also participated in a session organised by Socialist Register, an American Marxist journal. Its 2026 annual edition was entitled Late Stage Capitalism, in which I wrote one chapter called Capitalism in the 2020s and beyond. One co-editor of the SR is Steve Maher and Scott Aquanno was another speaker. Steve and Scott again presented their argument that US capitalism is not declining, but is still dominating with high levels of profitablity of capital and faster growth than other capitalist economies – of course all at the expense of American workers.
At the Socialist Register session on Late stage capitalism (Roberts,Maher, Aquanno and Meadway).
This view has got some empirical support. At HM, Bill Jefferies of SOAS presented some intriguing new empirical research. Jefferies kindly sent me his paper, as yet unpublished. In examining the corporate accounts of the top 100 US companies and in particular, the so-called Magnificent Seven technology giants — Amazon, Tesla, Nvidia, Apple, Meta, Alphabet, and Microsoft. Jefferies found that ”their very high rates of profit have driven the rising trend in overall US profitability since the 2008–2010 Great Recession, bringing it to, or near, historic highs.” Jefferies contrasts his approach to the usual Marxist estimates of profitability that are based on national accounts, which he considers are false because they overestimate capital stock levels. I have previously rejected Jefferies’ claim before, here.
But that does not mean he is wrong about the rising profitability of US capital, particularly since the end of the pandemic slump of 2020. In previous posts on my blog, I have shown that US corporate gross profit margins (the ratio of profits to sales) have risen to 19.4%—the highest recorded in the US since the 1940s! In contrast, labour’s share of US national income has plummeted to a new low. So the rate of surplus value in the US economy has risen by 17%, to reach near its immediate post-war peak.
Source: BLS, author
These trends translate into a key counteracting factor to Marx’s law of the tendency of the rate of profit to fall. Using the ‘top down’ data of the US national accounts that Jefferies rejects, I find a sharp rise of over 20% in US profitability since the pandemic, which is pretty much the same rise for the average rate of profit of the top 100 companies that Jefferies found using his ‘bottom-up’ calculations from corporate accounts.
Source: Federal Reserve, author’s calculations.
Indeed, my latest estimate of global corporate profits (US, Japan, UK, Germany, China) shows an accelerating rise of 18.3% yoy in Q2 2026, with the US and Japan leading the way.
Source: National accounts, author
Does this herald a new era for capitalism, particularly US imperialism, over the next ten years? In another session Nathan Kim and Justine Zhang argued the opposite of Maher and Aquanno. They reckon “things will get worse”, the rate of profit would not recover or be saved by AI and so US capitalism would continue to decline. In the Socialist Register session and in my article in Socialist Register, I put forward three scenarios for the future of global capitalism.
The first was that the low level of investment and productivity growth experienced by the major economies, in what I have called the Long Depression of 2010s, would continue through the 2020s, with stagnation morphing into stagflation, given that inflation is back followng the COVID pandemic.
The second was that after a slump (perhaps triggered by the bursting of the AI financial investment bubble), the survivors will regroup with higher levels of profitability to enable them to invest productively in the new AI technology. So capitalism could enjoy a lengthy period of expansion.
The third scenario was that continued depressed stagnation with declining world trade and productivity growth would lead to an intensification of conflicts over trade and technology and the further rise of reactionary forces in the major imperialist countries, leading even to world war.
Assuming working class action does not replace capitalism and imperialism in the next decade, if I had to choose, I would opt for the second scenario as best for working people. But of course, many sessions at the conference were concerned with the likelihood of the third scenario of reaction and war. In his session, Pete Green referred to the resurgence of mercantilism and war capitalism as Engels and others has observed in the late 19th century. Bill Dunn also argued that imperialism is now engaged in endless wars and even genocide. In another session, entitled Imperialism’s Systemic Drive to War, the presenters argued that imperialism’s debt crisis was shaping the drive towards World War III.
The third scenario can still be avoided by struggle globally. The final plenary at HM conference concentrated on just that: how labour could resist and defeat imperialism. But time is running out.
____________________________
KEEP RESISTANCE MEDIA ALIVE IN 2027
Be the first to comment