Daron Acemoglu and Simon Johnson, Pearson France, November 2024, 632 pages, 29.90 euros. Translated from English by Christophe Jaquet. Original title: Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity (PublicAffairs, May 2023).
The two 2024 Nobel Prize winners in economics trace a thousand years of relationships between technology and prosperity to demonstrate a simple and well-documented thesis: technological progress only improves living conditions for the majority when institutional and social counter-powers force the sharing of gains. Without these counter-powers, the benefits of innovation concentrate in the hands of those who control technology.
The thesis: technological progress does not trickle down
Acemoglu and Johnson start from an empirical observation. In Europe, the standard of living of the peasantry remained stable from the 11th to the 17th century, despite agricultural innovations of the Middle Ages (heavy plow, three-field rotation, water mill). Productivity gains were captured by feudal lords and the Church. The mechanism repeats during the first industrial revolution: between 1760 and 1840, British textile production increased twelvefold, but real wages of workers stagnated for eighty years. Friedrich Engels documented in 1845 living conditions in Manchester inferior to those of pre-industrial countryside [1].
The authors call this phenomenon the "narrative trap": as soon as a discourse automatically associating technology and prosperity becomes dominant, it neutralizes any questioning of the distribution of gains. This trap operates today through the discourse of major technology companies, which present automation as inevitable and universally beneficial.
A thousand years in ten chapters
The book covers five periods. Chapters 1 to 3 analyze the Middle Ages and the first industrial revolution. Chapters 4 and 5 examine the second industrial revolution (1870-1914), where electrification and chemistry created for the first time entire categories of new jobs (technicians, foremen, office workers) in addition to automating old ones.
Chapters 6 and 7 focus on the period 1945-1975, which the authors consider the historical exception. During these three decades, the working class in industrialized countries captured up to 70% of added value. Two factors explain this result: technologies that generated new tasks (not just automation), and an institutional structure that strengthened unions, collective bargaining, and regulation of monopolies. The existence of the Soviet Union as an alternative system played a leverage role in this configuration.
Chapter 8 marks the rupture. From 1980 onwards, digital technologies concentrated gains among a restricted elite. The authors document the stagnation of the American median wage between 1979 and 2019, while productivity per worker increased by 59.7% over the same period [2]. The gap between productivity and wages, virtually non-existent between 1948 and 1973, became structural.
Chapters 9 and 10 focus on artificial intelligence. Acemoglu and Johnson distinguish two possible trajectories: AI as an automation tool (worker replacement, surveillance, control) or AI as an augmentation tool (creation of new tasks, decision support, enhancement of human capabilities). The choice between these two trajectories depends on political and institutional power relations, not on the technology itself.
The proposals
The authors advance five recommendations. First, tax automation to rebalance fiscal incentives that favor the replacement of human labor by machines (labor is taxed, capital is not proportionally). Second, break up technology monopolies through antitrust regulations adapted to digital markets. Third, redirect public research toward technologies that create new tasks rather than simply automating existing ones. Fourth, strengthen workers' counter-powers (unions, works councils, sectoral bargaining). Fifth, democratize technological choices by including citizens and workers in adoption decisions.
The limits
Several criticisms deserve to be noted. Dean Baker, from the Center for Economic and Policy Research, observes that the authors underestimate the role of intellectual property monopolies (patents, copyrights) in the concentration of technological gains [3]. Intellectual property rules are not a side effect of innovation: they actively structure the distribution of benefits. The five billionaires from Moderna, enriched by patents on mRNA vaccines developed with public funds, illustrate this mechanism.
Noah Smith, economist and essayist, contests certain historical examples. The Haber-Bosch process (ammonia synthesis, 1909) allowed feeding billions of people without any social struggle being necessary to diffuse its benefits. The Panama Canal benefited the world economy without forced redistribution. Japan in the Meiji era industrialized its economy with gains widely shared in an authoritarian framework [4]. These counter-examples suggest that the authors' thesis, while capturing a dominant dynamic, is not universal.
The book also remains relatively silent on contemporary developing countries. Most of the analysis focuses on Western Europe and the United States. The Chinese experience (800 million people lifted out of poverty between 1978 and 2020 in an authoritarian framework) or Indian experience (growth driven by IT services) is only treated in passing.
Why read this book now
Power and Progress provides an analytical framework for ongoing debates about AI regulation. The European Parliament adopted the AI Act in March 2024. The United States hesitates between sectoral regulation and voluntary approach. China deploys AI in mass surveillance. In each case, the question posed by Acemoglu and Johnson remains the same: who decides the orientation of technology, and for whose benefit?
The 2024 Nobel Prize awarded to Acemoglu, Johnson, and Robinson (for their work on "how institutions influence prosperity") validates the book's institutionalist approach. Power and Progress extends Why Nations Fail (2012) and The Narrow Corridor (2019) by applying the same theoretical framework to technological history. The question is not whether technology progresses, but who profits from it and through what political mechanisms.
References
- [1] Friedrich Engels, The Condition of the Working Class in England, 1845.
- [2] Economic Policy Institute, "The Productivity-Pay Gap", updated October 2022. https://www.epi.org/productivity-pay-gap/
- [3] Dean Baker, "We Can Do Better with a Thousand Years: Review of Power and Progress", CEPR, 2023. https://cepr.net/publications/we-can-do-better-with-a-thousand-years-review-of-power-and-progress-by-daron-acemoglu-and-simon-johnson/
- [4] Noah Smith, "Book Review: Power and Progress", Noahpinion, 2023. https://www.noahpinion.blog/p/book-review-power-and-progress
- Receive analyses from the Journal directly in your inbox.
- China's CO2 emissions decreased by 0.3% in 2025. It's a modest figure. But it extends a trend that has lasted for 21 consecutive months: since March 2024, emissions from the world's leading polluter have been "stable or declining."
- Does AI destroy jobs? Two studies published in March 2026 — one by Harvard Business School, the other by Anthropic — provide the first solid empirical data. And the answer is more nuanced than public debate suggests.
- Malaria killed 608,000 people in 2022, 95% of whom were in sub-Saharan Africa and 78% were children under five years old. And for the first time, a vaccine deployed on a large scale shows measurable results in the field.
- Wanting to be 20 years old today. An independent media outlet that documents progress with rigor, without naivety or catastrophism.
- Structured reading summaries: central thesis, key arguments, limits, and verdict.
- JdP is an independent editorial project founded on data, counter-narratives, and lucid optimism. Each article is sourced, nuanced, and open to discussion.
- The Journal d'un Progressiste uses cookies to improve the reading experience and understand how the site is used. No data is collected for commercial, advertising, or resale purposes. Cookies necessary for site operation are always active. Optional cookies are only activated with your explicit consent, in accordance with GDPR.