Book Review: Daron Acemoglu, What Happened to Liberal Democracy? Remaking a Politics of Shared Prosperity

Professor Daron Acemoglu is a bona fide academic superstar. The publication of any new book by Acemoglu is hence always an important intellectual event; the publication of his latest and much anticipated work is no exception. It is a book offering important insights for the economic future of Northern Ireland in an era of rapid technological and political change.

Acemoglu added the most prestigious award to his phenomenal resume when he in 2024, along with his coauthors Simon Johnson and James Robinson, won the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel. The citation for the award states that the prize was awarded to Acemoglu, along with his coauthors, “for studies of how institutions are formed and affect prosperity”. This citation goes to the heart of Acemoglu’s contribution: his formal academic writings are concerned with the relationship between institutions (or ‘the rules of the game’) and their economic outcomes. As in What Happened to Liberal Democracy? the way institutions shape technological choices forms a large part of the story.

Acemoglu’s analysis within What Happened to Liberal Democracy? illustrates that even if we assumed technology fell from the sky like “manna from heaven”, the metaphor earlier economists used to describe technical progress, not all would have equal access to such gifts. He reiterates that technical change is not a gift: it instead is a product of institutional choice and can be steered in ways that are more or less socially beneficial. Acemoglu observes that the number of robots per an industrial worker in Germany is around twice that found within the United States. Yet he also notes that lower skilled jobs in Germany have been far better protected, as when introducing new technologies German firms retrained blue collar workers for technical and supervisory posts, whereas American corporates resorted to layoffs. Indeed, Acemoglu has instead invoked another metaphor: he suggests the dystopia in H.G. Wells 1895 science fiction classic The Time Machine, where technology can segregate populations rather than produce shared prosperity, is closer to the historical truth.

Acemoglu’s earlier work

In three previous books - Why Nations Fail (2012), The Narrow Corridor (2019) and Power and Progress (2023) – Acemoglu along with his coauthors distils the technical, Nobel worthy insights into surveys intended for lay audiences. In Why Nations Fail, coauthored with Robinson, the focus is on the balance between inclusive institutions, which promote shared prosperity, and extractive institutions, which might enrich elites but do nothing for shared prosperity. While inclusive institutions arise out of dispersed political power and pluralism, extractive institutions arise from inequitable power distributions. This framework is applied to explain divergent economic histories (e.g. the comparative success and failure of the two Koreas).

In The Narrow Corridor, again coauthored with Robinson, the conceptual framework shifts to three political-economic outcomes: a despotic Leviathan in which the power of the state tramples the citizenry, an absent Leviathan in which the state is enfeebled relative to its population. Only r the so-called shackled Leviathan is the state simultaneously powerful enough to guarantee beneficial property rights and public goods, yet restricted (‘shackled’) enough by society to not to regress into a mere expropriator or mafia. Therefore, long-run economic success is only attained by entering and then remaining within the corridor.

In Power and Progress, coauthored with Johnson, the discussion of long-run economic history shifts to the relationship between technology, institutions and prosperity. Their cases examined range from the Middle Ages to the emerging era of Artificial Intelligence (AI). The observation made throughout the book - in the discussion of the Industrial Revolution for instance – is that any growth generated by technological progress has not always been shared equally, rather elites have often captured the gains.

The Golden Age (c.1950-73) is presented within Power and Progress as a period of shared prosperity (within the global North at least). This outcome was underpinned by among other things: labour friendly technology, increased education and training budgets combined with powerful unions. The creation of future “productivity bandwagons”, in which rising prosperity can be equitably shared, is viewed as a formidable challenge in an era of large corporations, including Google, Facebook, Apple, Amazon and Microsoft, unconstrained by worker power.

After the Golden Age

The observations in Power and Progress regarding the Golden Age are developed in terms of mass production and shared prosperity in What Happened to Liberal Democracy? Acemoglu, presents the Golden Age as an industrial compact under which a substantial share of firm profitability spilled over into wage and employment growth; this spillover ensured prosperity was shared. The resulting compact was self-regenerating: worker incomes were spent on domestically-produced goods that sustained employment opportunities (an economic guarantee), while democracy gave protection to the compact at the ballot box (a political guarantee). Acemoglu notes that unskilled workers did particularly well under these guarantees.

 Acemoglu contends that just as the compact was self-regenerating, so deindustrialisation has undone it in ways that are hard to reverse. Since the 1980s clusters based on educated workforces offering high value services have thrived while smaller former urban industrial production centres have declined. Automation breaks the relationship between mass production and shared prosperity as demand for workers need not increase even if firm output grows. This diagnosis leads Acemoglu to prescribe a working-class liberalism in which AI expands worker capabilities ‘and becomes a hand-maiden for creating new tasks and jobs for workers of all skills’.

Applying Acemoglu’s analysis to Northern Ireland

Understood in the light of Acemoglu’s framework the unfortunate predicament for Northern Ireland was that while it did not benefit fully from the upward (manufacturing-led growth) phase of the industrial compact, it suffered badly during the downward (deindustrialisation) unravelling of the compact. As economic historians of the Golden Age have demonstrated, the opportunities for a productivity bandwagon afforded elsewhere was diluted. Headline manufacturing employment shrank between 1949 and 1966 and a consequent tendency towards high levels of unemployment existed. Furthermore, such was the power of vested interests in the staples of linen and shipbuilding that investment was skewed; new industries struggled to gain traction until the 1960s.

Acemoglu’s book - with its message of how successful democracies and economies are self-regenerating, rejection of silver-bullets as well as recognition that under some circumstances public services and unions can support positive economic outcomes - provides a counterpoint to purely market-based thinking. This outlook it shares with Paul Collier’s Left Behind, which I reviewed previously for Pivotal. My review of Left Behind noted that in the Northern Irish context, the place-based policies advocated by Collier make sense; that the experience of devolution illustrates equally that translating decentralisation into shared prosperity requires serious attention to what is sometimes termed ‘institutional geography’ (i.e. design issues).

Regarding institutional geography, the creation and maintenance of power sharing have long been the understandable focus of policymakers, yet while the political settlement of the Agreement has since 1998 produced a marked reduction in violence, much still remains to be achieved by the political institutions and reforms look necessary. Acemoglu argues that powerful interests can skew incentives in their own favour to the detriment of wider society. The failure of the Renewable Heat Initiative (RHI), mentioned in that review of Left Behind, and subsequently the focus of my journal article co-authored with Esmond Birnie, is just the highest profile reminder of what happens when devolution suffers from design weaknesses. The observed outcomes in the RHI case were inefficient, inequitable and far from environmentally optimal; they revealed problems in the way a range of actors all behaved. Readers can judge themselves the extent to which any design weaknesses have been resolved.

Regarding the relationship between AI, institutions and economy, the evidence suggests that Northern Ireland is more reliant on occupations that are both low productivity and vulnerable to AI. Given additional evidence that management quality and AI adoption are also linked, policy in the private sector might best be focused on worker training and business leadership, while public sector spending (e.g. healthcare) might additionally need to be reallocated towards such improvements, as well as ensuring that the planning system and utilities infrastructure do not throttle AI investment. However, AI regulation is a supranational rather than purely devolved challenge, so the best approach for the Executive might be to nurture a culture of adopting international best practice. Any future evolution of AI strategy should evolve further is this direction. In short, a smart second mover strategy, quickly adopting what works from larger markets, might be the best course of action given the devolved settlement.

Dr Graham Brownlow, FREcon is Reader in Economics at Queen’s Business School. In December 2025 he became the first Fellow of the Royal Economic Society to be based in Northern Ireland. Graham has previously held full-time appointments at Dublin City University, University of East Anglia and Auckland University of Technology; prior to these academic appointments he worked with the Government Economics Service (GES) in Whitehall and within the investment management industry in Edinburgh.