Daron Acemoglu, the esteemed Nobel laureate in economics for 2024 and the third most cited economist in history, presents an intriguing paradox in his latest tome, What Happened to Liberal Democracy?. Despite his reputation, the book’s most significant shortcomings lie in its economic analyses. Acemoglu thoughtfully examines the growing distance between modern liberalism and the working class, yet when he attempts to propose solutions, the foundational elements of economic reasoning become conspicuously absent. He juxtaposes flawed markets with an idealized government, advocates for expansive bureaucracies while sidestepping the well-documented risks of bureaucratic capture, and while he is adept at intertwining institutions with economic theory, he occasionally lauds democratic consensus without acknowledging the elite structures that underlie it.
Acemoglu positions himself firmly on the left of the liberal spectrum, asserting that liberalism has historically fostered both material and social advancement. He advocates for a âworking-class liberalismâ rooted in tradition, which, in his view, entails leveling the economic playing field and promoting technological advancements that enhance rather than replace labor. His intentions are commendable; he acknowledges that elites are unreliable cultural stewards, yet he fails to clarify why these same elites should be trusted to manage technology, redistribute wealth, or steer the administrative state.
He contends that the postindustrial economy favors cognitive skills relevant in sectors like finance, health, and education, while diminishing the role of manufacturing. Acemoglu claims that automation has severed the once-strong connection between mass production and shared prosperity, leading to a dearth of quality jobs for those without a college degree. This perspective aligns with the economic rationale for populist dissatisfaction, a notion popular among left-liberals.
What sets Acemogluâs narrative apart is his willingness to acknowledge the complicity of left-liberals in the alienation of the working class. Much like in my own book, Why Democracy Needs the Rich, he suggests that educated liberals are drawn to social engineering as a means of enhancing their status and influence. He recognizes that this cohort has misused its power to impose cultural values without democratic endorsement, further alienating the very working class they aim to uplift. He expresses regret over the contemporary Left’s tendency to believe that safeguarding minorities and vulnerable groups necessitates silencing dissenting opinions. This acknowledgment is refreshing, yet Acemoglu’s critique of elite power appears to vanish when he endorses the very policies that would empower those elites.
His advocacy for increased redistribution and a more expansive government role in economic matters stems from his principle of ânondomination,â which he defines as the ability to act freely without being subject to the arbitrary authority of individuals or institutions. This stands in stark contrast to classical liberalism, which advocates for a limited government focused on preventing coercion and fraud, safeguarding property rights, and delivering public goods.
In Acemoglu’s framework, nondomination necessitates some degree of redistribution and the provision of public servicesâhealthcare and infrastructure, for instanceâto ensure equitable economic opportunities. He argues that an active government is essential for directing new technologies to benefit all workers. However, he does not adequately address the potential pitfalls surrounding this principle. When does reliance on an employer, or even family, cross the line into domination? Which social norms qualify as oppressive, and who holds the authority to decide? Additionally, how can a government that supposedly liberates citizens from oppressive norms avoid engaging in the very social engineering that Acemoglu criticizes?
The ambiguity surrounding the nondomination concept also raises pertinent concerns from the realm of public choice theory. This field examines how political dynamics operate amid the competing interests of various groups. Nondomination could easily be co-opted by any number of interest groups claiming victimhood, enabling politicians, bureaucrats, and unions to define the term in ways that ultimately favor their agendas, even at the expense of others.
While Acemoglu warns against letting working-class liberalism overlook the needs of the disadvantaged to win over the middle class, itâs essential to recognize that government policies skewing towards the interests of the affluent, particularly older demographics, are not merely moral failings. They are predictable outcomes of voting behavior, where older citizens, possessing more time and resources, are more likely to engage in the electoral process. Additionally, the educated elite, adept at navigating administrative systems, often wield far more influence than the working class due to their lack of free-rider dilemmas. When Acemoglu suggests a new agency to ensure AI benefits workers, he neglects to specify how he would establish structures or norms to counteract the predictable incentives steering it away from his desired outcomes.
What safeguards pluralism when elites seek to utilize courts or agencies to forge a consensus that serves their interests?
Acemoglu’s case for a pro-worker AI agency lacks a robust economic rationale. He posits that such an agency should ensure AI enhances worker productivity, asserting that neither digital advertising nor the sale of automation software will inherently drive the development of worker-friendly AI. Nevertheless, he fails to pinpoint the market failure necessitating this intervention. If AI can aid electricians in diagnosing issues or nurses in conducting complex tasks, firms will inherently seek to provide such tools as a competitive advantage. Incumbent companies do not monopolize the market, and startups are incentivized to fill any gaps and enhance worker capabilities. The market’s strength lies not in its need for direction but in its natural drive to meet demands effectively.
Acemoglu also advocates for a tax system that treats capital and labor neutrally, a proposal that avoids market distortions and minimizes bureaucratic expansion. However, it underscores his inability to demonstrate a clear market failure. Why is neutrality insufficient? He faces a dual challenge: first, he must clarify why the market for upskilling workers through AI is flawed, and second, even if it is, he inadequately compares it to a government agency whose imperfections he has yet to model.
To illustrate the economic challenges posed by the AI agency he proposes, consider this: Acemoglu suggests that AI will bolster demand for teachers by facilitating smaller group instruction following AI-generated assignments and lesson plans, implying an increase in the teaching workforce rather than a reduction. However, why should the government prioritize boosting demand for a specific occupation instead of allowing the market to dictate these changes?
Occupational demands fluctuateâjust ask the agricultural sector. Labor-saving technologies can lower prices, and teachers, as consumers and taxpayers, are also affected by these changes. Historically, such innovations have led to the emergence of new job types we could hardly have envisioned. While some predict that we are on the brink of artificial general intelligence that could displace nearly all human labor, Acemoglu himself remains skeptical of such projections.
Acemoglu quotes F. A. Hayek, who cautioned that economic âdataâ cannot be fully comprehended by a single mind capable of extrapolating its implications for society. This observation highlights the knowledge problems that arise when government initiatives replace market mechanisms. Which tasks warrant automation? What technologies will cultivate new, complementary jobs? Will innovations that displace labor simultaneously create new opportunities?
Moreover, government agencies are prime breeding grounds for rent-seeking behaviors. This omission is particularly noteworthy given Acemoglu’s acknowledgment that âthe speed and effectiveness of public service provision have diminished while costs have soared due to excessive regulation.â Teachers’ unions, for instance, possess significant power and may leverage the agency to safeguard their interests, while other professional organizations, like the Association of American Law Schools, could impose barriers to competition. Even in matters of national security, politicians often redirect funds to favored constituencies and regions. Consequently, an agency with such broad and ambiguous objectivesâlike promoting pro-worker AIâcould easily become a tool for political gain rather than societal benefit.
The most compelling aspect of Acemogluâs book is his insistence on the necessity of broad democratic approval for government-led cultural transformations. He argues for a âpractical social compactâ that leverages self-governance and secures buy-in from a significant portion of society’s communities. He is rightly concerned that elites may harness government institutions to impose their values on the general populace.
However, he falls short in recognizing how elite institutions manufacture democratic consent. His example of consensusâthe 2022 Respect for Marriage Act, which mandates state recognition of same-sex marriagesâfails to account for the complex dynamics leading to that consensus. This statute followed the Obergefell v. Hodges ruling, which overturned bans in thirty states, suggesting that same-sex marriage was not merely a product of grassroots democratic agreement but rather a top-down judicial decision that alienated many citizens who felt their preferences disregarded by a court comprised of elite legal minds.
Acemoglu rightly emphasizes the importance of âcommunity-level experimentationâ for social adaptation to changing circumstances, acknowledging that âexperimentation is powered by variation.â Yet he overlooks federalism as a crucial mechanism for fostering such diversity and gradually building democratic consensus. Federalism permits varying policies across jurisdictions, enabling open discourse on the outcomes. This is how national consensus, if it is to be achieved, can best be cultivated. When citizens can voice their opinions without preemption, they are far more likely to accept the final verdict. Federalism serves as the political knowledge equivalent to competition in economic knowledgeâa decentralized discovery process.
Acemogluâs analytical shortcomings offer broader lessons for left-liberals. Like their classical liberal counterparts, they cannot evade the necessity of addressing comparative institutional questions. Do concepts like nondomination inadvertently empower state domination? What specific market failures hinder the market from enhancing long-term welfare? What mechanisms can prevent new regulatory agencies and redistributions from being co-opted by vested interests? What safeguards exist to preserve pluralism when elites seek to utilize courts or agencies to promote an elite-centric consensus? Acemoglu correctly identifies the potential for elites to distort left-liberalism into a vehicle for oppression. However, he fails to clarify why the new AI agency he advocates for would not merely replicate the same elite-driven dynamics.
John O. McGinnis is the George C. Dix Professor in Constitutional Law at Northwestern University and a senior writer at Law & Liberty. He is the author of Why Democracy Needs the Rich (2026), Accelerating Democracy (2012), and coauthor, with Mike Rappaport, of Originalism and the Good Constitution (2013). He is a graduate of Harvard College; Balliol College, University of Oxford; and Harvard Law School. He has contributed to prestigious law reviews including Harvard, Chicago, and Stanford, as well as opinion journals like National Affairs and National Review.
This analysis originally appeared at Law & Liberty.
Daron Acemoglu, Nobel laureate in economics, delivered his prize lecture in economic science in Aula Magna, Stockholm University, Sweden, on December 8, 2024. (Pontus Lundahl/TT/Code 10050 Credit: TT News Agency/Alamy Live News)
