How Meritocracy Became Self-Defeating

The Rise of Market Meritocracies

This formulation immediately raises a set of questions that every meritocracy must answer: What constitutes merit, and how (by what metrics) should merit, so understood, be measured? Meritocracies have answered these questions in various ways. The ancient Chinese meritocracy that began in the Tang dynasty and lasted for more than a millennium, reaching its peak in the Song dynasty, associated merit with erudition understood in terms of literary style and classical learning, and it measured merit, so understood, by extremely rigorous and highly formal exams. The meritocracy that ruled the Mongol empire, by contrast, associated merit with conquest, framed in terms of martial valor and skill and measured in terms of success on the field of battle.

Today, meritocracies all across the world, and especially in the United States, find it most natural to make markets the measure of merit. They associate merit with utility, so that a person’s merit reflects, in Greg Mankiw’s words, “the value of what [they] contributed to society’s production of goods and services” (Mankiw 2010, 295; see also Mankiw 2013, 32–33; Mankiw 2014). They measure this value in terms of market prices (Mankiw is, unsurprisingly, an economist) and especially the person’s wage (which, in a competitive labor market, equals the marginal product of their labor). And they emphasize that the skills needed to command high wages—the accumulated human capital that a worker can mix with her contemporaneous labor effort to yield marketable goods and services—are acquired in schools (and colleges), which are increasingly understood as factories for building human capital at scale. These schools, in turn, deploy their own metrics—course-specific grades and standardized test scores, calibrated to predict future labor-market success—to measure how much human capital each of their graduates has built.

Market meritocracy is so charismatic because it claims to vindicate the ideals that captivated Jefferson, Boutmy, and Shaw: to square the inevitable hierarchy that attends the imperatives of governing with the democratic equality of all citizens and to nurture the elite required for administering complex societies in the common interest.

First, market meritocracies square hierarchy with democratic equality by providing, universally, a fair opportunity to get ahead. It is, by now, not news that competitive markets can coexist with, and even generate, enormous inequalities in economic outcomes. In the United States, the Gini coefficient for household income has increased by nearly a fifth over the past 50 years (Federal Reserve Bank of St. Louis 2025; Markovits 2019, figure 4) while the share of national income captured by the richest 1 percent of households has roughly doubled (World Inequality Database 2024). But even as it has magnified inequality of outcomes, meritocracy has removed important barriers to equality of opportunity. Buckley had seen a future he couldn’t prevent, as meritocratic admissions did indeed prefer the talented outsider.

The overwhelmingly most significant reason why the American elite is more diverse today than it was 50 years ago is meritocracy, which replaced a sclerotic hereditary elite composed of lazy and untalented rich, white, Protestant men with a dynamic and diverse elite that includes many talented people who would, just three generations ago, have been categorically excluded on account of their class, race, religion, or gender. This diversification is everywhere and undeniable, evident in broad data and vividly captured in case studies: A prime-aged Black man whose income put him at the 90th percentile of the Black distribution would have been at roughly the median of the white distribution in 1960 but near the 75th percentile of the white distribution today (Bayer and Charles 2018, 1459); Princeton, which didn’t produce its first Black graduate until 1947 (Armstrong, n.d.) and didn’t include women until the class of 1973 (Yang, n.d.), today has a student body that includes roughly 50 percent people of color and 50 percent women (Princeton University Undergraduate Admission 2025); and the hyper-elite Cravath Law Firm, whose partnership consisted exclusively of Christian men from its founding in 1819 (Cravath, Swaine & Moore LLP 2026) until 1958 (Auerbach 1976), is today led by Presiding Partner Faiza J. Saeed, a woman of Pakistani origin (de la Merced 2016). These changes, and a thousand others like them, are the direct result of the end of openly unmeritocratic, categorical subordination. Even those who argue that the diversification of the elite has gone too far—critics who claim that diversity, equity, inclusion and affirmative action are unmeritocratic, something that is far from clear (Durlauf 2008)—must acknowledge that the policies they criticize operate on the narrow outer edge of a deeply meritocratic revolution, which has vastly expanded opportunity to groups that were previously excluded.

Second, consider how market meritocracies construct an elite based on virtue and talent—an elite whose superior skills serve the common good. Meritocracies of erudition create an indulgent, even snobbish, elite that measures accomplishment by values that others don’t and in fact can’t share or even appreciate. Martial meritocracies, for their part, measure accomplishment in terms of domination, certainly over outsiders and, at least potentially, over the citizens whose equality modern meritocracy is supposed to serve. Market meritocracy, by contrast, measures the elite’s accomplishment by the service that it renders the rest of society. Moreover, the prices that fix the market meritocrat’s worth themselves have a democratic component: In an idealized market, every good’s price literally equals the sum of what everyone else must give up, as measured from all their points of view, for its owner to possess it (Markovits 2012). This mechanism applies to the meritocrat’s labor also: A meritocrat’s wages, in a market meritocracy, equal what society would lose if she withheld her labor, as measured from everyone else’s point of view. As Mankiw says—calling this the “principle of just deserts” (Mankiw 2013, 32–33)—“a person who contributes more to society deserves a higher income that reflects those greater contributions” (Mankiw 2010, 295; see also Mankiw 2014). Market meritocracy implements this intuition in an immediate and precise way, so that market meritocrats do well by doing good.

Both arguments are plausible. They powerfully announce market meritocracy’s moral advantages over the aristocracies of race, breeding, and gender that meritocracy displaced. For a time, in the immediate post–World War II generations, the case for market meritocracy was irrefutable and irresistible.

But as meritocracy advances, it becomes self-defeating: Increasingly, each argument fails—indeed, fails on its own terms. The failure of the first argument—that meritocracy justifies unequal outcomes by creating equality of opportunity—is more familiar, although interestingly less complete. The failure of the second argument—that meritocracy creates an elite whose quest for private advantage serves the public good—is not commonly appreciated. Nevertheless, it is the fundamental reason why meritocracy now defeats its own values and fails to live up to its initial promise.

Just Deserts?

The argument that meritocracy justifies unequal outcomes—because fair equality of opportunity means that everyone gets what they deserve—falters in generically familiar ways. But the specific role that meritocracy itself plays in blocking fair equality of opportunity—which is what makes meritocracy self-defeating—is less familiar and therefore worth elaborating.

Mankiw’s principle of just deserts contains a moral gap. Even if it’s true that the meritocrat’s private advantages track her contributions to the public good, this isn’t by itself sufficient to justify her exceptional wealth and power. Instead, she will deserve these advantages only insofar as she deserves to be in the circumstances that yield them—to hold the jobs that provide her income and status and to possess the training and skills that allow her to do those jobs. But meritocrats typically get jobs on account of their social capital, their natural talent, and their training, and people do not deserve (nor typically do they even choose) any of these attributes. They obviously don’t deserve the networks that they are born or fall into. It’s almost as obvious that they don’t deserve the luck in the genetic lottery that accounts for whatever unusual natural talents they possess. Finally, meritocrats at most partly deserve their training, which depends enormously on investments in education made by others (most notably their parents, but also schools and other institutions)—investments that the meritocrats again haven’t chosen and can’t plausibly be said to deserve.

The undeserved advantages that underwrite every meritocrat’s success are themselves anything but random, moreover. Instead, they are distributed according to mechanisms that undermine—indeed, make a mockery of—meritocracy’s claims to legitimate unequal outcomes by sustaining equality of opportunity. Even in countries—like the United States—that have formally abolished aristocratic titles, social class continues to determine social capital, and social capital (even when it’s totally unrelated to the skills needed to perform a job) continues to exert a massive influence over who gets hired, as any number of close demographic studies of elite hiring today document in exquisite detail (Rivera 2015; Ho 2009).

Moreover, the educational investments, paid for by others, through which children build the skills that will determine what jobs they get as adults are enormously unequal, with children of rich parents receiving massively greater investments than middle-class or poor children do, especially in the United States. An Organisation for Economic Co-operation and Development (OECD) survey of 34 advanced economies reveals that the U.S.—whose public schools are overwhelmingly funded by local property taxes—is one of only three member states whose public education system invests more per pupil in schools that serve rich students than in schools that serve poor students (OECD 2013; Reich 2014; Porter 2013). And the differences can be enormous: Scarsdale, New York, where the median household income is more than $250,000 and the median house price exceeds $1.7 million (U.S. Census Bureau 2024), spends more than $35,000 per student per year on its public schools (NCES 2022), compared to a national average of just over $15,000 (NCES, n.d.). Elite private schools spend even more extravagantly on educating the overwhelmingly rich children who attend them: A typical school from Forbes Magazine’s list of the top 20 private schools, where roughly 70 percent of students come from households in the top 4 percent of the income distribution (Gaztambide-Fernández 2009, 35), might spend $75,000 per student per year (Markovits 2019, 126).

These unequal investments produce dramatic differences in the academic achievement between children from rich families and all other children: not just between rich and poor but also between the rich and the middle class. The academic achievement gap between rich and poor students now exceeds the gap between white and Black students in 1954—the year Brown v. Board of Education was decided—so that economic stratification produces greater educational inequality today than American apartheid did at mid-century (Reardon 2011, 99; 2013). Educational inequalities add up over the years, so that by the time students reach the end of high school, the achievement gaps between the most and least privileged children have grown enormous: In a recent year, for example, high school seniors who had at least one parent with a graduate degree were 150 times more likely to achieve SAT verbal scores at the Ivy League median than children neither of whose parents graduated high school (Markovits 2020b). Moreover, the achievement gaps between the rich and the middle class are much bigger than the gaps between the middle class and the poor: For example, the gap between the average SAT scores of children from households in the top 5 percent of the income distribution and children from median-income households is roughly twice as big as the gap between the average scores of median-income children and those of children from households at or below the poverty line (Goldfarb 2014; Carnevale and Strohl 2010).

It is unsurprising then that rich children dominate not just poor but also middle-class children in the college admissions competition, especially at elite colleges: According to one study, the student bodies of the roughly 150 most selective colleges comprise 14 times as many students from households in the top quarter of the income distribution as from the bottom quarter and six times as many as from each of the middle two-quarters (Carnevale and Strohl 2010). Another study of the 91 most selective colleges found 24 times as many rich as poor students, and there have been recent years in which Harvard, Princeton, Stanford, and Yale (which each spend more than $100,000 a person per year educating their students) have enrolled more students from the top 1 percent of the income distribution than from the bottom 60 percent (Chetty et al. 2017, 1, 14; Aisch et al. 2017).

Whereas aristocratic families passed their privilege down through the generations by accumulating land, factories, or even stocks and bonds and then distributing this physical or financial capital to children as bequests on the death of parents, meritocratic families pass privilege down their generations by securing exceptional investments in the human capital of their children while the parents are still alive. These unequal educational investments—paid from preschool through graduate school—cumulate: If a typical U.S. one-percenter family took the excess investment made in educating its children over and above what is invested in typical middle-class (not poor) children and, each year, invested this excess in the S&P 500 to create a fund that might pass to the children on the death of the parents (mimicking a traditional aristocratic inheritance), it would yield a meritocratic inheritance of between $10 and $15 million per child (Markovits 2019, tables 1 and 2).

The elite’s investments in human capital pay off. One recent study has estimated the purely economic rate of return on elite university education at between 13 and 14 percent per year (Economist 2014), and super-elite bachelors of arts (BAs) more than double the income gains yielded by average schools (Owen and Sawhill 2013; Schneider 2009, 4). Fewer than one in 75 high school dropouts, only about one in 40 workers with no education beyond high school, and only about one in six workers with a BA only will accumulate lifetime earnings equal to the median professional school graduate’s (Carnevale et al. 2011, 6). Elite educations matter even more for getting to the very top of the American hierarchy. More than half the partners at the five most profitable law firms in the country have graduated from law schools conventionally ranked in the “top ten” (Markovits 2019, 11). And a recent survey—citing numbers almost too extreme to credit—reports that nearly 50 percent of America’s corporate leaders, 60 percent of its financial leaders, and 50 percent of its highest government officials attended only 12 universities (Dye 2014, 180). Meanwhile, the richest 1 percent of American households owe roughly three-quarters and the richest 0.1 percent owe roughly two-thirds of their incomes to selling their (super-skilled) labor and therefore to their enormously intensive and expensive educations. ¹

When inequalities in outcomes grow large enough, equality of opportunity becomes impossible; and whatever its initial hopes, meritocracy has now established a new technology for the dynastic transmission of privilege. Rich parents buy vastly unequal schooling for their children, which produces extravagant skills. The children then grow up to deploy their meritocratic inheritances in the labor market, thus dominating the competition to secure the most elite jobs. The jobs pay extravagant salaries, which the new generation of meritocrats deploys to fund extravagant educations for its children. And the cycle begins again in the next generation. In spite of its original promise to provide opportunity to everyone and diversify the elite, meritocracy today constitutes the single biggest obstacle to equality of opportunity in the new world that it has brought into being.

This dynastic cycle, moreover, stretches the gap in Mankiw’s principle of justice deserts into a massive, and conspicuous, moral chasm. Left liberals—who tend to view justice as the problem of how fairly to share the total benefits of social cooperation—unsurprisingly put the problem at the center of their economic thought. As John Rawls (1971) has pointed out, nobody deserves their position in either the natural lottery, which distributes genetic talents across babies, or the social lottery, which distributes babies across families. Indeed, the problem is so conspicuous that even libertarians, who resist collectivizing the social product, often accept that individual desert cannot provide the principle behind their resistance. Hayek (1960, 94), for example, acknowledges that a person’s talents “do not depend on any credit due to him for possessing them” and consequently that “the value which a person’s capacities or services have for us and for which he is recompensed has little relation to anything that we can call moral merit or deserts.”

Meritocracy has become self-defeating. Meritocracy undermines its own promises to square unequal outcomes with equal opportunities and to render the hierarchy needed to manage a complex society consistent with the egalitarian imperatives of a democratic age. But perhaps these observations don’t defeat the argument for meritocracy absolutely. For one thing, there are important ways in which the difference between meritocratic and aristocratic privilege survives them. Most obviously, the greater racial, ethnic, and gender diversity of the meritocratic elite isn’t undone by the fact that privilege may still be passed down along these (previously subordinated) family lines. Moreover, even if the meritocratic technology is no less effective at sustaining elite dynasties than the aristocratic technologies that it replaced, the mechanism by which advantages are passed from parents to children might make dynastic transmission more or less morally offensive.

Once, aristocrats automatically assumed their titles on the death of their parents, so that rank depended expressly and immediately on accidents of birth, and elites wore the undeservedness of their advantage on their sleeves. ² Today, by contrast, meritocrats must devote immense diligence and skill to acquiring and sustaining their privilege, which is built in and through enormously competitive and intensely demanding schools and jobs. Acceptance rates of 5 percent or less across the Ivy-Plus colleges entail that even if rich parents are a nearly necessary condition for admission, they are far from a sufficient condition. However much privilege they come from, elite students will have secured their places at top schools by beating out a reserve army of similarly privileged competitors in an intense meritocratic competition. And 70-plus-hour workweeks render the “extreme jobs” (Hewlett and Luce 2006) ³ through which superordinate workers capture their extravagant incomes the opposite of sinecures.

Even the most successful meritocratic dynasties can’t simply bequeath or assume their meritocratic inheritances but must, instead, devote enormous sweat equity to rebuilding privilege in each new generation. This conspicuous and genuine display of effort, diligence, and skill clothes meritocratic privilege in a thick, almost impenetrable layer of virtue. Of course, even virtue might not be enough to secure desert: As Rawls (1971, 104) observes, even a person’s character, including her capacity and even inclination for hard work, “depends in large part upon fortunate family and social circumstances for which [she] can claim no credit,” so that “the notion of desert seems not to apply to these cases.” ⁴ But this argument—that people don’t deserve and therefore shouldn’t benefit from their characters and virtues—runs the risk of proving too much, of sweeping so broadly that it sweeps its own foundations out from under it. Accounts of personality that treat character as just another kind of circumstance are so deflationary that they leave it unclear why human beings are the kinds of creatures whose hierarchy and subordination are worth worrying about to begin with.

All in all, then, considerations concerning opportunity and desert leave the case for meritocracy wounded but not killed. Meritocracy is self-defeating, in a way. The rich and the rest don’t compete on a level playing field, but unlike under formal aristocracies of race or breeding, they do compete, and the rest can sometimes win; and the meritocratic elite’s virtues, even if not self-made, remain really admirable. Meritocracy’s promise to justify unequal outcomes through open and equal opportunities may be narrowly formal, but it is not entirely false.

Unmasking Merit

This state of play raises the stakes concerning the second justification for market meritocracy—the idea that it constructs an elite around real rather than false virtues, an elite whose exceptional skills don’t just sustain private advantage and privilege but instead contribute to the common good. The most important objection to market meritocracies isn’t the familiar argument that meritocrats don’t deserve their virtues and therefore don’t deserve to capture the benefits that these virtues confer, but rather the unfamiliar argument that meritocrats aren’t in fact virtuous at all—that the skills built in elite schools and deployed in elite jobs, when assessed in proper context and measured against the right counterfactual, don’t in fact contribute to the general welfare.

Unmasking merit requires a two-step argument. The first stage rehearses the recent history of elite work and wages, to identify the root causes of the explosion of meritocratic inequality. The second stage places the historical facts in an economic and moral context, to reveal that elite skills contribute much, much less to the common good than champions of meritocracy suppose and that they may even, when accurately measured, make society overall poorer rather than richer. Accordingly, what we conventionally call merit is actually an ideological conceit, constructed to hide unjustified hierarchy and subordination under a veneer of the common good. Meritocracy is self-defeating because meritocratic inequality makes merit itself into a sham.

There is no doubt that the private returns to elite training and skill are greater today than they have ever been. Over the course of market meritocracy’s recent career—in the roughly 60 years since Kingman Brewster and Bill Buckley did battle at Yale—elite wages have exploded, even as mid-skilled, middle-class wages have stagnated. It is well known that whereas a typical large-firm CEO at the middle of the last century made roughly 20 times a production worker’s income, his counterpart today makes roughly 300 times as much (Mishel and Davis 2015; AFL-CIO 2026). But the scale and breadth of the transformation can still surprise and even shock. Meritocracy has produced executive compensation that’s enormous compared not just to production workers’ wages but also to corporate income writ large: In a recent year, the five highest-paid employees of each of the firms in the S&P 1500 (7,500 elite executives) together captured income equal to 10 percent of the total profits of the entire S&P 1500 (Bebchuk and Grinstein 2005, 284, 297), while a somewhat broader elite composed of professional finance-sector workers each year capture wages that account for nearly half of a typical Wall Street firm’s net revenue (Ho 2009, 255; McDonald 2005). Moreover, the shift of income from mid- to super-skilled workers isn’t limited to management and finance. A cardiologist was paid about four times a nurse’s income in the mid-1960s but makes more than seven times as much today; law firm partners made a little less than five times a legal secretary’s income then but make more than 40 times as much now; and whereas David Rockefeller made roughly 50 times a bank teller’s income for running Chase Manhattan Bank in 1969, Jamie Dimon now makes well over 1,000 times a bank teller’s wages for running the bank’s successor, JPMorgan Chase (Markovits 2019, 18). These ratios literally cash out the educational segmentation of the labor market documented earlier—the fact that fewer than one in 75 high school dropouts will enjoy lifetime incomes equal to the median professional school graduate’s.

Many mechanisms underlie this transformative shift of wages to favor meritocrats, but the most important—the most powerful independent mechanism and the root cause behind some of the other mechanisms—is skill-biased technical change. Finance-sector workers, for example, were at the middle of the twentieth century neither better educated nor more highly paid than other workers—The Economist magazine even asked, in a 1963 article headlined “Has Banking a Future?,” whether banks were “the world’s most respectable declining industry” (Rosenthal 2013). Mid-century bankers deployed common sense and personal networks to allocate capital through relatively straightforward debt and equity contracts, based on reasoning not altogether different from that deployed in managing the finances of a small business or even a middle-class household. But The Economist could hardly have gotten the future more wrong: Over the subsequent 50 years, finance grew to roughly double its share of total GDP, without having increased its share of total employment. And today, finance workers are twice as likely to have gone to college and 70 percent more highly paid than others, and elite finance-sector incomes account for as much as a quarter of rising wage inequality economy-wide (Philippon and Reshef 2012, figure 1). The elite workers manage this by deploying new financial technologies—elaborate asset pricing models, complex derivatives, and intricate trading platforms—that allow a few super-skilled workers using digital computers to replace many mid-skilled workers.

Management has traversed a similar path: While mid-century firms were run by elaborately layered armies of mid-skilled, modestly paid middle managers, the corporate restructurings of the 1980s and 1990s eliminated middle management, so that today’s firms are run by a small cadre of super-skilled elite executives that captures 10 percent of total profits for itself and outsources tasks that they can’t do themselves to super-skilled, highly paid management consultants (Markovits 2020a). Once again, the elite executives deploy elaborate new technologies to monitor, analyze, organize, and command their complex business organizations without depending on middle management. Finally, while finance and management represent the most prominent and most consequential cases of technologically induced labor-market polarization, they are far from alone. Even in manufacturing, the economic sector most closely associated with mid-skilled, middle-class jobs—and in which total employment fell by roughly a third between 1992 and 2012—the number of jobs for workers with a BA rose by 2.4 percent and for workers with a graduate degree by 44 percent (Shapiro 2013; Manufacturing Institute 2014; Torpey 2014). ⁵ The new workers deploy new technologies—robots and algorithms—so efficiently that manufacturing’s share of real GDP has held roughly steady even in the face of the massive decline of its employment share (Baily and Bosworth 2004, 4; Chien and Morris 2017; Wehner and Beschel Jr. 2012; Nutting 2016). ⁶ Overall, these technological changes have eliminated roughly a quarter of the American economy’s mid-skilled jobs since 1980, while the share of jobs for super-skilled workers has increased by more than a third, and technical and professional workers’ share of total employment has more than doubled since mid-century (Jaimovich and Siu 2020, figure 3).

In the economic sectors that dominate meritocratic inequality, and indeed across the economy as a whole, skill-biased new technologies have complemented super-skilled human workers and substituted for mid-skilled human workers, and in this way simultaneously driven an explosion of economic inequality and reframed that inequality to give it a meritocratic cast. New technology hasn’t always operated in this skill-biased way, however. The innovations that transformed the world by producing the first Industrial Revolution, for example, were biased against skilled workers—as factory methods broke intricate final products and integrated production processes down into simple components and distinct stages of production, which allowed unskilled factory workers to replace highly trained craftspeople and artisans (Markovits 2019, 248). (The artisans resisted new technology for precisely this reason, and from the very beginning: The Luddites were skilled weavers who organized a campaign of petitions, sabotage, and even riots to resist the “scribbling machines” that threatened their jobs.) Innovation biased against skill spread across industries, right alongside industrialization: “Many of the major technological advances of the nineteenth century . . . substituted physical capital, raw materials, and unskilled labor . . . for highly skilled artisans” so that by the century’s end, not just weavers but most skilled artisans, including “the butcher, baker, glassblower, shoemaker, and smith . . . [had their] occupations profoundly altered by the factory system, machinery, and mechanization” (Goldin and Katz 2008, 122). Moreover, this bias against skill endured into the twentieth century, when the Ford Motor Company, in the 1910s, deployed modularization and assembly-line methods to build cars without needing artisan mechanics or coachmakers. ⁷

These patterns make it essential to ask: What has driven these technological changes? Why have recent innovations, unlike their predecessors, so dramatically favored elite workers? Why is the arc of innovation bending toward skill just now?

The answer centrally invokes the trends in schooling discussed earlier. Innovation, rather than being driven by an independent technical or scientific logic, is instead embedded in broader social developments. Interested innovators have always created inventions to mix with a society’s other available resources, so that they might generate economic returns. This pattern means that the arc of innovation bends toward the resources that are available for innovators to exploit. In early agriculture, for example, innovators in arid places invented drip irrigation, while innovators in flood plains invented paddy field farming. More recently, societies in which labor was scarce relative to land (such as the United States) developed very different agricultural techniques from societies in which land was scarce relative to labor (such as Japan). In each case, the arc of innovation bent to meet the prior resource base that new technologies might exploit to increase production.

The same forces applied to influence innovation in subsequent eras—save that the main resource available to innovators has changed, increasingly emphasizing not land or even other physical capital but rather human capital, that is, the skill and effort of free workers (Acemoglu 2002). The arc of subsequent innovation therefore bent to meet the character, and also the distribution, of the human capital that might be deployed to implement the new technologies that innovators invent. The first Industrial Revolution, for example, arose against the backdrop of a new mass of unskilled workers—first from the countryside and, later, fleeing famine-stricken Ireland—that concentrated in England’s cities beginning in 1800 and led London, Manchester, Birmingham, and Liverpool to grow by factors of between five and 10 over the course of the nineteenth century (Markovits 2019, 250). This historically unprecedented mass of unskilled labor accumulated in English cities, waiting for a technology to exploit it, even as social norms (which constituted elites as what Thorstein Veblen [1899] called a “leisure class”) led educated elites and even skilled artisans to abjure intense or industrious labor.

Small wonder, therefore, that the technological innovations that drove the Industrial Revolution—coal, steam, and the techniques of factory production—were biased against skill, tending to replace skilled artisanal with unskilled factory labor (Markovits 2019, 250). Just as early agrarian innovations responded to the balance of sun, water, and land that they might engage, so early industrial innovations responded to the balance of human skill and effort that they might engage. Sterling Bunnell, caustically surveying the industrialized labor force, remarked that while “highly skilled men need little outside of their tool chests” to be productive, “cheap men need expensive jigs” (Bunnell 1914, 611; Montgomery 1979, 188). Bunnell’s ire both disguised and reflected the central lesson of early industrial innovation, namely that the expensive jigs arose, and were shaped, by the prospect that they might be used to exploit the newly available cheap men.

The very same mechanism, moreover, remains in effect today and drives present-day innovation’s overwhelming bias in favor of skilled and even super-skilled workers. The expansion of college education that began in the 1960s, and the later (and still ongoing) intensification of both selection and training at the very top schools and universities, have combined with new social norms that displace leisure with busyness as the badge of honor (Gershuny 2005) and produce an unprecedented reserve of super-skilled and Stakhanovite workers, again waiting for new technologies that might exploit them. Innovators rushed to invent technologies to meet the new supply of super-skilled workers, in ways that are visible in the microhistories of industries and even individual firms. ⁸

In finance, for example, theoretical ideas from the 1950s through the early 1970s (including the capital asset pricing model and the Black-Scholes model for pricing derivatives) were rendered practicable in the late 1970s and 1980s, in order to exploit a surplus supply of super-skilled physicists and engineers who found themselves without academic or other scientific jobs after the U.S. won the space race and détente slowed the arms race (Kaiser, forthcoming; Mulvey and Nicholson 2014). (Older finance workers, made in the less-skilled mid-century mold, called the super-skilled newcomers “rocket scientists,” because they literally were.) In management, consulting firms and business schools invented new ways of running businesses—MIT’s version was called “corporate reengineering,” McKinsey’s was “Overhead Value Analysis”—that openly sought to “foment a stratification within companies and society” (Kiechel 2010, 9) and replace a mass of mid-skilled, middle-class middle managers with a narrow cadre of intensively educated, super-skilled, extravagantly paid elite executives, capable of running large and complex firms directly, without relying on elaborate layers of middle managers, and willing to work with an intensity that the old elite would have found degrading and intolerable (Markovits 2020a).

Across sector after sector, skill-biased innovations tailored to the new elite caused mid-skilled, middle-class workers to lose out to that elite. A superordinate working class arose—one that yields labor hours and effort that resemble and even exceed the 70-hour workweeks of the nineteenth-century industrial working class (Woytinsky 1935; Zeisel 1958) but now, on account of its power and status, captures the gains from its own exploitation and so has grown rich. Once again, the richest 1 percent of households today capture more than double the share of national income that they did in 1970, even as extravagant wages, paid in exchange for super-skilled labor, account for roughly three-quarters of their total income.

These developments—through which the extravagant incomes that superordinate workers now command depend on technological innovations triggered by the rise of the very same super-skilled elite—undermine the second justification for meritocratic inequality. They reveal that even if the elite have worked for rather than been given their status, allegedly meritocratic inequality remains unjustified. Even if the elite are responsible for their hard-won training and the capabilities it confers, these raw capabilities turn out not to be true excellences, of the sort that can legitimate wealth and status. Instead, the outsized wages that elite skills presently command are an artifact of prior inequality, and the seemingly enormous addition that the elite makes to the common good is an illusion, produced by an artificially narrow calculus of contribution—an elite-imposed tyranny of no alternatives (Unger 2009).

Meritocracy aspires to make meritocratic inequality appear natural—almost inevitable. Superordinate workers possess talent and training that others can’t match, and these yield skills that make their labor exceptionally productive. Meritocratic elites, as Mankiw (2010, 295) would have it, “[contribute] more to society” than other workers do, and the market naturally—and rightly—pays people according to “the value of what [they] contributed to society’s production of goods and services.” But raw capabilities depend on context to become valuable skills, and different contexts value different capabilities. Certainly the capacities to evaluate and trade short-term risks, to manipulate intricate systems of formal rules in the service of rationalizing preferred substantive outcomes, and to marshal vast quantities of data to coordinate the efforts of large numbers of workers toward an intricate production process—capacities that make finance workers, lawyers, and business executives enormously productive today—would not contribute much to economic production in agrarian or even early industrial societies. Moreover, these raw capacities have become so valuable in superordinate workers today only on account of the enormous economic inequality that our world contains—inequality without which the technologies that the superordinate workers use their skills to deploy would not have been invented.

This reveals that justifications of meritocratic inequality like Mankiw’s commit a—very short—circle. Superordinate workers are said to deserve their highly unequal incomes because they are exceptionally productive, but the productivity of their raw capabilities turns out to depend on precisely the economic inequality that it is supposed to justify. Homer’s Odysseus may have been “skilled in all ways of contending,” but the meritocratic elite is skilled only at contending in economic inequality’s shadow. And so even if the meritocratic elite has worked hard and diligently—with genuine virtue—to cultivate its raw capacities, it can deserve the outsized returns to these raw capacities only if it deserves the economic inequality that makes these returns so very large. But that inequality is precisely what remains in need of justification. Meritocracy’s effort to produce an elite that supports the common good is self-defeating. Elite “merit” is a sham.

A second argument that merit is a sham gives the historical relationship between skill and inequality a different but overlapping gloss, one that challenges the calculus of contribution that makes elite labor appear to add so much to the common good. Even supposing that superordinate workers deserve to be paid in proportion to their contributions to the common good, it remains to determine what those contributions are. The conventional accounting of contribution—which Mankiw had in mind when he observed that the meritocratic elite contribute more to society than others do—measures each meritocrat’s contribution according to the difference between the social product when she works and when she doesn’t, where everyone else keeps doing the same thing regardless of whether the meritocrat works or not. But the historical dynamic through which superordinate workers induce innovations that favor their own skills invites an alternative accounting, which more accurately measures the meritocrats’ contributions to society, especially taken as a class.

This accounting credits meritocrats with the difference between total output with and without their labor, but now allowing the rest of society to reorganize their work to optimize production in the meritocrats’ absence. The alternative measure of elite executives’ contributions to their firms’ production, for example, looks not to the difference between what firms that have been stripped of middle management can produce with and without top leadership, but rather to the difference between what firms can produce when the management function is concentrated in a narrow executive elite and what they can produce when it is dispersed across the firms’ broader workforce. Insofar as corporate restructurings redirected firm revenues from workers to executives rather than increasing overall productivity, the alternative measure of meritocratic managers’ contributions will be smaller—indeed, much smaller—than the conventional measure.

More generally, elite workers may be massively productive according to the conventional calculus of contribution but not productive at all—indeed, they may have a negative product—according to the alternative measure. This will happen whenever elite workers produce direct gains from their own efforts that are smaller than the indirect losses that they impose by preventing others from working as productively as they otherwise might. And insofar as the meritocratic elite have bent the arc of innovation to favor their own peculiar skills at the expense of alternative innovations that favor middle-class skills, then this is exactly what has happened. Superordinate workers may be essential for productivity using the technologies that now exist, but these technologies are not natural or inevitable but rather induced by the very elite that they now benefit. At the same time, their rise has blocked the invention of alternative technologies that might allow mid-skilled workers to drive an equally, and perhaps even more, productive economy, one with much lower inequality.

While the precise accounting of the meritocratic elite’s true contribution to the common good—the precise net result of balancing superordinate workers’ direct contributions against the losses they cause by suppressing the contributions of all other workers—necessarily remains speculative, there are good reasons to believe that it is in fact very small, or even zero. The technologies that concentrate the management function in a narrow executive elite seem not to have improved the overall performance of American firms—although they have increased CEO pay and perhaps also investor returns (Markovits 2019, 267). And the financial innovations that support today’s super-educated, extravagantly paid finance workers seem to have reduced neither the total transactions costs of financial intermediation nor the share of fundamental economic risk borne by the median household (Philippon and Reshef 2007, figure 11). Most generally, total factor productivity did not grow more rapidly in the more intensively meritocratic decades after 1980 than in the less meritocratic decades before 1970 (Federal Reserve Bank of St. Louis, n.d.). Again, even if superordinate workers owe their raw capabilities to their own virtuous efforts, these raw capabilities may not in fact contribute much to the common good, properly measured. Even if the meritocratic elite deserve to be paid in accordance to their social product, a morally accurate accounting may shrink this social product, perhaps even to zero.

The ancien régime fell not just because aristocracy’s birthright privilege came to be seen as unfair, but also because aristocrats lost their luster. Aristocratic manners and habits, once admired for combining élan with stability, came to be ridiculed as pretense—pompous and sclerotic. Most important, the leisure and conspicuously useless extravagance that once marked status increasingly appeared indulgent, sybaritic, and even weak, as industry came to be valorized and busyness became the badge of honor. Aristocracy, that is, came to be seen not just as unfair but also, and more important, as wasteful, indulgent, and ridiculous.

The arguments just rehearsed reveal that meritocracy is vulnerable to an analogous attack. These arguments unmask merit much as critiques of the ancien régime unmasked aristocratic pretense: They show that raw capacities that meritocrats cultivate in school, deploy at work, and celebrate as the foundation for their income and status don’t actually promote the general welfare and aren’t in fact virtues at all.

The arguments, moreover, formalize a sensibility that precedes them. They elaborate and render theoretically articulate complaints that popular culture and politics—in particular, the rising populist anger and resentment against elites and their institutions—already intuits. Nowhere is this clearer than in the current wave of attacks on universities, and especially elite ones. These attacks recognize that elite university admissions are unfair and that elite graduates’ dominance in the competition for the highest-paying jobs compounds the unfairness. But the center of the popular case against universities lies elsewhere, as critics emphasize not who gets in but what universities teach to those who do get in and what values universities promote in society more broadly. Attacks on “woke” elites draw from many sources, including male resentment at the empowerment of women and white resentment at rising racial diversity. But a central source of the attacks’ broader political power is a widespread sense that elite culture and values (especially inside elite universities) are self-indulgent and self-serving—that elite institutions (most especially in finance, law, and education) promote elite interests instead of, and even against, the common good. Even the focus on purifying elite institutions of nonmeritocratic biases—on diversity and inclusion—carries an odor of self-dealing. The meritocratic elite’s extreme vigilance against prejudice reflects sincere, profound, and salutary moral convictions. But the siren call of an elaborate and fragile identity politics also serves a meritocratic elite’s self-interest, as a conspicuous openness to including accomplished people regardless of their diverse backgrounds helps both to launder the economic advantages that helped build these accomplishments and to legitimate the further economic advantages that inclusion confers. Together, these attacks frame meritocratic institutions as a conspiracy of the privileged and the excluded, aimed against the middle class. And they frame merit as an ideology designed to launder elite self-dealing through false claims about serving the common good.

Populists, following these logics, aim their resentments not at the remaining vestiges of the ancien régime—neither at the hereditary superrich whose aristocratic privilege has survived nor at entrepreneurial oligarchs who aspire to amass enough wealth to found new hereditary lineages—but rather at the professional classes and the meritocratic institutions through and in which they rise and thrive. Donald Trump accepts Mark Zuckerberg and Jeff Bezos and embraces Elon Musk, even as he attacks Harvard and Kirkland & Ellis. The populist posture gains traction because it reflects a profound social logic. Even if the welcome offered oligarchs is corrupt, the superrich play only a glancing role in most middle-class people’s lives, as a symbol or even just a fantasy. And even though the attacks on elite institutions are malign (and motivated by authoritarian prerogative rather than concern for the middle class), they exploit genuine vulnerabilities—that the meritocratic elite has itself become a special interest. Authoritarian populism, in other words, isn’t simply an eruption of malevolence; it is also an apt reaction against meritocratic inequality.

A pattern recognized by development economists weaves these two chronicles of self-defeat into a single story. It commonly happens that countries that are rich in natural resources—oil, diamonds, gold, and so on—flounder even as countries without natural resource wealth flourish. The Democratic Republic of Congo and Singapore were roughly equally poor when the each became independent countries in the 1960s. But the Democratic Republic of Congo, which has among the world’s richest mineral deposits, remains one of the poorest and most unstable countries in the world today, with a GDP per capita of less than $1,000. Meanwhile Singapore, which has no natural resources at all (and became independent only by expulsion from Malaysia), is today one of the world’s richest and most stable countries, with GDP per capita above $60,000. These are extreme cases, but the examples may be multiplied—so commonly that development economists have come to speak of a “resource curse.”

Resource-rich countries might flounder for many reasons: Mineral wealth makes a country an attractive target for conquest and colonial extraction, for example. But one important class of reasons concerns the effects that natural resources have on a country’s internal social, economic, and political arrangements. Resource-rich countries focus their collective lives on extractive industries—drilling, mining, and so on. These industries concentrate wealth in a narrow landowning elite and require a large mass of subordinate workers to do the arduous and dangerous tasks of extraction, and they therefore discourage investment in education and the development of a flourishing middle class, both of which produce domestic powers that might resist and even threaten the landowning elite. These patterns, in turn, suppress widespread political participation and democratic accountability and encourage oppression and corruption. Slow growth and political instability naturally follow.

The meritocratic processes just described inaugurate a new and historically unprecedented version of the resource curse—in which the cursed resource isn’t oil, gold, or any physical capital but rather the skill and effort of free workers, i.e., human capital. Meritocracy distorts the societies and economies that embrace it, inducing technological innovations that concentrate production, and therefore also income and status, in a narrow superordinate working class. In this way, meritocracy in effect concentrates economic life around extractive industries, only with the variation that the resource from which they extract their income isn’t physical but rather human capital. At the same time, meritocratic elites, as ownership castes always do, close ranks to concentrate human capital within their own narrow class. The labor market concentrates work, income, and status in a narrow class of super-educated superordinate workers, who dominate an underemployed and stagnant middle class. And the elite uses its enormous incomes to buy its children schooling that nobody else can afford, so that rich children monopolize the spots at top school and elite universities and then monopolize the top jobs in the next generation. Each of these patterns sustains and entrenches the other, through a series of feedback loops.

The pathologies that plague resource-rich societies naturally follow, as meritocratic inequality grows and snowballs, producing not just economic but social stratification, undemocratic politics, corruption, and slow growth. Meritocracy casts a human resource curse.

Conclusion

Meritocracy is self-defeating because meritocratic successes distort the social and economic order in ways that undermine the values that give meritocracy its appeal. Superordinate workers devote their income and status to outsized investments in education, and investments in human capital establish a new technology for the dynastic transmission of privilege down through the generations of meritocratic families. In this way, meritocracy undermines the equality of opportunity that it was embraced to promote. At the same time, superordinate workers bend the arc of innovation so that new technologies complement their own skills and substitute for everyone else’s, so that meritocracy produces an elite that does well without doing good. This undermines the confluence of elite accomplishment and the general interest that meritocracy was embraced to promote.

It is tempting for meritocrats—given meritocracy’s considerable charisma and their equally considerable self-regard—to respond to growing meritocratic inequality and rising populist discontent by doubling down on meritocracy. Such a reform program seeks to purify our economic and social life of its remaining aristocratic elements—for example, by eliminating legacy admissions preferences at elite universities and by policing monopoly power, insider self-dealing, and oligarchic corruption in commercial life. This program isn’t mistaken in any straightforward way: Its targets are all immoral and inefficient, and the reforms it contemplates will make society fairer and more prosperous.

But insofar as meritocracy is self-defeating, such meritocratic reforms will not seriously reduce the economic and social inequalities that lie at the root of populist discontent, because these inequalities result from meritocracy itself—from the achievement rather than the shortfall of meritocratic institutions. And any response to populist authoritarianism that seeks merely to correct such shortfalls will quickly discredit itself, as its successes will leave the gravamen of the populist complaint unaddressed. Insofar as meritocracy is self-defeating, more meritocracy can’t cure, but will instead exacerbate, the current failings. Instead, our institutions will survive the populist attack only if they can overcome their meritocratic exclusiveness to remake themselves in a fundamentally more democratic image.