The Meritocracy Trap

by Daniel Markovits

  The Industrial Revolution’s bias against skill—early industrial technology’s tendency to replace artisanal with factory production—shows this effect in action. Over the first half of the nineteenth century, England witnessed an immense and utterly unprecedented migration of unskilled workers from the countryside (and from famine-stricken Ireland) into its cities. Between 1811 and 1911, Greater London grew from around 1 million to over 7 million inhabitants, Greater Manchester grew from around 400,000 to 2.5 million, Greater Birmingham from around 250,000 to 1.75 million, and Greater Liverpool from around 150,000 to 1.4 million.

  The central innovations of industrial production all targeted and exploited (and at the same time stimulated) this new labor source. The new style of production used standardized outputs, composed of interchangeable parts, to fragment previously integrated manufacture into discrete steps. This allowed unskilled workers, doing simple repetitive tasks coordinated by industrial engineers, to make goods whose production previously required the integrated efforts of a skilled artisan. Along the way, the innovations displaced older artisanal methods and the highly skilled workers who once deployed them.

  Early industrial technology’s bias against skill therefore responded directly to the shape of the human resources that it engaged—to the balance of skilled and unskilled labor available in industrializing England—just as early agricultural technologies responded to the balance of natural water resources that they might engage.

  Present-day innovation’s overwhelming bias in favor of skilled workers reflects the same mechanism. Only now the incentives to innovate push in the opposite direction. The boom in college education beginning in the 1960s produced a steep increase in the supply of skills, and the subsequent (and still ongoing) rise of an extraordinary concentration of training provided by the very top schools and universities produced a still steeper rise in super-elite skills and superordinate workers, capable of performing tasks of unprecedented complexity. At the same time, the transformation of social norms to celebrate hard work and deplore leisure primed newly super-skilled workers to deploy their training with intense industry. Moreover, the number and tenacity of the new elite workers allow technologies that mix with super-skilled labor to build on one another.

  These trends provided economically fertile ground for new technologies that might productively be mixed with intensely effortful, super-skilled labor. Interested innovators responded to the new terrain—to the new labor supply that would make skill-biased innovations profitable. Their inventions focused on, and even fetishized, the newly available elite skills. At the same time, they neglected the more ordinary skills on which production had previously relied. Capital, naturally attracted to superordinate labor that it might exploit, funded the innovations. (The concentration of venture capital firms in Silicon Valley is just the most obvious instance of this phenomenon.)

  Innovation acquired its now-famous bias in favor of skill not out of the blue, nor even on account of technology’s necessary logic, but rather by tracking, even chasing, the new supply of skill that meritocratic education unleashed. The skilled-biased innovations—in finance and management, and also in retail, manufacturing, and across the economy—refocused production around super-skilled labor. This suppressed middle-class wages and increased elite ones. And meritocracy’s trademark high-end inequality ensued.

  Writing at the close of the Industrial Revolution, Sterling Bunnell remarked caustically that “highly skilled men need little outside of their tool chests” to be productive, while “cheap men need expensive jigs.” Innovation’s recent career turns this remark on its head. Today, not just in finance and management but across the broader economy, expensive workers induce innovators to invent jigs that make other workers cheap.

  The growth in the supply of skill accelerated beginning in roughly 1970, as the post–World War II baby boom and the midcentury investment in college education intersected to produce an unprecedentedly large cohort of college graduates. The college wage premium unsurprisingly dropped precipitately in the subsequent decade, as the new supply of college skills outstripped the existing demand for them. But then, entering the 1980s, the wage premium took a surprising turn, increasing rapidly and almost without interruption through the present day. Moreover, the premium continued to rise steeply in subsequent decades, even as the relative supply of college skills also increased (with only a modest reduction in the pace of growth). This suggests a sudden, rapid increase in the demand for college skills, beginning roughly a decade after the sudden rise in the supply of college-educated workers. (An independent estimate finds that the demand for college skills grew more than one and a half times faster in the 1980s than over the previous four decades.) The best explanation for this pattern—for the lagged timing of the rising demand for college skills—is that the rising supply of college-educated workers induced the innovations that would make their skills valuable and raise the wage premium that they enjoy.

  These patterns and relationships replay themselves at the very top of the distribution also, in the premium captured by super-skilled workers with elite BAs or professional degrees. Meritocratic reforms at top universities, together with the elite’s newfound embrace of industry instead of leisure, caused the supply of super-skilled labor to explode beginning in the early 1970s. Once again, the new supply initially suppressed the returns to super-skill, and the top 1 percent’s income share reached its bottom not in the middle but rather at the very end of the midcentury era, falling from around 12 percent in the late 1960s to 10.4 percent in 1976. But then top incomes grew rapidly, beginning in the late 1970s and accelerating dramatically in the early and mid-1980s (so that by 1988, the top 1 percent’s income share had increased by half to reach nearly 15 percent) and continuing to rise into the new millennium. The broader super-elite economy extends the microhistories of finance and management rehearsed earlier. The best explanation for this pattern, once again, is that the innovations that biased work and wages in favor of super-skills were induced by the new supply of super-skilled workers that the meritocratic revolution in elite education created.

  Finally, international comparisons corroborate the historical lesson that connects skill-biased innovation to the rise of a super-skilled workforce. The division of labor into gloomy and glossy jobs is starker in the United States than in other rich countries, and elite education is more intensive and exceptional. Meritocratic inequalities at work and at school feed on each other.

  Germany is, after the United States, the second-richest large country in the world, with a population of roughly eighty million and a GDP per capita of about $50,000. Indeed, the United States and Germany are (by one measure) the only two countries in the world with populations greater than fifty million and per capita GDPs greater than $50,000. But in spite of their being the only members of this exclusive club, the recent histories of education and work in Germany and in the United States have taken nearly opposite paths. The interconnected differences open a window into the relationship between training and labor, and in particular between elite education and the economic returns to super-skill.

  On the one hand, U.S. and German education targets different populations by different means. The United States has concentrated its educational investments in an increasingly narrow elite. And it has delivered education increasingly in university settings, effectively eliminating workplace training. Germany, by contrast, has spread education increasingly broadly, over a larger and larger segment of its population. Moreover, all members of the broad German elite receive effectively equivalent educations: Germany has virtually no private schools or universities, and while there are elite faculties in the German public university system, there are virtually no exceptionally competitive or distinctively elite student bodies. Germany also provides intensive vocational training to those outside of the university-educated elite. Finally, the German state promotes egalitarian education from earliest childhood and backs its commitments by law
. In Berlin, for example, the city government has gone so far as to enact an ordinance making free daycare available to all city residents and making elite daycare effectively illegal, by forbidding daycares from subjecting parents (no matter how rich) to surcharges greater than 90 euros a month.

  On the other hand, American and German employers have in recent decades focused new investments and innovations on very different segments of the labor market. American firms allocate new investments in plants and machinery disproportionately to complement high-skilled workers. German firms, by contrast, channel new capital toward sectors in which unskilled or mid-skilled labor dominates production.

  When a firm buys new equipment, this makes the workers who use that equipment more productive and therefore increases their wages. In this way, decisions about whether to invest in equipment used by mid-skilled or by super-skilled workers directly influence the pay of both types, and therefore also the wage premium that accompanies training and skill. Between just 1975 and 1991, for example, new investments—capital deepening—in manufacturing and retail in the United States increased the skill premium by about 8 percent. By contrast, capital deepening in these sectors decreased wage differentials in Germany. Overall, capital deepening in the United States is associated with increased wage dispersion and rising returns to skill, while in Germany it is associated with wage compression and falling returns to skill. Indeed, the effect was felt even in banking. American banks chase elite talent and have become some of the most economically unequal workplaces in the world. But in Germany, new investments in banking chased mid-skilled workers and decreased wage inequality in that sector by fully a third.

  Adding other countries into the mix reinforces the connection between inequality at school and at work and reemphasizes that the U.S. elite’s exceptional schooling and exceptional wages feed off of each other. Across developed economies, the college wage premium and the gap between elite and middle-class educational investments tend to rise and fall in tandem.* Where education concentrates training, firms focus investments on super-skilled workers and inflate the skill premium, and where education disperses training, firms focus investments on mid-skilled workers and reduce the skill premium. Meritocracy at school and meritocracy at work go together.

  A meritocratic working elite, by its very existence, stimulates the demand for its own skills. The skill fetish that dominates the labor market today is a contingent and designed response to meritocratic developments in elite education. The elite then spends its enormous incomes on exceptional educations for its children, whose own skills induce further skill-biased innovations. And the cycle continues. One prominent commentator has gone so far as to speculate—more suggestively than literally—that “the Vietnam War draft laws and the high college enrollment rates of the baby boom cohorts . . . induced the development of computers.” Simply put, glossy jobs were created in response to the emergence of super-educated workers who might do them.

  Meritocratic inequality grows by feeding on itself.

  THE HUMAN RESOURCE CURSE

  If society is a ladder, then opportunity to reach the top can be unequal in two ways.

  First, a person’s odds of reaching higher rungs can depend on the rung at which she starts. When meritocratic education concentrates training in children of rich parents and skews elite student bodies toward wealth, it damages equality of opportunity in just this way. Rich children exclude middle-class children from elite schools, and superordinate workers render mid-skilled workers redundant. Elite opportunities are middle-class obstacles, and the elite blocks the middle class on account of realizing rather than betraying meritocratic ideals.

  Second, the value of any given climb depends on how far apart the rungs of the social and economic ladder lie—on how large the absolute differences of income and status across social classes actually are. It is one thing for a person to be confined to his birth rank in a narrowly compressed economic distribution, in which the classes lead materially and socially similar lives. It is quite another to be confined in a widely dispersed society, in which even adjacent ranks experience material and social conditions that render their lives mutually unrecognizable. When meritocracy polarizes work, replacing middle-class jobs with gloomy and glossy ones, it increases the separation between the rungs of the ladder. Not just individuals but entire castes come apart. This second development makes inequalities of opportunity in the first sense much worse.

  The feedback loops that drive meritocratic inequality forward connect these two failures. The mechanisms through which meritocracy stretches the social and economic ladder, to increase the gaps between the rungs, also cause a person’s chances of climbing to depend on where she starts. (The division between gloomy and glossy jobs causes elites to give their children the extraordinary investments needed to get the glossy jobs.) And the mechanisms through which meritocracy makes a person’s chances of climbing the ladder depend on where she starts also drives the rungs of the ladder farther apart. (The rise of an elaborately educated elite induces innovations that bias work and income to favor the skills that this elite possesses.)

  Together, these two effects have a devastating impact on absolute social mobility—the odds that a person will enjoy a greater income than her parents did. As with so many of meritocracy’s burdens, the greatest decline is again concentrated in the broad middle class.

  Virtually the entire generation born in 1940, which came of age during the postwar middle-class boom, grew richer than its parents. In this respect also, all of America resembled St. Clair Shores. A baby boomer had to be born into the richest tenth of households before her odds of becoming richer than her parents fell below nine in ten, and she had to be born into the richest 1 percent before her odds fell below half.

  The generation born in 1980 faced a much bleaker future. Only the poor had a better than even chance of growing richer than their parents. Moreover, and crucially, the intergenerational decline in absolute mobility is by far biggest for children whose parents’ incomes lie between roughly the 20th and roughly the 95th percentiles of the income distribution—that is, for the (very) broad middle class. This is the group whose opportunities were most expanded by open and democratic education at midcentury and whose outcomes most benefited from shared economic growth.* The broad middle class is also, of course, the group whose complaints of exclusion ring loudest and most discordantly today.

  This lens focuses attention on the caste order that meritocratic inequality produces. Elite privilege and middle-class exclusion are not just individual but also, and essentially, collective affairs. The comprehensive divide described earlier in static terms also has a dynamic expression. When meritocracy guides innovation to favor meritocrats, the elite closes ranks. Exclusive education and the fetish for skill favor individual meritocrats, and the feedback loop between meritocracy’s two movements favors the meritocratic elite as a class.

  Deep social and economic forces drive this process forward. Inequality’s critics commonly accuse elites, having got rich, of pulling the ladder of opportunity up behind them. But this is not quite accurate. At least, it does not capture the core objection to meritocratic inequality.

  The critics are of course right that elite meritocrats are no more honorable than anyone else and therefore engage in all the familiar forms of self-dealing. But the principal mechanisms behind snowballing meritocratic inequality involve individually innocent choices—to educate children, to work industriously, and to innovate—that accumulate and feed on themselves, in ways that cause collective harm. Highlighting individual actions ignores the deeper structures within which people act. Emphasizing personal morality neglects politics.

  A deeper, structural view now reveals that meritocratic inequality reenacts a familiar economic paradox, only in a novel setting.

  Economists have long wondered why countries that are rich in natural resources—such as oil, gold, or diamonds—are often less rich overall than countr
ies with fewer resources. Part of the reason is that natural resources distort the economies of countries that are blessed with them. Such countries focus production on extractive industries, which get the resources through drilling, mining, and so on. These industries tend to concentrate wealth and power in a narrow caste of land and mineral owners, and they often also require a large class of oppressed labor to do the arduous and dangerous work of getting the wealth out of the earth.

  Resource-rich countries therefore fail to invest in mass education and even suppress commerce and the professions, and they never develop a productive and dynamic middle class. They tend to develop undemocratic and even corrupt social and political institutions, designed to protect the private interests of their powerful elites at the expense of the public good. As a result, resource-rich countries grow less quickly than resource-poor ones—not always, but often enough that economists speak of a resource curse.

  The feedback loops that drive meritocratic inequality forward present an unprecedented version of the resource curse. In the United States today, the cursed resource is not oil, gold, or diamonds, or any other kind of physical wealth, but rather human capital. The exceptional skills of superordinate workers distort economies that rely on them. Concentrated human capital induces innovations that refocus production on industries and jobs—finance, elite management—that use superordinate labor. And these industries concentrate wealth and power in an increasingly narrow caste of meritocratic workers, doing glossy jobs, who dominate a large class of subordinate workers doing gloomy jobs. They are in effect extractive industries, with the twist that they extract income not from natural wealth but rather from the human capital of the superordinate workers.