The Meritocracy Trap

by Daniel Markovits

  The new managerial elite therefore possesses elaborate educations. Executives at midcentury might (like Ed Rensi) have worked their way up a firm’s hierarchy instead of attending university. But today’s top managers (like McDonald’s current CEO) come from elite backgrounds and typically hold MBAs or have equivalent postgraduate training. And when top executives need managerial assistance today, they look not to middle managers but rather to management consultants: external advisers who enable firms to contract out management functions once performed internally. Consultants are again intensively educated and hyper-elite: the leading firm, McKinsey & Company, brags of its “university-like capabilities,” and illustrates this by observing that its proprietary research enables the firm “with the push of a button, [to] identify the top 50 cities in the world where diapers will likely be sold over the next ten years.”

  Finally, the technological advances that concentrate management in a narrow executive elite also inflate that elite’s economic value. As top executives monopolize the management function, and firms come to depend on them for internal coordination, they capture virtually all of management’s economic returns. Income streams that were at midcentury shared widely across all of a firm’s middle-class middle managers become concentrated in elite executives. The status and income that CEOs and other elite managers—including at McDonald’s—now command places them right next to financiers in meritocratic inequality’s pantheon.

  In typical recent years, the single highest-paid American CEO has taken home almost $100 million, and the mean incomes of the two hundred highest-paid CEOs have been roughly $20 million. The three-hundred-times-median work incomes that large-firm CEOs capture today is about fifteen times the CEO-to-worker income ratio from 1965. A slightly broader managerial elite, which includes the rungs just below the top, also captures massive incomes: as when the five highest-paid officers of S&P 1500 firms—seventy-five hundred workers overall—collectively receive income equal to 10 percent of the entire S&P 1500’s profits.

  The corporate reorganizations that decimated the middle class therefore did not simply improve American business, encouraging management to shape up, becoming lean and fit. Instead, they reconstituted firms, inserting new forms of hierarchy that have made American management, in a memorable phrase, fat and mean.

  This third age of American management reverts, in important ways, to the artisanal model of work embraced by the first. Both production workers and managers understand themselves as tied not to a particular employer but rather to a set of tasks and skills. And a renovated guild system—constructed out of the schools and universities that dominate elite education and training today—provides the skills that pair workers with tasks and determine status and pay.

  Management, which played a central role in building the midcentury middle class, eviscerates the middle class today. Many workers, often not even employees, do gloomy production jobs at the bottom; while at the top, a few superordinate executives, operating in delayered organizational hierarchies, exercise vast powers of command in jobs that the new technologies have polished to a bright and glossy sheen.

  THE HOLLOW MIDDLE

  Finance and management are not the only industries to succumb to these new divisions at work. Other sectors of the economy have also shed mid-skilled jobs and polarized their workforces.

  Midcentury retail, for example, was dominated by small, independent stores. In 1967, single-store firms still made 60.2 percent of all sales (and large chains accounted for only 18.6 percent). These stores, in turn, employed mid-skilled workers. As the New York Times observed, in 1962, “In the small independent establishment where the store owner doubles as sales clerk as well as buyer, he often will produce a high rate of selling productivity.” Alternative models, the Times also noted, found it hard to compete: “In the larger establishments, retailers often find themselves in trouble because of a general low quality of personnel.”

  Today, retail is dominated by massive chains of enormous stores with names familiar to anyone who shops: Dollar General, Family Dollar, Walgreens, CVS, 7-Eleven, Kroger, and of course Walmart and Amazon. The large chains deploy new technologies for selling that eliminate mid-skilled workers and replace them with a gloomy mass and a glossy elite. Street-level, subordinate retail workers now perform narrow and overwhelmingly menial tasks: as shelvers, checkout operators, janitors, or even as greeters (at Walmart), and as effectively mechanized warehouse workers (at Amazon). They are paid commensurately poorly: Walmart (now the nation’s largest employer) pays a median wage of just $17,500 according to one estimate, and $19,177 according to another; and the firm unsurprisingly has full-time employees who fall below the poverty line and rely on public assistance, including through holiday food drives hosted, ironically, by Walmart’s stores. An Oklahoma store, for example, collected cans of food in a bin labeled “let’s succeed/by donating to associates in need.” Meanwhile, in 2017 Walmart’s CEO made 1,118 times as much as the firm’s median worker.

  Superordinate retail workers deploy new technologies to centralize the tasks that successful selling requires—ranging from big-data-driven analyses of shopper behavior now provided by firms like Percolata of Silicon Valley, to price optimization programs that give discounts where consumers notice them and hike prices where they do not, to branding techniques that help customers identify goods without in-store assistance. The elite workers who develop and administer these innovations are, unsurprisingly, intensively educated and enormously skilled. Jeff Bezos, who is the founder and CEO of Amazon and the richest person in modern history, graduated summa cum laude and Phi Beta Kappa from Princeton and in the firm’s early days recruited employees from among American Rhodes Scholars studying at Oxford.

  Mid-skilled clerical jobs—for example, telephone operators, typists and word processors, travel agents, and bookkeepers—have also been disappearing, as elites use computers to do some clerical tasks themselves and farm others out to now-subordinate data entry workers. Law firms have over the past fifteen years shed over one hundred thousand support jobs for workers with less than two years of college even as they have added the same number of jobs for workers with JDs and BAs. And computer-assisted design programs displace mid-skilled draftsmen and empower super-skilled architects and engineers to produce more intricate and inventive designs.

  No sector is immune. Even in the arts and entertainment, new technologies allow a few “superstar” entertainers to capture global audiences, displacing many only slightly less skilled performers who previously served local audiences by being the best entertainment within traveling range. In 2017, Beyoncé, LeBron James, and J. K. Rowling each made nearly $100 million. This is perhaps one hundred times more than their midcentury counterparts. It is roughly one thousand times more than backup singers, players in the NBA’s development league, and television scriptwriters—all skilled people, near but not at the tops of their professions—make today.

  Finally, of course, technology has transformed manufacturing. Conventional wisdom emphasizes that technology has destroyed the traditional mid-skilled manufacturing jobs that helped to build the midcentury middle class, not just in St. Clair Shores but around the country. General Motors, the country’s largest employer at midcentury, paid its unionized workers $60,000 per year, plus substantial benefits. Today, the U.S. auto industry deploys over twelve hundred robots per ten thousand employees, and the trend toward robot production is accelerating. (Robots play more prominent roles yet in manufacturing in Europe and Asia.) Overall, the United States has lost nearly eight million manufacturing jobs since the late 1970s. And to employ the same percentage of the American workforce in manufacturing today as worked in the sector in the mid-1960s, the economy would require perhaps twenty-five million more manufacturing jobs than it currently provides.

  At the same time, although less familiarly, the new technologies have created a new group of glossy jobs staffed by super-skilled industrial workers who design,
program, and manage automated production processes. Even as overall domestic manufacturing employment fell by roughly a third between 1992 and 2012, the number of manufacturing jobs for workers with college degrees increased by 2.4 percent, and the number of manufacturing jobs for workers with graduate degrees increased by 44 percent.

  These super-skilled workers are more productive—much more productive—than the mid-skilled workers they replace, and their productivity has allowed manufacturing’s share of real gross domestic product to hold steady even as its share of employment declined. They are also better paid: between 2007 and 2012, the average income of manufacturing workers increased by over 15 percent. In extreme cases, the shift from mid-skilled to glossy jobs can rival the effect in finance. Kodak at its peak employed 140,000 mid-skilled light manufacturing workers making cameras and film; and its founder, George Eastman, famously embraced the midcentury middle-class model of work, providing lifetime employment and extensive opportunities for workplace training and promotion. Today, the firm has been effectively displaced by digital alternatives such as Instagram, which employed a grand total of thirteen super-skilled workers when it was sold to Facebook for $1 billion. These thirteen became fabulously wealthy, of course.

  All these examples (and many others not recounted here) tell and retell the same basic story, as many variations on a single theme. The democratic regime that governed the midcentury American workplace has yielded to meritocratic inequality today. A raft of technological innovations has eliminated the mass of mid-skilled, middle-class jobs that once dominated production and replaced these jobs with varying combinations of gloomy and glossy ones.

  The economic sector most closely associated with middle-class work at midcentury—manufacturing—has suffered a steep decline in absolute and proportional employment, and the sectors most closely associated with both subordinate and superordinate work today—retail on the one hand and finance on the other—have experienced massive expansions. In addition, work has polarized within each sector: mid-skilled loan officers yield to subordinate clerks and superordinate analysts; middle managers yield to subordinate contract workers and superordinate executives; mid-skilled independent retailers yield to massive chains that employ subordinate checkout cashiers and superordinate e-commerce software developers; and mid-skilled tool- and diemakers yield to robots and superordinate engineers.

  Aggregate data confirm and quantify the overall hollowing out of the middle of the labor market.* In the three and a half decades beginning in the early 1980s, the share of all jobs principally devoted to mid-skilled tasks has fallen at a steep and accelerating rate: by roughly 5 percent over the 1980s, roughly 7 percent over the 1990s, and nearly 15 percent since 2000. Over the same period, the share of jobs principally devoted to super-skilled tasks has risen steeply, by nearly 10 percent per decade, and the share principally devoted to unskilled tasks has also grown, mostly since 2000.

  Altogether, fully a quarter of the economy’s mid-skilled jobs have disappeared since 1980, and the share of jobs allocated specifically and exclusively to high-skilled workers has increased by more than a third, while the share of the overall workforce composed of technical and professional workers has more than doubled since 1950 and is today nearly 20 percent. Moreover, state-by-state comparisons suggest that the ascent of the rich and the stagnation of the middle class go together—that rising top incomes (as measured by the richest 1 percent’s share of total income) produce falling middle-class incomes. In addition, the trends are not peculiar to the United States. Rather, meritocracy undermines democratic equality at work throughout the world’s rich societies.

  These trends only grow stronger as the years advance. Low- and high-wage employment, once again, both increased during the first decade of the new millennium even as mid-wage employment declined. A similar pattern arose during the Great Recession and the subsequent recovery: with middle-class jobs accounting for three times as large a share of recession losses as recovery gains, while subordinate and superordinate jobs each accounted for greater shares of gains than of losses. In addition, the Bureau of Labor Statistics predicts that over the coming decade, the fastest-shrinking job categories will all be mid-skilled, and the ten fastest-growing will all be either low- or super-skilled. The McKinsey Global Institute—the consulting firm’s research arm—forecasts an even more dramatic transformation, predicting that nearly one-third of the U.S. workforce, overwhelmingly in mid-skilled jobs, will be displaced by automation by 2030.

  These developments, taken all together, constitute not a ripple but a tidal wave—even a sea change. The labor market has, bluntly put, abandoned the midcentury workforce’s democratic center, and this has fundamentally transformed the nature of work.

  Whereas work once underwrote midcentury America’s apt self-image as an economy and society dominated by the broad middle class, work today underwrites the equally apt sense of a rising division between the rich and the rest. At midcentury, work united Americans around a shared, democratic experience, and the unionized middle-class autoworkers at General Motors embodied the labor market. Today, work divides Americans, in a labor market epitomized by Walmart greeters and Goldman Sachs bankers.

  TRAINING PAYS OFF

  The democratic workplace that dominated the midcentury economy suited the habits of the American middle class. The many mid-skilled jobs and the associated opportunities for training that once characterized the workplace connected workers across ranks and skill levels. When pinboys in St. Clair Shores could walk into lifetime employment—at middle-class wages and with workplace training and opportunities to advance—elaborate, competitive, and elite educations had no compelling purpose.

  By contrast, the meritocratic workplace that dominates the economy today suits the habits of the American elite. The hollow middle of today’s labor market segregates workers by type, and in particular isolates super-skilled workers from all others. Rich families run the gauntlet of elite education (which has displaced continuous workplace training in favor of discreet degrees, provided in universities) to give their children the exceptional training and skills required to get and to do superordinate jobs, so that they might land on the right side of the meritocratic divide.

  Elite training succeeds. The investments that rich families make in their children’s human capital pay off. Children from the richest fifth of households are roughly seven times more likely than children from the poorest fifth to end up in the top quintile of the income distribution as adults, roughly nine times more likely to end up in the top quintile of the wealth distribution, and roughly twelve times more likely to end up in the top quintile of the education distribution.

  Education has become the labor market’s preferred sorting mechanism, and the economic returns to schooling, especially at the best schools, have become astronomical. Education, that is, nearly perfectly maps the fault line that separates subordinate from superordinate workers in the newly polarized labor market, sorting workers into income classes that barely overlap. Intensive educations and glossy jobs run quite generally together; meritocratic inequality makes elite students and superordinate workers one and the same.

  The completeness of the segmentation amazes.* The median college graduate will make more money over his lifetime than 93 percent of workers without a high school degree and than 86 percent of workers with a high school degree only; and the median professional school graduate will make more money than nearly 99 percent of high school dropouts, 98 percent of workers with a high school degree only, and 83 percent of workers with a BA only. This means that only about one worker in fifty from the bottom half of the educational distribution makes more than the median worker from the top tenth.

  The absolute numbers at stake are massive. The median male worker with no more than a high school degree makes about $1.5 million over a lifetime, the median male college graduate makes about $2.6 million over a lifetime, and the median male professional degree holder makes more tha
n $4 million over a lifetime. For women, the relevant numbers are $1.1 million, $1.9 million, and just over $3 million. The absolute numbers are also relatively large—compared both to America’s own past and to other rich countries in the present. The college income premium is perhaps twice as large today as it was in 1980, and the present discounted value of a BA, net of tuition fees, is nearly three times greater today than it was in 1965. (The purely economic rate of return on university education—one prominent estimate yields a 13 to 14 percent return to a year of college—about doubles the long-run returns provided by the stock market.) The college premium is one and a half times bigger in the United States than in Britain and France and three times bigger than in Sweden.

  As with grades and test scores, the top/middle lifetime income gap far exceeds the middle/bottom gap. A closer look at still more refined educational elites amplifies this pattern. BAs from even modestly higher-ranked schools boost incomes by 10 to 40 percent more than BAs from lower-ranked schools and nearly double the rate of return on the tuition. Super-elite BAs generate still greater income boosts, more than doubling the gains produced by an average BA, and the top incomes from super-elite schools more than triple the incomes of the top earners with average BAs. (The highest-paid 10 percent of Harvard graduates average salaries of $250,000 just six years after graduation.) A recent broader survey reports—incredibly—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 twelve universities.