2

The Foundations of the New Upper Class

In which are described the conditions that led to the emergence of the new upper class.

FOUR DEVELOPMENTS TOOK us from a set of people who ran the nation but were culturally diverse to a new upper class that increasingly lives in a world of its own. The culprits are the increasing market value of brains, wealth, the college sorting machine, and homogamy.

At the Bottom of It All: The Increasing Market Value of Brains

In the early 1990s, Bill Gates was asked what competitor worried him the most. Goldman Sachs, Gates answered. He explained: “Software is an IQ business. Microsoft must win the IQ war, or we won’t have a future. I don’t worry about Lotus or IBM, because the smartest guys would rather come to work for Microsoft. Our competitors for IQ are investment banks such as Goldman Sachs and Morgan Stanley.”1 Gates’s comment reflected a reality that has driven the formation of the new upper class: Over the last century, brains became much more valuable in the marketplace. The evidence for that statement took two long chapters to present in The Bell Curve, but the reasons why it happened are not mysterious.2

The Effect of Cognitive Ability on Vocational Success

Cognitive ability is only one of many factors that explain why some people rise to the top of their professions. Assets such as industriousness, motivation, self-discipline, and interpersonal skills play crucial roles. But that truth is easily misinterpreted to mean that cognitive ability is unimportant.

The analogy originated by sociologist Steven Goldberg helps keep things in perspective: For the professions, creative work, and the management of large and complex organizations, cognitive ability plays the same role in determining success that weight plays in determining the success of offensive tackles in the National Football League. The heaviest tackle is not necessarily the best. In fact, the correlation between weight and performance among NFL offensive tackles is probably quite small. But to have a chance of getting the job, you had better weigh at least 300 pounds.3 Similarly, the correlation of IQ scores with performance among those people who are attorneys, screenwriters, and biochemists is modest. But to be a top attorney, screenwriter, or biochemist, you have to be very smart in the ways that IQ tests measure.

First, the higher-tech the economy, the more it relies on people who can improve and exploit the technology, which creates many openings for people whose main asset is their exceptional cognitive ability. What was someone with exceptional mathematical ability worth on the job market a hundred years ago if he did not have interpersonal skills or common sense? Not much. The private sector had only a few jobs such as actuary that might make him worth hiring. His best chance was to go into academia and try to become a professor of mathematics. His options were not much wider in 1960. What is a person with the same skill set worth today? If he is a wizard programmer, as people with exceptional mathematical ability tend to be, he is worth six figures to Microsoft or Google. If he is a fine pure mathematician, some quant funds can realistically offer him the prospect of great wealth.

Second, the more complex business decisions become, the more businesses rely on people who can navigate through labyrinths that may or may not call upon common sense, but certainly require advanced cognitive ability. Consider the prospects for a lawyer. A hundred years ago, lawyers mostly practiced law for individual clients and made the amounts of money that individuals could afford to pay. Those who were corporate lawyers made corporate salaries—good, but not the stuff of dreams. As the size of business deals grew and regulatory law became more complex, the need for lawyers who never see the inside of a courtroom increased. Today, if a first-rate attorney can add 10 percent to the probability of getting a favorable decision on a regulatory ruling worth hundreds of millions of dollars, he is worth his many-hundreds-of-dollars-per-hour rate. If he can work out the multidimensional issues that enable the merger of two large corporations, he may be worth a commission of millions of dollars. The same thing happened in the financial industry, as technology has made possible new and complex—but also fabulously profitable—financial instruments.

Third, the bigger the stakes, the greater the value of marginal increments in skills. In 1960, the corporation ranked 100 on the Fortune 500 had sales of $3.2 billion.4 In 2010, the 100th-ranked corporation had sales of $24.5 billion—almost an eightfold increase in constant dollars. That kind of supersizing in the corporate world occurred across the range—the corporation ranked 500 in 2010 was about eight times larger than the 500th-ranked corporation in 1960. The dollar value of a manager who could increase his division’s profitability by 10 percent instead of 5 percent escalated accordingly.

To some degree, the demands on the cognitive skills of managers also grew over that half century because of the increasing complexity of choices that often accompanies huge size. But that’s not the main point. Even if the skills required of the manager of a corporate division in 1960 and 2010 were the same, and raw brainpower did not play a more important role, cognitive ability is nonetheless an all-purpose tool. Given the same interpersonal skills, energy, and common sense, the manager with higher cognitive ability has an edge in increasing profitability by 10 percent instead of 5 percent—and that, combined with the larger stakes, also made brains worth more in the marketplace.

The Enabler: Wealth

Given that backdrop, it is no surprise that the people working in managerial occupations and the professions made a lot more money in 2010 than they had made in 1960, and that their growing wealth enabled the most successful of them, the members of the new upper class, to isolate themselves from the rest of America in ways that they formerly couldn’t afford to do. Figure 2.1 shows the median income of families at various points on the income distribution, starting with those at the 25th centile and going all the way up to the 99th centile. The data are based on American families of all races and ages.

FIGURE 2.1. AMERICAN FAMILY INCOME DISTRIBUTION

Source: Integrated Public Use Microdata Series (IPUMS). Unit of analysis is the family. The initial year is 1959 because the income variables in a decennial census or March Current Population Survey (CPS) interview refer to the preceding calendar year.

Shelves of books and academic articles have been written about the material in Figure 2.1. Narratives can be told for every level. Real income for the bottom quartile of American families fell after 1970. The poor didn’t actually get poorer—the growth of in-kind benefits and earned-income tax credits more than made up the drop in pretax cash income—but they didn’t improve their position much either.5 Real family income for families in the middle was flat. Just about all of the benefits of economic growth from 1970 to 2010 went to people in the upper half of the income distribution.

Centiles and Percentiles

Centile means the same thing as percentile. But in a text that often refers to percentages and percentage changes, using centile makes for clearer sentences.

The increase was most dramatic at the very top of the distribution. From 1960 through the early 1990s, the top centile of American families had incomes that began at around $200,000. Then in 1994–95, the bottom end of the top centile careened up from $233,000 to $433,000. Whether the change happened within that single year is open to debate—an analysis using IRS data shows the leap occurring from the late 1980s through the late 1990s—but there is no doubt that a phenomenal growth in top incomes occurred sometime during that period.6 In the March 2010 Current Population Survey (CPS), reflecting income data for calendar year 2009, the 95th centile for working-aged Americans began at $199,000 and the 99th centile at $441,000. And remember: $441,000 is the bottom of the top centile.

The top five centiles are important for our purposes because they contain almost all of the new upper class. It is possible to draw that conclusion because of the nature of accomplishment required to be part of the new upper class as I have operationally defined it. In twenty-first-century American society, meeting that definition also means that you almost certainly have a family income that puts you into the top five centiles of family income ($199,000-plus).

For many of the positions that qualify you for the new upper class, $199,000 is far below the average. If you are a partner in a major law firm or the president of an important university or foundation, you are likely to be making several hundred thousand dollars a year. If you are the CEO of a major corporation, you are making millions per year in total compensation.

Even if you are in a position that doesn’t allow you to become truly wealthy, your family’s income was almost surely close to the top five centiles of income, and often deep into them. As of 2010 a cabinet officer made $191,300, a Supreme Court justice made $208,100, and the Speaker of the House made $217,400. Ordinary members of Congress made $169,300, and deputy secretaries and heads of major agencies made $172,200. And that’s just the salary for the people holding those jobs, not their family’s total income. Many of them have spouses who bring in large salaries as well. Furthermore, people often come to those positions after making a lot of money in their previous careers, and the value of the perks of high government office rival those of CEOs.

Other “poor” members of the new upper class are journalists, academics, and public intellectuals in general. David Brooks calls their plight status-income disequilibrium, a psychological condition that occurs, for example, when an eminent Columbia faculty member goes home after giving his speech at the Plaza Hotel to admiring Wall Street executives. While his audience is dispersing in their limos to their duplex cooperatives on the Upper East Side, he catches a cab home to his cramped apartment near the Columbia campus, his standing ovation still ringing in his ears, only to be told by his wife that the shower drain is clogged and he must take care of it before the children get up for school the next morning.7 Brooks speaks to a disorienting reality that many well-known faculty members, journalists, and guests on the Sunday news shows can relate to. Their status is a lot higher than their income. But these same people have family incomes of more than $150,000 a year even if there’s just one income (the big names make at least that much at major universities, newspapers, and think tanks). Even if they don’t write best sellers, they often supplement their salaries with book advances and speaking engagements. If the spouse is also providing an income, their income can easily reach the top centile. Those who suffer from status-income disequilibrium feel strapped for money because taxes take a big chunk of their income, and most of them live in New York or Washington, where they have to spend much of the rest on housing, child care, and tuition for private schools. But they can feel poor only because their professions so often throw them in contact with the truly rich. With the rarest exceptions, even the poorest members of the new upper class are in the top few centiles of the American income distribution.

These observations are borne out by the incomes reported in the CPS. I have used the top 5 percent of people in managerial occupations and the professions as a working definition of the new upper class. In 2009, the cutoff for the 95th centile of family income for people working in those occupations was $287,000.

Wealth enabled the development of an isolated new upper class. It did so first by enabling the new upper class to become spatially isolated. The price of houses in a neighborhood screens the people who can live there. Even the Columbia professor, poor as he feels and cramped though his apartment may be, is living in a neighborhood priced so that only people in the top few centiles of income can afford to live there. The higher the price that a new-upper-class couple can pay, the more precisely they can define the kind of neighborhood in which they live. The higher the price that they can pay, the more privacy they can buy—in the form of a concierge and security guards in the lobby of their urban apartment building, a literally gated community, or a high-end suburb insulated from the rabble by distance.

Wealth also enabled the development of a distinctive lifestyle among the new upper class. Markets supply demand, and if there is a demand for the goods and services that underpin an alternative lifestyle, the market will provide those goods and services.

The Mechanism: The College Sorting Machine

The initial mechanism whereby people with distinctive tastes and preferences are brought together is the college sorting machine.

Exceptions like Bill Gates and Steve Jobs notwithstanding, almost everyone in the new upper class has finished college. But the simple possession of a bachelor’s degree does not come close to capturing the complicated relationship between education and the nature of the new upper class. The key to understanding why the new upper class has formed and why it has such a distinctive culture is the interaction between high cognitive ability and education in general, and more specifically the interaction of high cognitive ability and elite colleges.

Cognitive Ability as a Natural Incentive for Segregation

The human impulse behind the isolation of the new upper class is as basic as impulses get: People like to be around other people who understand them and to whom they can talk. Cognitive segregation was bound to start developing as soon as unusually smart people began to have the opportunity to hang out with other unusually smart people.

The yearning for that kind of opportunity starts young. To have exceptional cognitive ability isolates a young person as no other ability does. The teenager with exceptional athletic ability who becomes the star quarterback has lots of people who are eager to be his friends even if he is shy or socially awkward. The teenager with exceptional interpersonal ability is one of the most well-liked kids in school—that’s what exceptional interpersonal ability does for you. But the math star who possesses only average interpersonal ability is seen as an oddball. He has just one or two classmates he can talk to about what he’s good at, if he’s lucky, and he may have no one at all. The teenage girl with average interpersonal ability and exceptional verbal ability has the same problem. If she has fallen in love with the poetry of T. S. Eliot, she is hard-pressed to find anyone else who will understand why. Her classmates already don’t get her jokes and are put off by her vocabulary. She knows that if she were to try to talk about Ash Wednesday, she would first get blank stares and then be teased unmercifully.

When cognitively talented children are forced to deal with that situation, they usually find ways to cope. They study topology or read Ash Wednesday in the privacy of their bedrooms. The boy learns to talk about sports with the other guys and the girl learns not to use vocabulary that will attract ridicule. Making that effort often produces surprising results, as the cognitively gifted children realize that the other kids are smarter and more interesting than they had thought. In all cases, the need to make the effort tends to encourage flexibility, maturity, and resilience among cognitively talented youth.

Still, it amounts to one of those things that people are glad they have done, but did only because they had to. Either they figure out a way to fit in or else they are lonely (see Sinclair Lewis’s Main Street for the revenge of the lonely small-town smart boy). Those same young people would have jumped at the opportunity to be around other people like themselves. Over the last half century, the opportunities to do so opened up. The expansion of college education in general played the single most important role in that process. But another development was almost as important.

Cognitive Stratification Among Colleges

Cognitive stratification among colleges occurred extraordinarily fast.8 As of 1950, elite colleges did not have exceptionally talented student bodies. By 1960, they did.

Before World War II, most of the freshmen in an elite college were drawn from the region’s socioeconomic elite—from the Northeast for the Ivy League, the West Coast for Stanford and USC, and the South for Duke and Vanderbilt. Some of those students were talented, but many were academically pedestrian. In a study done in 1926, the average IQ of students at the most prestigious schools in the country, including Columbia, Harvard, Princeton, and Yale, was 117, barely above the 115 that has been the average of all college graduates, and denoting the 88th centile of cognitive ability.9 That same year, the Carnegie Foundation conducted a study of all of Pennsylvania’s colleges and universities, using the same measure of IQ that produced the 117 mean for the high-prestige schools. In Pennsylvania alone, ten colleges had freshman classes with mean IQs that put them at the 75th to 90th centiles, making those classes cognitively indistinguishable from those of the elite schools.

That situation persisted through the 1930s and 1940s. As late as 1952, the mean SAT verbal score (now known as the Critical Reading score) of incoming Harvard freshmen was just 583, above the national mean but nothing to write home about.10 Then came the revolution. By 1960, the average SAT verbal score among incoming Harvard freshmen had jumped to 678. The progenitors of the revolution were aware of how momentous the shift had been. William J. Bender, Harvard’s dean of admissions, summed up the preceding eight years. “The figures,” he wrote, “report the greatest change in Harvard admissions, and thus in the Harvard student body, in a short time—two college generations—in our recorded history.”11 The average Harvard freshman in 1952 would have placed in the bottom 10 percent of the incoming class by 1960.

The same thing happened throughout the college system, as shown in Figure 2.2. The colleges are representative of the ones clustered at various SAT verbal scores. The backdrop is the distribution of SAT scores in 1960 if all eighteen-year-olds had taken the test.

FIGURE 2.2. COGNITIVE STRATIFICATION AMONG COLLEGES AS OF 1960

Sources: Siebel, 1962; College Entrance Examination Board, 1961. Adapted from Herrnstein and Murray, 1994.

Georgia Southern is representative of a school in the second tier of a state university system, while North Carolina State is typical of a large first-tier university as of 1960. Then come successively more selective private schools. But even among them, the differences in the mean verbal score of the incoming freshmen were substantial, with Harvard anchoring the high end.

The stratification became still more extreme during the 1960s. In 1961, 25 percent of Yale’s entering class still had SAT verbal scores under 600. Just five years later, that figure had dwindled to 9 percent, while the proportion of incoming students with SAT verbal scores from 700 to 800 had increased from 29 to 52 percent.12

The situation at the end of the twentieth century may be conveyed through the work of Roger Geiger, who studied how the students in the top five centiles of test scores (1400-plus on the SAT or 30-plus on the ACT) were distributed among the nation’s universities. Using data for 1997, he calculated the approximate numbers of such students who attended the top 35 public universities, the top 35 private universities, and 35 highly selective small colleges in 1997. Figure 2.3 shows the cumulative percentage of students with these high scores who were soaked up by just 105 colleges. I have extended the horizontal axis to 100 percent to give a visual sense of the concentration of such students relative to the entire student population of four-year institutions.

FIGURE 2.3. CONCENTRATION OF TOP HIGH SCHOOL STUDENTS IN 105 SELECT COLLEGES

Source: Geiger, 2002, tables 3.2, 3.3, 3.4.

Together, just 10 schools took 20 percent of all the students in the United States who scored in the top five centiles on the SAT or ACT. Forty-one schools accounted for half of them. All 105 schools, which accounted for just 19 percent of all freshmen in 1997, accounted for 74 percent of students with SAT or ACT scores in the top five centiles.

Given this concentration of academic talent in a relatively few colleges and universities, the original problem has been replaced by its opposite. Instead of feeling sorry for the exceptionally able student who has no one to talk to, we need to worry about what happens when exceptionally able students hang out only with one another.13

What Are the Elite Schools?

There is no shortage of rankings of colleges and universities, with the Barron’s and U.S. News & World Report lists being the most famous. You should think in terms of the usual suspects. At the top are the iconic schools—Columbia, Harvard, Princeton, and Yale. Right behind them (some would say beside them) are consensus high-prestige schools such as Stanford, Duke, MIT, the rest of the Ivy League, and some of what used to be known as the Seven Sisters, followed by another two or three dozen schools whose names are familiar to all parents who aspire to get their children into selective schools. The top twenty-five national universities and top twenty-five liberal arts colleges in the U.S. News & World Report rankings, which can be viewed online, include all of the schools that everyone agrees are elite, plus a few with less glittering reputations.14

The segregation of the college system now means that the typical classroom in a third-tier public university is filled with students who are not much brighter than the average young person in the nation as a whole, whereas the typical classroom in an elite school has no one outside the top decile of cognitive talent, and many who are in the top hundredth or thousandth of the distribution. Both sets of students are technically “college educated” when they get their BAs, but that’s where the similarity stops. The cognitive pecking order of schools is apparent to everyone—to employers looking at applicants’ résumés, to parents thinking about where they want their children to go to college, and to high school students thinking about how to best make their way in life.

Incentives on the demand side have interacted with incentives on the supply side. More and more of the best students want to go to the elite schools, and the elite schools, eager to maintain their status, search ever more assiduously to fill their incoming class with the best of the best. The competition on both sides to achieve the same end has proved to be irresistible. In the early 1990s, when Richard Herrnstein and I were writing The Bell Curve, he sat on the Harvard undergraduate admissions committee. One day when we were on the phone discussing the latest draft, he told me happily that Harvard had snagged more exceptionally qualified students for the next entering class than ever before. “But Dick,” I said, “we’re writing about all the problems that causes.” Herrnstein, who loved Harvard, replied (with a smile, I am sure), “I want ’em all.”

The Dominance of the Upper-Middle Class in Elite Schools

The concentration of high-ability students wouldn’t be so bad if those students had only their ability in common. In the ideal of the meritocracy, the new Yale or the new Princeton is now peopled by all those lonely high school students I was just describing, gathered from small towns and inner cities, the children of bakers and bankers, nurses and insurance agents, showing one another how much they have to learn about the full spectrum of American life.

It didn’t work out that way. The opening of elite schools to the academically talented of all backgrounds was not accompanied by socioeconomic democratization of those schools. On the surface, it looked as if things had changed. The proportion of students coming from socially prominent families dropped. The proportion that came from exclusive prep schools dropped. The de facto quotas on the number of Jews who would be admitted were dropped. Affirmative action increased the representation of African Americans and Latinos on elite campuses. The numbers of Asian American students increased manyfold through the force of their superb credentials.15 But despite these changes, the student bodies of the elite schools were still drawn overwhelmingly from the upper-middle class. According to sociologist Joseph Soares’s analysis in The Power of Privilege, consistent with other such analyses, 79 percent of students at “Tier 1” colleges as of the 1990s came from families in the top quartile of socioeconomic status, while only 2 percent came from the bottom quartile.16 For Soares, these numbers are evidence of obvious bias against the most able students who are not from the upper-middle class and above. “Unless one believes that only rich people can be smart,” he writes, “we have a staggering distance to travel to achieve a fair opportunity for all to reach every level of our educational system.”17

The bias that Soares set out to investigate could occur in two ways. The first is that the pool of applicants is biased. The second is that the admissions process continues to give preferential treatment to the children of the affluent.

Soares presents compelling evidence that the applicant pool is biased. Using the National Educational Longitudinal Study (NELS), Soares demonstrates that the applicant pool is heavily loaded toward youths who come from the high-income professional class, especially those who come from the Northeast United States. In a logistic regression analysis of the NELS data that controls for other influences on the decision to apply, Soares finds that students with the same gender, race, and SAT scores are more than three times as likely to apply to a selective school if they come from one of those professional high-income families in the Northeast, and twice as likely if they come from a professional high-income family outside the Northeast.18 Other things being equal, Asians were almost twice as likely to apply as non-Asians, and students from private schools were four times more likely to apply than students at public schools.19

So the applicant pool is skewed. But from among this pool, are admissions officers giving preferential treatment to those who possess the right cultural capital? Without doubt, certain applicants get an edge that has nothing to do with merit. In The Price of Admission, journalist Daniel Golden documents the ways in which elite schools manage to find room for the children of alums, big donors, celebrities, athletes, the elite college’s own faculty, and wealthy parents whose estates might eventually make their heirs into big donors.20 The question is this: What would the freshman class look like if all of these considerations were eliminated and the decisions were made purely on the basis of test scores, extracurricular achievements, teacher recommendations, and high school transcripts? Answer: Socioeconomically, the change in the class profile would range from minuscule to zero. The elite schools are turning away at least two-thirds and often 80 or 90 percent of their highly qualified applicant pool. The applicants on the cusp who would be admitted if all the preferential treatment were eliminated would still be dominated by children from the upper-middle class, because those are the young people who dominate the applicant pool.

When Soares turns his multivariate analyses to the admissions decisions, his results are notable for what they did not reveal. The most basic indicators of the socioeconomic status of the applicant—parental income and occupation—did not produce significant relationships with probability of admission after controlling for measures of the student’s real abilities.21

The reason that upper-middle-class children dominate the population of elite schools is that the parents of the upper-middle class now produce a disproportionate number of the smartest children. For example, one of the basics for having a decent chance of getting into an elite school is a high SAT score, with “high” defined as at least 700 on the SAT verbal and SAT math. Among college-bound seniors who took the SAT in 2010, 87 percent of the students with 700-plus scores in the math and verbal tests had at least one parent with a college degree. Fifty-six percent of them had a parent with a graduate degree.22 This is not a function of coaching—the dispassionate studies of coaching show average gains of only a few dozen points—but of ability to do well in a challenging academic setting.23 That ability is reflected in the other measures—grades, teacher evaluations, and many types of extracurricular accomplishments—that admissions committees use.

In that glaring relationship of high test scores to advanced parental education, which in turn means high parental IQ, lies the reason that the tests aren’t the problem and bias in the admissions process isn’t the problem. The children of the well educated and affluent get most of the top scores because they constitute most of the smartest kids. They are smart in large part because their parents are smart.24 That brings us to the role of homogamy.

The Perpetuator: Homogamy

Homogamy refers to the interbreeding of individuals with like characteristics. Educational homogamy occurs when individuals with similar educations have children. Cognitive homogamy occurs when individuals with similar cognitive ability have children.

The Increase in Cognitive Homogamy

Before the age of mobility, people commonly married someone from the same town or from the same neighborhood of an urban area. The events that threw people together seldom had anything to do specifically with cognitive ability. Similar cognitive ability was a source of compatibility between a young man and young woman, and some degree of cognitive homogamy existed, but it was a haphazard process.25 Meanwhile, educational homogamy was high, because hardly anyone went to college. In large proportions of married couples, both had less than a high school education or both had a high school diploma.

As the proportion of college graduates increased, so did the possibilities for greater educational homogamy at the top, as college graduates found they had more potential marriage partners who were also college graduates. Drawing on the extensive technical literature and the CPS, sociologists Christine Schwartz and Robert Mare examined trends in “assortative marriage,” as it is known in the jargon, from 1940 to 2003.26 They found that homogamy has increased at both ends of the educational scale—college graduates grew more likely to marry college graduates and high school dropouts grew more likely to marry other high school dropouts.

For our purposes—trying to understand how the new upper class came to be—the effects of increased educational attainment may be seen in a simple measure. In 1960, just 3 percent of American couples both had a college degree. By 2010, that proportion stood at 25 percent. The change was so large that it was a major contributor to the creation of a new class all by itself.

But increased educational homogamy had another consequence that the academic literature on homogamy avoids mentioning. Increased educational homogamy inevitably means increased cognitive homogamy.

A college education, starting with admission and continuing through to graduation, is a series of cognitive tests.27 To be able even to begin a major in engineering or the hard sciences, students have to be able to do advanced calculus, and that in turn requires logical-mathematical ability in roughly the top decile of the population. To be able to cope with genuine college-level material in the social sciences and humanities requires good linguistic ability—in the top quartile of the distribution if you’re content with scraping by, closer to the top decile if you want to get good grades in a moderately demanding college.28 To graduate means passing all these tests plus a general test for perseverance.

Which Comes First, Education or IQ?

Educational attainment is correlated with IQ, but education does not have much effect on IQ after the child enters elementary school. By that, I do not mean that the absence of any education after age 6 wouldn’t make a difference, nor that exceptions do not exist. Rather, I mean that if a thousand children are administered a good IQ test at age 6, and those children then attend a wide variety of elementary and secondary schools, their IQs at age 18 will be very similar to what they were at age 6, and statistical analysis will not show that the children who went to the expensive private schools got an IQ boost as a result. This finding goes back to the famous Coleman Report in the 1960s.29 Scholars still debate whether additional years of education are associated with increments in general mental ability or just increments in test scores, but no one contends that education routinely transforms average children into intellectually gifted adults.

The result is that each level of educational attainment—high school diploma, AA, BA, MA, and professional degree or PhD—implies a mean IQ for people attaining that level that has been remarkably stable among whites at least since the beginning of the 1980s. I must limit the numbers to whites as I present these data, because aggressive affirmative action has produced means for African Americans and Latinos at each level of educational attainment that are substantially lower and more variable than the white means.30 But since we are talking about the new upper class, there are good reasons to think in terms of the white means—partly because African Americans and Latinos who enter the new upper class have passed a number of career tests signifying that they approximate the white means on cognitive ability for each level of educational attainment, and partly because the new upper class is still overwhelmingly white.31

Table 2.1 shows the evidence for these stable means. The data for persons reaching adulthood in the 1980s and 2000s come from the 1979 and 1997 cohorts of the National Longitudinal Survey of Youth (NLSY), which are used to establish national norms for the Armed Forces Qualification Test (AFQT), which measures the same cognitive abilities that IQ tests measure.32

TABLE 2.1. MEAN WHITE IQ FOR LEVELS OF DEGREE ATTAINMENT IN THE NLSY-79 AND NLSY-97

  Years when subjects reached age 25
  1982-89 2005-9
Mean IQ for persons completing no more than …
No degree 88 87
High school diploma/GED 99 99
Associate’s degree 105 104
Bachelor’s degree 113 113
Master’s degree 117 117
PhD, LLD, MD, DDS 126 124

Source: NLSY-79, NLSY-97. Sample limited to whites.

The stability of the scores over the three decades from the 1980s through the 2000s is remarkable, considering that the number of bachelor’s degrees, expressed as a percentage of twenty-two-year-olds, increased from 22 percent in 1981 to 37 percent in 2008. But the country was also becoming steadily more efficient at getting the best students into college over that period, so that the greater size of the college population didn’t mean a markedly less able population.

If the mean IQs at the higher levels of educational attainment have been stable, then the growth of two-degree couples has meant, inevitably, greater cognitive homogamy at the top. But that’s just the beginning. The college sorting machine has also been at work.

College brings people together at the time of life when young adults are beginning to look around for marriage partners, and the college sorting machine brings the highest-IQ young women and young men together in the most prestigious schools. As if that weren’t enough, graduate school adds another layer of sorting, so that the brilliant young woman who went to a state university goes to Harvard Law School, where she is brought into the elite pool. For the prospective members of the new upper class who don’t find a marriage partner as an undergraduate or at grad school, the names of the schools they attended give them badges that signal their status to prospective mates. The substance of their education also sorts them into occupations that increase the likelihood that they will eventually marry people with similar characteristics.33

So it’s not just that college graduates are likely to marry college graduates, but that graduates from elite colleges are likely to marry other graduates from elite colleges.34 Back in the days when Harvard men and Wellesley women were more likely to be rich than to be especially smart, this meant that money was more likely to marry money. In an era when they are both almost certainly in the top centiles of the IQ distribution, it means that very smart is more likely to marry very smart.

Shared Culture

Put people with greater educational and cognitive similarity together, and you have the makings of greater cultural similarity as well. When one spouse is a college graduate in the top centiles of cognitive ability and the other is a high school graduate with modestly above-average cognitive ability, they are likely to have different preferences in books and movies, different ways of spending their free time, different friends, and differences in a dozen other aspects of life. Those differences carry with them a built-in measure of cultural dispersion within marriages. In 1960, two-thirds of families in managerial and professional occupations had that built-in educational heterogeneity. In 2010, when three-quarters of the most financially successful couples both had college degrees, the demand for the goods and services to supply the distinctive tastes and preferences of very bright and well-educated people had been concentrated.

Transmission of Cognitive Ability to the Next Generation

Another consequence of increased educational and cognitive homogamy is the increased tenacity of the elite in maintaining its status across generations. The adage “shirtsleeves to shirtsleeves in three generations” grew out of an observed reality: If the children and grandchildren are only average in their own abilities, money from a fortune won in the first generation won’t keep them at the top of the heap. When the parents are passing cognitive ability along with the money, the staying power of the elite across generations increases.

Specific numbers can be attached to such statements. The stability of the average IQs for different levels of educational attainment over time means that we can predict the average IQs of children of parents with different combinations of education, and we can also predict where the next generation of the smartest children is going to come from.

On average, children are neither as smart nor as dumb as their parents. They are closer to the middle. This tendency is called regression to the mean. It exists independently of genes. Regression to the mean is a function of the empirically observed statistical relationships between the tested IQs of parents and children. Given the parameters in the note on this page, the expected value of the IQ of a grown-up offspring is 40 percent toward the population mean from the parents’ midpoint IQ.35

Suppose we have four white couples with the same level of education, plugging in the average IQs for those levels of education as given in table 2.1 (splitting the difference between the NLSY-79 and NLSY-97 figures when necessary). I add a fifth couple who both have degrees from elite colleges, with a midpoint IQ of 135.36 Here is what we can expect as mean IQs of the children of these couples:

Parents’ Educations Expected IQ of the Child
Two high school dropouts 94
Two high school diplomas 101
Two college degrees (and no more) 109
Two graduate degrees 116
Two degrees from an elite college 121

These represent important differences in the resources that members of the next generation take to the preservation of their legacy. Consider first a college graduate who marries a high school graduate, each with the average cognitive ability for their educational level (113 and 99, respectively). Their expected midpoint IQ is 106. Suppose they have built a small business, been highly successful, and leave $5 million to their son. If their son has the expected IQ of a little less than 105, he will have only about a 50 percent chance of completing college even assuming that he tries to go to college. Maybe he inherited extraordinary energy and determination from his parents, which would help, but those qualities regress to the mean as well. Shirtsleeves to shirtsleeves in three generations is a likely scenario for the progeny of that successful example. Compare that situation with the one facing the son of two parents who both graduated from elite schools. If he has exactly the expected IQ of 121, he has more than an 80 percent chance of getting a degree if he goes to college. These percentages are not a matter of statistical theory. They are based on the empirical experience of both the 1979 and 1997 cohorts of the National Longitudinal Survey of Youth—if you had an IQ of 105 or one of 121 and entered college, those are the probabilities that you ever got a degree.37

In addition to those differing chances of graduation are qualitative differences between young people with IQs of 105 and 121. First, the reasons that someone with an IQ of 105 doesn’t finish college probably include serious academic difficulties with the work, whereas the reasons a person with an IQ of 121 doesn’t finish college almost certainly involve motivation or self-discipline—no one with an IQ of 121 has to drop out of college because he can’t pass the courses. Second, there is a qualitative difference in the range of occupations open to those two young persons. The one with an accurately measured IQ of 105 cannot expect to be successful in any of the prestigious professions that are screened for IQ by their educational requirements (e.g., medicine, law, engineering, academia). It is unlikely that he can even complete those educational requirements. Someone with an accurately measured IQ of 121 can succeed in any of them if his mathematical and verbal talents are both strong, or succeed in the ones geared to his talents if there is an imbalance between mathematical and verbal ability.38

Now think in terms of an entire cohort of children. Where will the next generation of children with exceptional cognitive ability come from? For purposes of illustration, let’s say that “exceptionally high cognitive ability” means the top five centiles of the next generation of white children. More than a quarter of their parents may be expected to have a midpoint IQ of more than 125.39 Another quarter may be expected to have midpoint parental IQ of 117–125. The third quarter may be expected to have midpoint parental IQ of 108–117. That leaves one quarter who will be the children of parents with midpoint parental IQ of less than 108. Only about 14 percent of that top five centiles of children are expected to come from the entire bottom half of the distribution of white parents.

Therein lies the explanation for that startling statistic I reported earlier about SAT scores: In 2010, 87 percent of the students with 700-plus scores in Critical Reading or Mathematics had a parent with a college degree, and 57 percent had a parent with a graduate degree. Those percentages could have been predicted pretty closely just by knowing the facts about the IQs associated with different educational levels and the correlation between parental and child IQ. They could have been predicted without making any theoretical assumptions about the roles of nature and nurture in transmitting cognitive ability and without knowing anything about the family incomes of those SAT test-takers, how many test preparation courses their children took, whether they went to private schools, or how ingenious the educational toys in the household were when they were toddlers.

In an age when the majority of parents in the top five centiles of cognitive ability worked as farmers, shopkeepers, blue-collar workers, and housewives—a situation that necessarily prevailed a century ago, given the occupational and educational distributions during the early 1900s—these relationships between the cognitive ability of parents and children had no ominous implications. Today, when the exceptionally qualified have been so efficiently drawn into the ranks of the upper-middle class, and when they are so often married to people with the same ability and background, they do. In fact, the implications are even more ominous than I just described because none of the numbers I used to illustrate the transmission of cognitive ability to the next generation incorporated the effects of the increased educational homogamy of recent decades. In any case, the bottom line is not subject to refutation: Highly disproportionate numbers of exceptionally able children in the next generation will come from parents in the upper-middle class, and more specifically from parents who are already part of the broad elite.