So while race-based segregation has been slowly declining, class-based segregation has been increasing. In fact, the trend toward class segregation has been true within each major racial group, so affluent and impoverished black (or Latino) families are less likely to be neighbors now than they were 40 years ago.
Earlier in this chapter we mapped the increase of class-based segregation on either side of East Harbor Road in Port Clinton, and in subsequent chapters we shall see how exactly this same process—a kind of incipient class apartheid—has worked itself out in towns and cities across the country. It is a process that has powerful consequences for whom our children encounter in their daily lives—both in school and out, in terms of both peers and potential mentors, as we shall see in Chapter 5. Whether we are rich or poor, our kids are increasingly growing up with kids like them who have parents like us.
Since the 1970s, increasing class-based residential segregation has been translated into de facto class-based school segregation. Schoolchildren from the top half of the income distribution increasingly attend private schools or live in better school districts. Even when poor and wealthier schoolchildren live in the same school district, they are increasingly likely to attend separate and unequal schools. And often within a single school, AP and other advanced courses tend to separate privileged from less privileged kids. Later on, kids from different class backgrounds are increasingly sorted into different colleges: for example, by 2004, kids from the top quarter of families in education and income were 17 times more likely to attend a highly selective college than kids in the bottom quarter.40
Once again, this educational segregation has consequences far beyond the classroom, in terms of friendship networks and other social resources. As we have already seen, in Port Clinton in the 1950s kids from all sorts of backgrounds attended the same classes, played on the same teams, and went to the same parties. Today, however, even though Chelsea and David live only a few miles apart, they are unlikely even to encounter each other. Educational segregation is so important that we will devote an entire chapter to it (Chapter 4).
People mostly tend to marry others like themselves, but the degree of intermarriage across various social boundaries changes over time. For two reasons, intermarriage rates are useful indicators of how strict boundaries are in social life. First, we mostly marry people that we’ve met. Therefore, the more permeable a boundary (for example, in neighborhoods or schools), the more likely that young people will meet mates on the other side. Second, a high intermarriage rate implies that interaction across that boundary will be more frequent in the future, at least within extended families. In short, high intermarriage rates today imply low social segregation yesterday, and still lower social segregation tomorrow. In recent decades, for example, increasing religious and racial intermarriage has reflected and reinforced the gradual lowering of religious and racial barriers throughout American society. What about intermarriage across class boundaries?
Trends in marriage across class lines throughout the twentieth century turn out to mirror almost precisely the Great U in income inequality.41 During the first half of the century, marrying outside one’s own social class became steadily more common. After mid-century, however, that trend reversed itself. In the second half of the century Americans increasingly married people with educational backgrounds similar to their own, with the most educated especially likely to marry one another.42 In other words, as the gap between rich and poor narrowed in the first half of the twentieth century, more and more Romeos and Juliets jumped across, but as the economic and educational gulf has widened in more recent decades, fewer and fewer people are finding partners on the other side.
The decline in cross-class marriages has implications for the composition of extended families. Two generations ago, extended family gatherings might bring together small businessmen and manual workers, professors and construction workers, but the ripple effects of increasing endogamy (marrying within your own social class) ensure that one’s kin networks today—and even more, tomorrow—are likely to be from the same class background as oneself, further reducing cross-class bridging. Fewer and fewer working-class kids will have rich uncles or well-educated aunts to help them ascend the ladder.
Ultimately, growing class segregation across neighborhoods, schools, marriages (and probably also civic associations, workplaces, and friendship circles43) means that rich Americans and poor Americans are living, learning, and raising children in increasingly separate and unequal worlds, removing the stepping-stones to upward mobility—college-going classmates or cousins or middle-class neighbors, who might take a working-class kid from the neighborhood under their wing. Moreover, class segregation means that members of the upper middle class are less likely to have firsthand knowledge of the lives of poor kids and thus are unable even to recognize the growing opportunity gap. One reason, in fact, for including life stories of young people in this book is to help reduce that perception gap—to help us all to see, in the words of Jacob Riis, a social reformer during the previous Gilded Age, “how the other half lives.”44
To what extent has the American Dream of a fair start for everyone in fact been characteristic of American history? The answer to that question turns out to depend in part on the standard of comparison—the myth of perfectly open upward mobility, the reality of our own past, or the reality of similar countries. It also depends on the important distinction between absolute and relative upward mobility.
In a growing economy with rising educational levels, everyone could in principle do better than his or her parents in absolute terms, even if the relative standing of every family remained unchanged—if the children of college graduates got postgraduate degrees, for example, while the children of the illiterate graduated from elementary school. In such a world, a rising tide would lift all boats, even if no one ever moved from a dinghy to a yacht, so that relative mobility was zero.
Conversely, even if the economy as a whole were stagnant, in a system of perfect social mobility the more capable and ambitious children of lower-class parents on their way up would pass the ne’er-do-well scions of upper-class families on their way down. Such a world would have equality of opportunity, because where people ended up would not depend on where they started in life. So a society might have low absolute mobility but high relative mobility, or the reverse.
Over the grand sweep of history, relative mobility accounts for only a small portion of total mobility experienced by individuals across generations, whereas absolute (or structural) mobility accounts for most of it. During periods of high growth in income and education, or thinning out of manual occupations, many people from lower-class backgrounds will experience absolute upward mobility, regardless of changes in relative mobility.
In principle, of course, a society might have both high absolute mobility (a rising tide lifting all boats) and high relative mobility (dinghies doing even better than yachts). My classmates in Port Clinton in the fortunate 1950s and 1960s benefited from exactly that happy state of affairs, and scholars have identified the same pattern nationwide.45 The underlying issue raised in this book is whether, by contrast, American youth now have the worst of both worlds—low absolute mobility and low relative mobility.
Most empirical studies of social mobility in America prior to the twentieth century focused on absolute upward mobility among white men, and used the national myth of perfect mobility as the standard of comparison. In other words, they asked how many upper-class men were strictly self-made—and the answer is “relatively few.” In that sense, these earlier studies ended up debunking the national myth, because mobility seems never to have been as great as the rags-to-riches narrative implies.
On the other hand, careful statistical comparisons by historians suggest that economic growth and successive expansions of the educational system did allow significant absolute mobility, perhaps especially in the first half of the twentieth century.46 In the decades after World War II, as I have said, absolute mobility (and to some extent even relative mobility) seems to have been unusually high, because economic growth and educational expansion allowed exceptional upward mobility.
The evidence now suggests, however, that absolute mobility has stalled since the 1970s, because both economic and educational advances have stalled.47 Much recent public commentary asserts that relative social mobility in America has fallen in the past quarter century as well, though hard evidence for those claims is weaker.48 In other words, Americans believe that income inequality has increased in recent years, and they are right about that. They aren’t so sure that equality of opportunity (or upward mobility) has changed much, and so far they seem to be right about that, too, even if they overestimate the odds of moving from the bottom to the top. But—and this “but” is crucial for this book—conventional indicators of social mobility are invariably three or four decades out of date.
The conventional method of assessing mobility compares a son’s (or daughter’s) income or education when they are in their 30s and 40s with their parents’ income or education when their parents were in their 30s and 40s. The rationale is that not until a given generation reaches early middle age do we know with confidence where they will end up on the socioeconomic ladder. However, that approach necessarily means that conventional measures of mobility are a “lagging indicator” of social change, since even the most recent conventional measures refer to a generation that was born 30 to 40 years ago. In assessing social mobility, therefore, policymakers and citizens who rely on the conventional method are like astronomers studying the stars: they have to contend with an information time lag, and can only see what has happened years or eons ago, not what is happening right now. David and Chelsea will not show up in statistical assessments of social mobility until the 2020s. Their childhood and adolescent experiences, compared with the experiences of my 1959 classmates, suggest that we’ve been veering further away from equality of opportunity for several decades—but if so, we won’t detect that slowing of upward mobility in our conventional measurements for another decade or so. Similarly, if Alpha Centauri, our nearest stellar neighbor, exploded last night, we won’t know about it for more than four years.
This book adopts a different approach, eschewing the conventional “rearview mirror” method and examining directly what has been happening to kids in the past three decades—the families into which they’ve been born, the parenting and schooling they’ve received, the communities within which they’ve been raised.49 We know that those experiences will inevitably have a powerful effect on how well they do in life. Whatever changes we can detect in these areas will foreshadow changes in social mobility—which, distressingly, according to the evidence I describe in this book, seems poised to plunge in the years ahead, shattering the American Dream.
If this book were a sociological text, we would need to distinguish among different conceptions and indicators of social class, such as occupation, wealth, income, education, culture, social status, and self-identity, and we would have to worry about inconsistencies among these measures—a well-educated but poorly paid librarian, for example, or a barely literate billionaire.50 For our purposes, however, and for the population as a whole, these various indicators are closely inter-correlated, and I know of no instances in which the core generalizations in this book depend entirely on the specific choice of indicator.
Education, and especially higher education, has become increasingly important for good jobs and higher incomes; in the language of economics, the “returns to education” have increased. While education and income are thus becoming more closely correlated, I generally prefer education as our indicator of social class, partly because income measures in most surveys are much “noisier” (error-prone or entirely missing) and because when both are available, education is typically the more powerful predictor of child-related outcomes. Here I follow the example of sociologist Douglas Massey, who operationalizes social class by “education, the most important resource in today’s knowledge-based economy.”51 Another practical reason is that few of the long-term studies on which we are forced to rely have good measures of family income.
For consistency and simplicity, in this book I typically report class breakdowns either by education alone (college degree or more versus high school or less) or by a composite measure of socioeconomic status (based on income, education, and occupational status), depending on what indicators are available for a particular topic or survey. Roughly speaking, the educational attainment of Americans can be divided in thirds, with the top third college graduates, the bottom third no more than high-school-educated, and the middle third with some postsecondary education. So when I speak of kids from “upper-class” homes, I simply mean that at least one of their parents (usually both) graduated from college, and when I speak of kids from “lower-class” homes, I simply mean that neither of their parents went beyond high school. Other breakdowns would produce essentially the same patterns. For the sake of variety, in the text I often use the shorthand of “high-school-educated” or simply “poor” to refer to anyone who has no more than a high school education, and “college-educated” or simply “rich” for anyone with a college degree.