The Son Also Rises

by Gregory Clark

  Inequality Given Slow Mobility

  This book shows time and again that social mobility is slow, is strongly inherited within families, and that there is little evidence of our ability, using feasible social programs, to increase it. Facing such a reality, the emphasis for societies should be on reducing the effects of inherited abilities and family ethos and aspirations on the rewards society generates for those of different abilities. If so much of social outcomes are determined at birth, then we can appeal to people’s sense of justice in other circumstances to argue for more redistribution. For example, there is widespread support for the resource transfers necessary to ensure that people born with physical limitations are not thereby impoverished. If social success and failure are strongly ordained at birth, then, by analogy, why not provide more aid for those who are unlucky in the familial random draw?

  If we cannot change the heritable advantages and disadvantages of families in the economic and social world, we should at least mitigate the consequences of these differences. For although there is no evidence that we can change social mobility rates, there is plenty of evidence that societies can reduce inequality in earnings, wealth, health, and relative social status. If low social mobility rates really are a law of nature, as incontrovertible as the gravitational constant, then we should spend less time worrying about them and instead worry about the institutions that determine the degree of inequality in social and economic outcomes.

  Some of the inequalities of income and wealth are, of course, the product of economic forces beyond the control of social institutions. But the tax system can mitigate the effect of these market forces on the distribution of rewards, and societies can control the degree of inequality they create in various social institutions.

  Some societies have used public interventions to compensate, to some degree, for the inherited disadvantages of poorer families. Sweden, for example, has much more extensive and effective educational and health interventions for poorer families than the United States does. Years of education are correlated with life expectancy in both societies, but to different degrees. In Sweden, the difference in life expectancy between high-school graduates and those with some postsecondary education was less than three years at age 30 in 2010.22 In the United States this gap, just looking at the white population, was seven years at age 25 in 2008.23 This difference is consistent with the idea that Sweden has narrowed the disparity in living conditions between rich and poor through universal access to health care and other social benefits.

  Such interventions to equalize life chances, of course, demand resources raised through the tax system. The average tax burden on wages, for example, is much higher in Nordic countries than in the more laissez-faire states of the Anglo-Saxon economic model. The OECD reports the average tax burden, counting all taxes and assessments on wages, to be 39 percent in Denmark and 43 percent in Sweden in 2012, compared to 30 percent in the United States and 32 percent in the United Kingdom.24

  We also see above that earnings are more equal in Nordic countries, though how much that is due to differences in labor supply as opposed to institutional choices is unclear. Unionization, however, is much more extensive in Nordic countries. The OECD reports a unionization rate of 68 percent of employees in Sweden in 2010 and 69 percent in Denmark, compared to 26 percent in the United Kingdom and 11 percent in the United States.25 While there is no official minimum wage in either Denmark or Sweden, union contracts typically set relatively high minimum wages for their respective economic sectors.

  Most economists, who value the free market as an economic regulator, would fear that such union and tax interventions would impose significant losses of output by creating disincentives to work and constraints on productive economic arrangements. It is indeed the case that output per person in the United States in 2010, at $42,000 (in 2005 dollars), was greater than in Sweden ($35,000) or Denmark ($36,000). However, the more free-market United Kingdom had an output of only $32,000, so there is no evidence that output is substantially affected by the much more equalizing Nordic social institutions.26 And work hours per adult are higher in the United States, so if output were measured per worker hour, the differences between the United States and Nordic countries would be even lower.

  Societies make other institutional choices that magnify or diminish inequalities in status. Consider university education, the path to careers and status for many of the next generation. Some societies—including the United States, the United Kingdom, China, and Japan—engage in extreme sorting of undergraduates into educational institutions. The most prestigious universities—Harvard, Princeton, Oxford, Cambridge, Peking, Tsinghua, Tokyo—select their students largely on merit, as defined within each society. They recruit a group of students from the apex of the ability distribution—a group that, as we saw, is concentrated in persistently elite families. In the United States, such universities may also have “legacy” admissions policies that favor students whose parents attended the same university.

  In the United States, the advantages of these elite universities are reinforced by a flow of private donations from alumni and others. In a perverse form of philanthropy, those who have get more, and those who have not get nothing.27Oxford and Cambridge have been working hard in recent years to emulate this successful U.S. funding strategy of obtaining alms for the rich.

  There is no evidence that this extreme sorting by ability is necessary to the operation of a productive university system. In other very successful societies, such as the Netherlands or Germany, universities are much less differentiated in their undergraduate composition. Thus in Germany, there is little distinction in status or student quality across the top ten or twenty German universities. Famous institutions, such as Heidelberg, do not have particularly selective undergraduate admissions. Less-popular undergraduate majors at Heidelberg, such as classics or ancient history, are still open to anyone with the university qualification, the Abitur. Other courses of study, such as medicine or law, are highly selective in their admissions, but this is equally true for many other German universities. In the Netherlands, under even greater egalitarian impulses, medical-school places since 1972 have been awarded not purely on merit criteria, such as high-school grades, but through a weighted lottery that was open to all students meeting minimum eligibility criteria.

  In the U.S. social model, extreme differentiation of status and outcomes has been permitted, and even encouraged, on the basis that it fosters and increases social mobility. We have seen that in all societies with intermarriage across social groups, social mobility will eventually equalize all social groups and expected outcomes for all families. But the pace of this mobility is very slow. Thus the Nordic model of using social institutions to reduce status differences and outcomes between elites and underclasses may look more attractive.

  Inequality across Countries

  We have seen in case after case that intergenerational mobility is slow. In particular, social groups converge on similar levels of social and economic success only after many generations, even in the most open societies. The case of the United States shows that strong forces of both positive and negative selection have operated on some of the groups arriving on its shores. Its elites now include a host of subpopulations from societies around the world: Coptic Egyptians, Indian Hindus, Iranians, Maronites, black Africans, and so on. These groups may represent as much as 5 percent of the population. Two percent of the population is Jewish, which, as we argue above, is an elite stemming from earlier selection processes. The underclasses of the United States include groups in which negative selection from the mother population is likely: New France descendants, Mexican Americans, and likely also the Hmong. Such groups may account for up to 18 percent of the U.S. population.

  This situation implies that the United States has, and will have for generations into the future, a much higher intrinsic level of social inequality than more homogeneous societies such as Germany, Poland, or Italy. Thus it will experience a greater variance of s
ocial outcomes for generations to come. Given the historic disparities in its constituent populations and the likelihood that immigration policies will sustain them, the United States needs to consider whether its commitment to social institutions that tolerate and even foster huge social inequalities is appropriate.

  1 See, for example, “Nomencracy” 2013.

  2 Heckman 2012.

  3 Murray 2012.

  4 Plomin et al. 1997, table 1. The correlation of intelligence with fathers is as high as with mothers, even though for the children studied, born in the years 1975–82, mothers would have interacted with children more than fathers.

  5 Plomin et al. 1997, 446.

  6 Plomin et al. 1997, 446. Two other adoption studies, in Minnesota and Texas, found more correlation between adoptive parents and children at age 18 than did the Colorado study. But these correlations averaged just 0.12 and 0.06, still very low. The overall average across the three studies is thus 0.07 (Richardson and Norgate 2006, 320).

  7 Sacerdote 2007.

  8 It is two times the correlation between genetically related siblings minus the correlation between genetically unrelated siblings.

  9 Black, Devereux, and Salvanes 2005.

  10 Björklund, Lindahl, and Plug 2006.

  11 The winners, however, did not obtain possession of the land until the removal of the Cherokee in 1838.

  12 Bleakley and Ferrie 2013a.

  13 Bleakley and Ferrie 2013b, table 6.

  14 Akee et al. 2010. All Cherokee children became eligible for the annual payment at age 21, regardless of educational status.

  15 Akee et al. 2010, tables 5 and 9.

  16 Løken 2010.

  17 Løken 2010, 128.

  18 Oreopoulos, Page, and Stevens 2008.

  19 Heckman 2012.

  20 Heckman et al. 2010a,b. However, the two programs enrolled, respectively, 58 and 57 treated children, and 65 and 54 controls. This constitutes only a modest evidential basis for the effects of early interventions (Campbell et al. 2012).

  21 Puma et al. 2012.

  22 Statistics Sweden 2011a.

  23 Olchansky et al. 2012.

  24 OECD 2013a, 15, table 0.1.

  25 OECD 2013b.

  26 Feenstra, Inklaar, and Timmer 2013. Expenditure-side real GDP at chained purchasing-power parities.

  27 In 2012 Stanford University received gifts of $1.04 billion, $157,000 per undergraduate (Stanford University 2012). Only a modest share of this money was earmarked for undergraduate education, but undergraduates also benefit from the prestige conferred by the faculty hired and the research accomplished with this money.

  SIXTEEN

  Escaping Downward Social Mobility

  MOST PARENTS, particularly upper-class parents, attach enormous importance to the social and economic success of their children. They spare no expenditure of time or money in the pursuit of these goals. In these efforts, they seek only to secure the best for their children, not to harm the chances of others. But the social world only has so many positions of status, influence, and wealth. Inevitably it seems that in pushing their own children up the social ladder, parents are stamping on the fingers of those climbing up from below. As a character in an Iris Murdoch novel says, “It is not enough to succeed; others must fail.”1

  Competition to enter the best private schools in Manhattan, for example, is so intense that it begins with kindergarten. The Dalton School on the wealthy Upper East Side, one of Manhattan’s Ivy League feeder schools, has such fierce demand for places in its kindergarten that four-year-olds undergo IQ tests and admission interviews. The selection process is so onerous that the deadline for applications for admission in all of 2013 was November 9, 2012. Thus recent moves to diversify the school and admit more students “of color” were greeted with a marked lack of enthusiasm by the parents of nonminority applicants who would thus face even more competition.2 Admission gained parents the privilege of paying annual fees of $38,710 for students in grades K–12 (though that does include school lunch).

  The common entrance exam for Manhattan private elementary schools, known as the ERB, costs $500 just to attempt. But an ancillary army of advisors and tutors is available for hire to ensure that your child has every advantage in getting into the right school and onto the right path in life. This preparation industry has become so expert that most of the private schools in Manhattan are expected soon to end their reliance on the ERB “because of concerns that the popularity of test-preparation programs and coaching had rendered its results meaningless.”3

  When it comes time for college admissions, another army of advisors awaits the call to arms to boost SAT scores, shape college admissions essays, and guide students in selecting from the appropriate armamentarium of extracurricular activities. Since sports provides an avenue of entry to elite universities for those with less than compelling SAT scores and GPAs, battalions of high schoolers are drilled in sports such as field hockey and lacrosse, which exist mainly as an adjunct to the college admission process.4

  For a long period, from at least 1880 to 1980, the rich and socially successful sharply limited their fertility. Their fewer children would thus each inherit more parental assets and gain a larger share of parental time and resources, than the abundant children of the poor. Yet despite a willingness to spend big in terms of time and treasure, we know that the law of social mobility exercised an inexorable pull, drawing families toward the mean. There is strong persistence of status, but those at the top of the social hierarchy in societies such as the United Kingdom, the United States, and Sweden will inevitably see their children, on average, move down.

  Further, the rate of regression downward to the mean is the same for the upper echelons of society, despite their considerable investments in their children, as is the rate of upward mobility for the lower echelons, even the ones who don’t bother to turn up for the PTA meetings.

  The forces of regression to the mean may seem glacially slow from the point of view of those at the bottom of the social ladder. But for the elites of Manhattan, Greenwich, or Silicon Valley, these forces exercise a death grip on dynastic ambitions. These are people used to getting what they want. Why should they be frustrated in this one primal ambition, for their children to enjoy the same rewards in life as their parents?

  The empirical evidence that middle- and upper-class parents can significantly boost their children’s human capital and economic outcomes through expenditure on children is weak, as Bryan Caplan recently emphasized in his book Selfish Reasons to Have More Kids.5 Even the pampered progeny of the lords of finance in Manhattan remain subject to the law of social mobility.

  This is all consistent with the idea that once parental inputs to children reach a certain basic level, which does not include Baby Einstein toys, playing Mozart to babies in the womb, or sending them to the Dalton School, parents can do nothing to improve outcomes for children. Beyond this point, social outcomes are potentially all in the genes, determined at the point of conception—or driven by a set of aspirations and values already embedded in parents and transmitted automatically to the children, who drink in their social class with their mother’s milk. Most likely, given the evidence above, the majority of status is actually genetically determined. You can hit the jackpot in the great genetic casino or go bust.

  Is there anything that this book can say to people who want the best possible income, wealth, education, and health outcomes for their children? The one scientific contribution we can make is to point out that with the appropriate choice of mates, a family can avoid downward mobility forever.

  The chapters above emphasize that one of the things that slows social mobility is the assortative nature of marriage. People in all societies tend to marry others of similar social status. Recently Charles Murray has argued that marriage has become even more assortative and that this trend will slow social mobility further.6 The reasoning behind Murray’s claim is that in earlier generations, women did not get much education, and thus potential ma
le partners had less information about their abilities, energy, and drive. But the rise of women’s education is permitting a much better matching of marital partners on these observed characteristics, thus slowing rates of social mobility. Even white society in America, on this view, is increasingly being segregated into lineages of prosperity and deprivation.7

  But no matter how assortative mating may become, downward mobility will continue. For downward mobility is driven by the fact that people typically select mates who resemble them on the basis of observed social characteristics—their achieved education, income, occupational status, wealth, height, weight, and health.8 This is their social phenotype, the sum of their observed characteristics. However, as we have seen above, we can usefully think of individuals as also having a social genotype, or underlying social status.9 Their social genotype produces the observed phenotype, but with random components in each dimension.

  This means that the people currently occupying the upper tails of the distribution of education, wealth, and occupational prestige tend to include disproportionately the lucky, the ones who benefited from happy accidents. Systematically, at the top, the phenotype is better than the genotype. Symmetrically, concentrated at the bottom are people who have experienced bad luck and unhappy accidents. There, the social genotype is much better than the observed phenotype. The curse of the elite is that they are surrounded by imposters, possibly including themselves, and thus the marriage market for the upper classes is full of prospects likely to underperform as carriers of a lineage. In contrast, the bottom of the marriage market is full of potential overperformers. Bad luck dominates, rather than bad social genotypes. So outcomes for the next generation tend to be better.