March 11, 2012
The Growing Class Divide
The Growing Class Divide—Posner
Charles Murray's recent book Coming Apart has gotten a lot of attention because of the data it presents concerning the growing social and economic gap between working-class and middle-class people (politicians call all nonwealthy Americans "middle class," but that is because politicians in a democratic society need to flatter the masses in order to get elected). The incomes of working-class people have fallen, family stability and employment have declined, and the crime rate has risen, creating a very large gap in these respects between working class and middle class, especially upper middle class. As summarized by the* Washington Post*'s economic analyst Robert Samuelson, "Among those 30 to 49 in the blue-collar community, 84 percent were married in 1960 and only 48 percent in 2010. In 1962, 96 percent of children were living with both biological parents; by 2004, the proportion was 37 percent. Meanwhile, the share of households with someone working at least 40 hours a week dropped from 81 percent in 1960 to 60 percent in 2008." The middle class and especially the upper middle class have escaped these trends. Murray's book is not very analytical, however, and also (though this is a limitation more than a defect) contains no constructive proposals for narrowing the class divide—both points argued in a recent online review of the book by David Frum. As Becker points out, the proposals that Murray made in a March 7* New York Times* piece are unhelpful.
A significant causal factor in the growing class divide reported in Murray's book may be the increasing return to IQ. In a society in which most work is manual (for example, agricultural labor, construction of buildings and roads, mining, assembly-line manufacturing, old-fashioned soldiering, and craft work such as plumbing and carpentry), the demand for manual labor is great, and the return to personal qualities that enhance the productivity of a manual laborer, such as physical strength and tolerance for dirty or dangerous working conditions, is substantial and favors men, because they have superior aptitude for most manual labor. (Male upper-body strength in particular is substantially greater than female.) Women have the superior aptitude for household labor, including child care, and so do not compete with men in the labor market in an economy dominated by manual labor.
With technological advance, the market value of the traditional male labor assets—assets that are not correlated with IQ—declines, and men find themselves competing with women for service jobs, which grow relative to mining, manufacturing, and other heavy-duty traditionally male jobs. Hence working-class male incomes fall, and at the same time middle-class and especially upper middle-class incomes rise because the return to IQ rises with the increase not only in technologically complex jobs but also in managerial and other intellectually demanding service occupations. Society becomes more complex and increasing complexity increases the demand for IQ even if the complexity is not technological.
IQ is normally distributed, with a standard deviation of 15 from its arbitrarily scored mean of 100 (not a "score," really, but just a round number to designate the mean). Hence two-thirds of the population has an IQ between 85 and 115. Persons with an IQ below 85 are pretty much limited to working-class jobs; but that is probably also true of a majority of persons with an IQ between 85 and 100—and persons with IQs below 100 comprise half the population. The percentage of the population with IQs between 70 and 130 (two standard deviations from the mean) is 95 percent, which means that only 2.5 percent have an IQ above 130. The increased return to IQ provides far more opportunities to that intellectual elite than in a society in which the overwhelming demand is for manual labor, and in which the demand for high-IQ persons is largely limited to priests and tyrants.
Of course there are people with modest IQs who are immensely successful, including professional athletes and entertainers (though only a small minority command high incomes in either group). There are even very successful businessmen who owe their success to charisma or even sheer luck rather than to a high IQ, and there are a fair number of dopes who are in the upper middle class by reason of inheritance. And there are high-IQ people who are held back by poor physical or mental health. All this is just to say that IQ is by no means a conprehensive index of personal success; it does not even measure all dimensions of intellectual prowess. There is certainly no one-to-one correlation between IQ and income or family stability or crime, but IQ is an important factor in success in these domains.
If this analysis is correct, then (it seems to me) an essential means of narrowing the class divide, if that is thought an important goal, is redistributive tax and spending policies (which are costly, hence not worth undertaking if narrowing the class divide is not deemed important). Improving our educational system, even if possible, which it doesn't seem to be, wouldn't do much, because education can't do much to increase intelligence. Milton Friedman used to say that the problem of poor people is that they don't have money, and he advocated a negative income tax (which we now have in the form of the Earned Income Tax Credit). Increasing the income of the depressed working class might go far to increase family stability in that class, lower the crime rate, and perhaps even increase employment, provided the redistribution was coupled with the creation of incentives for working.
On the Apparently Growing Class Divide and What can be Done About it-Becker
The US has traditionally stood for a large amount of equality of opportunity, at least among whites. This implies that the success of children would depend mainly on their ability and energy, and much less so on their parents' incomes and education. This has always been an aspiration rather than a fact, and the disturbing evidence in several studies indicates that equality of opportunity has declined by a lot during the past half century. In a recent book, Coming Apart, Charles Murray highlights this apparent fact with interesting statistics.
Another way to state what has happened during the past half century is that the degree of mobility between generations has apparently declined. The evidence suggests that children of successful parents are more likely to be successful relative to others of their own generation than was true in the first half of the 20th century. One important reason for this, I believe, is that education is now a much more important determinant of economic success than it was in 1960. Educated parents have always been much better than parents with relatively little education at preparing their children to succeed at school. This difference in preparation of children is now more important because greater education has become more necessary to succeed in the modern knowledge-driven American economy.
One approach to thinking about the causes of this trend is to divide employment opportunities into good and bad jobs. Some discussions assume, implicitly or explicitly, that the number of good jobs is rather fixed, that many individuals of different classes are capable of filling these good jobs, and hence that who gets the better jobs depends on contacts, influence, and credentials, like having a college degree. On this approach to labor markets, children of upper class parents-those with greater income and education- are more likely to succeed now than in the past because their overall education and other "credentials" have increased compared to those of children from the lower classes.
This seems to be the implicit view of the job market behind an op-ed piece this week in the NY Times by Charles Murray. In discussing what can be done to reduce the advantage of children from the upper classes, he advocates eliminating unpaid internships, eliminating the use of SAT scores in determining college admissions, and ending the ability of companies to list a college degree as necessary to apply for certain jobs. On the view that the number of good jobs is rather fixed, that many young persons of all backgrounds are capable of handling these jobs, and that having a college education does not generally signify greater knowledge and other skills, his suggestions might reduce some of the "artificial" advantages that children from the upper classes have in getting these jobs.
A very different view of the labor market is much more consistent with the substantial growth in the number of "good" jobs during the past century in every developed country. On this human capital inerpretation, the number of good jobs is not fixed but depends on the skills of workers, so that companies provide many more good jobs when the skills of workers increase. This approach implies that children from the upper classes are much more likely to get good jobs because they have much better skill sets than do children from the lower classes. This skill set includes not just knowledge and information, but also the ability to get to work on time, to start and finish tasks successfully, and to get along with colleagues. Children from the lower classes have fallen further behind in their earnings because their skill sets have fallen further behind those of children from the upper classes.
The reasons why this has happened since 1960 are not fully known, but some changes since then do seem crucial. Children from the lower classes are now far more likely to be raised by single parents than they were in 1960. Guidance from both parents is of great importance in preparing children for modern labor markets, and for all other aspects of life. Educated upper class parents start teaching their children when they are very young considerable knowledge and also other traits that are valuable when the children grow up. My colleague James Heckman has stressed that lower class children are by age 5 already considerably behind children from the upper classes, and that these differences grow as children age. As a result, children from lower classes are much more likely to drop out of high school, and much less likely to receive a college education. They end up with worse jobs, greater unemployment, and much higher crime rates.
From the prospective that the number of good jobs depends on the skills of workers, Murray's proposals make little sense. The goal should not be to take away unpaid internships and other ways that improve learning, sometimes mainly of children from the upper classes. Rather, the goal should be to increase the skills of children from poorer, less educated, and less stable family backgrounds, so that they become more productive in the work place.
I am not suggesting this goal is easily accomplished. For example, how do we make parents more caring about their children, or reduce the tendency of women from the lower classes to have children without being married or without being in other stable relationships with the fathers of their children? The two most attractive options appear to be greater emphasis on early childhood education and better schools, but much further research is necessary. This research is more likely to be forthcoming once it is generally accepted that limited knowledge and limited other work skills are the main reasons why children from low educated and low income families have over time fallen further behind children from educated and wealthier families.