James J. Heckman and Alan B. Krueger.
Inequality in America: What Role for Human Capital Policies?
mit Press. 384 pages. $40.00
John H. Laub and Robert J. Sampson.
Shared Beginnings, Divergent Lives: Delinquent Boys to Age 70.
Harvard University Press. 352 pages. $49.95
One assumption of both social liberals and limited-government conservatives in debates over social policy is that family issues are too personal and private for government intervention. Mutual care and the rearing of children may be essential for the perpetuation of society, but what happens in the family, the argument goes, is the exclusive province of individual members of the household. This assumption is part of the reason that some policymakers recoil at the thought of debates over such seemingly disparate issues as gay marriage, contraception, and President Bush’s healthy marriage initiative.
Not comfortable addressing the nature of the family, policymakers usually restrict social policy proposals to “preventative” measures such as Head Start, tuition subsidies, or vocational training. Of these, there is no shortage of proposals, but there is a dearth of evidence on their relative effectiveness. The problem with “preventative” measures is that sometimes they come too late to prevent anything. Some students in Head Start programs fail to graduate, and some young adults who play in midnight basketball leagues fall into delinquency.
The failure of such policies to stem the surge of inequality in income and wealth in the United States over the past 25 years is the subject of Inequality in America: What Role for Human Capital Policies? by James J. Heckman and Alan B. Krueger. The book is structured to reflect the symposium that was its source; it begins with one essay each by the book’s principal authors evaluating human capital policies meant to improve the skills of the least-skilled members of the work force. Brief comments follow from four noted economists, including the redoubtable Lawrence Summers, president of Harvard University.
Krueger, a professor of economics at Princeton, is among the most prominent scholars advocating significant reductions in class sizes as a method for improving public schools. His contribution to the volume prescribes the same policies pursued by the educational establishment over the past two decades — spend more to reduce class sizes, increase teacher training, and expand an array of programs from early childhood to post-secondary years, including Head Start, Job Corps, and tuition subsidies. Krueger marshals what evidence exists for the success of these programs, but his proposals for substantial investment in education and other human capital policies run up against a problem: The return is not worth the investment. The policies Krueger advocates have been pursued over the past 25 years, though perhaps not to the degree he would like. But consider this: Real spending per pupil in public schools (adjusted for inflation) was 240 percent higher in 2000 than in 1960. While school resources have increased, reliable measures such as the National Assessment of Educational Progress show that student performance has remained flat. All the while, an increasing number of low-income Americans lack the skills to succeed in the labor market.
The problem of diminishing returns for human capital investment is diagnosed in a chapter by James J. Heckman, a Nobel Prize-winning professor of economics at the University of Chicago, and economist Pedro Carneiro. Rather than asking whether the treatment supplied by various human capital policies produces significant effects, they analyze the costs of these policies compared to their benefits. Many are found wanting. The returns from job-training programs, it turns out, are marginal because an individual’s cognitive and noncognitive skills are nearly developed by the time he or she becomes eligible for these programs. There is a relatively small percentage of the population, about 8 percent, that is “credit-constrained” and could benefit from tuition subsidies, but in most cases such subsidies benefit the upper-middle class.
The difference between the policy recommendations of Heckman and Carneiro, on one hand, and Krueger, on the other, flows from the different levels of attention they give to the importance of skills developed early in life. For example, scholars agree that there is a strong relationship between family income and college attendance. The most common way to interpret this evidence is the obvious one: Income provides the ability to purchase goods that lead to educational success, and students who do not attend college are constrained by an inability to afford tuition. Heckman and Carneiro propose a different interpretation — better family resources during a child’s formative years are associated with a higher quality of education and a better environment for fostering cognitive skills such as verbal ability and noncognitive habits, including self-discipline, which improve life chances. They analyze data to show that for most people an environment in which a child can develop important skills and habits is quantitatively more important than resources alone.
In reading a book full of regression tables and careful social science analysis, it is all too easy for a view of the trees to obscure the forest. Heckman and Carneiro’s fundamental conclusion deserves emphasis. They argue that instead of providing funds to improve skills late in life, government gets more bang for its buck with early childhood programs. In their view, successful policies strengthen early childhood education and provide incentives to improve public schools. Less attention is paid to how to strengthen the family’s potential for instilling skills and habits necessary for success in the labor market, but surely that is an area for future research.
Both Krueger and Heckman and Carneiro share a concern about a growing underclass that lacks the ability to create wealth. Krueger’s policy recommendations, while not particularly novel, are refreshing in their focus on results. All too often policymakers propose continuous reform as if the process of reform itself were a solution, without bothering to evaluate what particular reforms accomplish. The papers contained in Inequality in America advance the debate over what to do about inequality by proposing some standards for human capital policy reform.
While there is much to recommend in Heckman and Carneiro’s skeptical economic analysis of the often excessive cost of human capital programs, there is a bit too much dismay in their conclusions, even for practitioners of the dismal science. Their analysis suggests a disturbing conclusion: It is either impossible or excessively expensive for human capital policies to significantly improve the life chances of the poor after early childhood. The authors go so far as to suggest that income subsidies rather than training programs would be a more efficient way to reduce inequality.
While the federal government may be littered with well-intentioned but hopelessly inefficient programs, the facts should not lead us to such extreme despair about the plight of the poor. Large-scale quantitative analysis of social problems, while useful, should be the beginning rather than the end of the debate about policy solutions. Traditional quantitative approaches focus on the relationship among variables across individuals. This method leaves something to be desired because there are important changes that occur within individual cases which have to do with the accumulation of experiences during a lifetime. For example, economists assume that static individual variables such as income, race, iq, or childhood delinquency are responsible for setting a young person on a criminal trajectory, and in the aggregate this may appear to be true. But some indigent children with poor verbal skills and limited self-control do achieve economic success, and some juvenile delinquents desist from crime and build stable and productive lives.
The best way to counter this despair is to look to cases in which individuals have left poverty or criminality. In their book Shared Beginnings, Divergent Lives: Delinquent Boys to Age 70, John H. Laub and Robert J. Sampson examine the life histories of juvenile delinquents and argue that major life events are better predictors of crime than the childhood risk factors that are the focus of so much economic and criminological analysis. Using a combination of quantitative and qualitative data, the authors ask, “Why do some juvenile offenders persist in committing criminal acts as they grow older while others desist?”
They find that the classic factors alone — low verbal iq, difficult temperaments, and parental criminality — do not distinguish chronic criminals from desisters. They write that “the fact remains that there are important differences in adult criminal trajectories that cannot be predicted from childhood, contra the ‘national summits’ of the policy world and apparently much yearning among criminologists.”
To determine what separates desisters from career criminals, the authors look closely at life histories and change within individuals. They find three major “pathways” out of delinquency, each of which provides discipline, informal social control, and connectedness. The first two — military service and a steady job — were crucial turning points for some men in their desistance from crime. Both of these turning points helped to break bad habits and provide alternatives to criminal behavior. The authors write that “the simple fact is that people who work are kept busy and are least likely to get into trouble.”
Before many of the men found steady work, however, they had to develop habits of self-discipline. For men who lacked self-control, work was hard to find. The military was a boon for some, but it seemed to exacerbate problems like alcoholism in others. Prison was either a neutral experience or a harm. By far, the most frequently cited path out of crime was marriage. Many of the men who were frequent offenders in their youth were involved in delinquent peer groups which reinforced criminal and, sometimes, alcoholic habits. Marriage broke the cycle for some men. Often, the wife would intervene and move the husband into new habits and peer groups through nagging or subtler means.
Marriage is important quantitatively, as the most frequently cited reason for desistence, but exactly how marriage is important comes through in the interviews the authors conduct as the former delinquents near age 70, nearly a half-century after the original study. Take the case of Leon, a drifter from a home with weak parental supervision who was given to theft and drinking binges. He claims that “If I hadn’t met my wife at the time I did, I’d probably be dead.” His wife says she took a chance on meeting him — she knew he was in trouble as an adolescent — but she was able to change his routines to center around home rather than around a delinquent peer group. As a married man, Leon held a steady job and did not go out with the guys.
Marriage, then, is a more potent remedy than incarceration or reform school, which in the study appear to do more harm than good. Preventative social policies may indeed come too late, but it is far from clear that the delinquent poor are unable to escape their situation. The authors do a service by showing how some individuals are able to make a dramatic break with a criminal past when they encounter turning points in life.
Aside from the novel policy implications of their work, Sampson and Laub’s data make a significant contribution to the criminology literature. They use a rigorous combination of quantitative and qualitative methods to analyze what may be the longest longitudinal dataset of crime in the world. They begin by performing regression analysis on data for 500 juvenile delinquents who were remanded to reform school in Boston in the 1940s. The original data, collected in Sheldon and Eleanor Glueck’s classic study Unraveling Juvenile Delinquency (Harvard University Press, 1950), tracked the men up to age 25 and followed up at age 32. Sampson and Laub go a step further by reanalyzing the original data and then locating and interviewing 52 of these men after they had last been seen in the 1960s.
Their study considers only white males, which is both a virtue and a vice: The study is able to hold more factors constant, which strengthens comparisons, but it also necessarily leaves out important groups. The book suffers the defect of much of social science in that it is occasionally hobbled by unwieldy jargon. There is redemption, though, in moments of glittering insight, such as when the authors note that while the former delinquents say they are quite happy, they retain a bit of melancholy about their status as family men: “bonded relations are fused with ambivalence — dependence, even when welcomed, entails a certain entrapment.”
Impressive as Sampson and Laub’s scholarship is, their book deserves a hearing beyond academe. While rigorous economic analysis in studies such as Inequality in America may cast doubt on the efficacy of social programs for people who have fallen into poverty and crime, Sampson and Laub offer some reason to think that individuals can be redeemed. Their work, like a growing body of social science, points to the stable family as the most important producer of the social norms and skills that produce self-discipline and success in the labor market. The family, it turns out, is important for both children and adults.
The question becomes what can public policy do, if anything, to strengthen this locus of social capital. Is it enough to encourage — or even coerce — people to enter into work and family relationships in order to stem criminal activity and foster life skills? Would coercion even produce lasting change? The authors find that it does, and they quote approvingly a classic work on crime:
We need not try to develop deep and lasting interests, be they values or personality traits, in order to produce the behavior we want. It is enough to create situations which will coerce people into behaving as we want them to and then to create the conditions under which other rewards will become linked to continuing their behavior.
The difficulty is that policymakers have an instinctive reaction against government involvement in family policy. This reaction may be warranted — after all, the federal government may be too much of a blunderbuss to be able to prescribe appropriate family behavior. But forbidding government policy from addressing the family should be a prudential decision, not a principled one. Connections to steady employment, a church, or an organization like the military may be enough for some people to develop the skills and habits necessary for success in the labor market, or at least to keep them out of crime, but the power of these groups is limited. For many other people, marriage and family — the source of so much inequality — are still also the most efficient and effective remedies for problems that plague the most disadvantaged people, the poor and delinquent.
