Sunday, September 14, 2008

The growing polarisation of American society and its implications for productivity

James J. Heckman
from VoxEU, a global economics forum
--originally posted August 25, 2008

America has a growing skills problem. This column emphasizes the importance of early environments in determining skills. It suggests that to promote skills, public policy should refocus attention to the early years of childhood and away from its current emphasis on the later years.

America has a growing skills problem. One consequence of this skills problem is rising inequality and polarization of society. A greater fraction of young Americans are graduating from college. At the same time, a greater fraction are dropping out of high school. Trends in the production of skills from American high schools coupled with a growing influx of unskilled immigrants have produced an increasing proportion of low-skilled workers in the US workforce. More than 20% of American workers cannot understand the instructions written in a medical prescription. A further consequence of the skills problem is a slowdown in growth of productivity of the workforce.

The origin of this skills problem lies in the decline of the family in American society. Dysfunctional families retard the formation of the abilities needed for successful performance in modern society.

The importance of cognitive and noncognitive abilities
American public policy currently focuses on cognitive test scores or “smarts.” Yet an emerging literature shows that much more than smarts is required for success in life. Motivation, sociability, the ability to work with others, the ability to focus on tasks, self-regulation, self-esteem, time preference, health, and mental health all matter. In an earlier time, these traits were part of what was called “character.” A substantial body of research shows that earnings, employment, labour force experience, college attendance, teenage pregnancy, participation in risky activities, compliance with health protocols, and participation in crime are all strongly affected by non-cognitive as well as cognitive abilities. Heckman, Stixrud and Urzua (2006) show that in many dimensions of social performance, non-cognitive traits are as important, or more important, than cognitive traits in predicting success.

Compelling evidence on the importance of non-cognitive skills comes from the GED (General Education Degree) programme (Heckman and Rubinstein, 2001, and Heckman and LaFontaine 2008). GED recipients are high school dropouts who pass a test to certify that they are equivalent to high school graduates.
Currently, 14% of US high school certificates are issued to GEDs. Previous research shows that the cognitive test scores of GED recipients and the cognitive test scores of persons who graduate from high school but do not go on to college are comparable. Yet GED recipients have the earnings of high school dropouts. GED recipients are as “smart" as ordinary high school graduates, yet they lack non-cognitive skills. GED recipients are the “wise guys” who cannot finish anything. They quit the jobs and marriages they start at much greater rates than ordinary high school graduates. Most
branches of the US military recognize this in their recruiting strategies. Until the recent war in Iraq, the armed forces did not generally accept GED recipients because of their poor performance in the military.

Ability gaps open up early in life
Gaps in both cognitive and non-cognitive skills between advantaged and disadvantaged children emerge early and can be attributed, in part, to adverse early environments into which an increasing percentage of US children are being born. Figure 1 shows the gap in cognitive test scores by age of children stratified by the mother's education. Similar patterns are found for non-cognitive skills (see Heckman, 2008, and Cunha, Heckman, Lochner and Masterov, 2006). Gaps in ability emerge early and persist. Most of the gaps in ability at age 18, which substantially explain gaps in adult
outcomes, are present at age five. Schooling plays a minor role in creating or perpetuating gaps, even though American children go to very different schools depending on their family backgrounds. Test scores for children with very different family backgrounds are remarkably parallel with age.

Figure 1. Trend in mean cognitive score by maternal education, IHDP study
Note: Using all observations and assuming that data are missing at random. Source: Brooks-Gunn, Cunha, Duncan, Heckman, and Sojourner (2006).

How do these early and persistent differences in abilities arise?
Is the difference due to genes as Herrnstein and Murray claimed in The Bell Curve?
Evidence from the recent literature in psychology and biology suggests that the genes versus environment distinction that was once much in vogue is obsolete. Extensive recent literature suggests that gene-environment interactions are central to explaining children’s intellectual development. For example, breast-fed children attain higher IQ scores than non-breast fed children. This relationship is moderated by a gene that controls fatty acid pathways. Identical twins are affected by life experiences that substantially differentiate the genetic expression of adult twins. Further, the impact on adult antisocial behaviour of growing up in a harsh or abusive environment depends on the absence of a variant of a particular gene. A substantial literature shows that family environments play an independent role in creating adult abilities. Adverse family environments of children
create problem adults.

The decline of the American family and the rise of social problems
The evidence on the importance of family factors in explaining ability gaps is a source of concern because a greater proportion of American children are being born into disadvantaged environments, where disadvantage is measured by the quality of parenting (Heckman, 2008). A divide is opening up in American society. Those children born into disadvantaged environments are receiving relatively less stimulation and resources to promote child development than those born into more advantaged families. Women who are more educated are working and earning more. Their families are more stable and mothers in these families are also devoting more time to child development activities than less educated women. Children in affluent homes are bathed in financial and cognitive resources. Those children born into less advantaged circumstances are much less likely to receive cognitive and socio-emotional stimulation and other family resources. The family environments of single parent homes compared to intact families are much less favourable for investment in children (Moon, 2008).

Enriching early environments can partially compensate for early adversity
Experiments that enrich the early environments of disadvantaged children demonstrate causal effects of early environments on adolescent and adult outcomes and provide powerful evidence against genetic determinism. Two of these investigations, the Perry Preschool Program and the Carolina Abecedarian Project, use a random assignment design and collect long-term follow-up data. They demonstrate substantial positive effects of early environmental enrichment on a range of cognitive and non-cognitive skills, schooling achievement, job performance, and social behaviours long after the interventions end. The Perry Program was administered to 58 disadvantaged African-American children in Ypsilanti, Michigan between 1962 and 1967. The treatment for this program consisted of a daily 2.5-hour classroom session on weekday mornings and a weekly 90-minute home visit by the teacher on weekday afternoons. The control and treatment groups have been followed through age 40. There is a consistent pattern of successful outcomes for treatment group members compared with control group members, even though an initial increase in IQ gradually disappeared within the four years following the intervention.

Such IQ fadeouts have been observed in other studies. Focus on cognitive skills alone misses the point. The Perry program operates primarily through improving the non-cognitive traits of participants (Heckman, Malofeeva, Pinto and Savelyev, 2008). At the oldest ages tested, treated individuals scored higher on achievement tests, attained higher levels of education, required less special education, earned higher wages, were more likely to own a home, and were less likely to go on welfare or be incarcerated than controls even though their IQs were no higher than those in the control group. In the similar, but more intensive and earlier starting Abecedarian program, IQ gains were found to last into early adulthood.

An estimated rate of return (the return per dollar of cost) to the Perry Program is around 10%. This high rate of return is higher than the post-World War II return on US stock market equity (5.8%) and suggests that society at large can benefit substantially from such interventions in the lives of disadvantaged children. Interventions in the later lives of disadvantaged children, such as job training, convict rehabilitation, and reduced classroom sizes, have much lower returns (Cunha, Heckman, Lochner and Masterov, 2006).

Using an empirically determined technology of skill formation, Cunha and Heckman (2006) simulate an early childhood intervention that moves children from the bottom 10% of family resources to the 70th percentile. This intervention achieves Perry results. To achieve similar results using adolescent interventions requires spending 35-50% more in present value terms (Heckman, 2008).


Fifty percent of the variance in inequality in lifetime earnings is determined by age 18 (Cunha and Heckman, 2007). The family plays a powerful role in shaping adult outcomes that is not fully recognized by current American policies. As programs are currently configured, interventions early in the lives of disadvantaged children have substantially higher economic returns than later interventions such as reduced pupil-teacher ratios, public job training programs, convict rehabilitation programs, adult
literacy programs, tuition subsidies, or expenditure on police. This is because life-cycle skill formation is dynamic in nature. Skill begets skill; motivation begets motivation. Motivation cross-fosters skill, and skill cross-fosters motivation. If a child is not motivated to learn and engage early on in life, the more likely it is that when the child becomes an adult, he or she will fail in social and economic life. The longer society waits to intervene in the life cycle of a disadvantaged child, the more costly it is to remediate disadvantage.


Brooks-Gunn, J., F. Cunha G. Duncan J. J. Heckman, and A. Sojourner. “A Reanalysis of the IHDP Program,” 2006. Unpublished manuscript, Infant Health and Development Program, Northwestern University.

Cunha, F. and J. J. Heckman. “Investing in our Young People.", 2006. Unpublished manuscript, University of Chicago, Department of Economics.

Cunha, F. and J. J. Heckman. “The Evolution of Uncertainty in Labour Earnings in the US Economy,” 2007. Unpublished manuscript, University of Chicago. Under revision.

Cunha, F., J. J. Heckman, L. J. Lochner and D. V. Masterov. “Interpreting the Evidence on Life Cycle Skill Formation,” eds. E.A. Hanushek and F. Welch, 2006. Handbook of the Economics of Education, 12, 697-812, Amsterdam: North-Holland.

Heckman, J. J. “Schools, Skills, and Synapses”, Fall 2008. Economic Inquiry, 289-324.

Heckman, J. J. and P. A. LaFontaine. “The GED and the Problem of Noncognitive Skills in America,” 2008. Unpublished book manuscript, University of Chicago, Department of Economics.

Heckman, J. J., L. Malofeeva, R. Pinto and P. Savelyev. “The Effect of the Perry Preschool Program on Cognitive and Noncognitive Skills: Beyond Treatment Effects,” 2008. Unpublished manuscript, University of Chicago, Department of Economics.

Heckman, J. J. and Y. Rubinstein. “The Importance of Noncognitive Skills: Lessons from the GED Testing Program,” May 2001. American Economic Review, 91(2), 145-149.

Heckman, J. J., J. Stixrud and S. Urzua. “The Effects of Cognitive and Noncognitive Abilities on Labour Market Outcomes and Social Behavior,” July 2006. Journal of Labour Economics, 24(3), 411-482.

Herrnstein, R. J. and C. A. Murray, 1994. The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press.

Moon, S. H. “Skill Formation Technology and Multi-dimensional Parental Investment,” 2008. Unpublished PhD thesis, University of Chicago, Department of Economics.

This article may be reproduced with appropriate attribution.


Post a Comment

Links to this post:

Create a Link

<< Home