15. G. W. Harrison, J. A. List, J. Econ. Lit. 42, 1009 (2004).
16. D. A. Dillman, Mail and Internet Surveys: The Tailored
Design Method (Wiley, New York, 2000).
17. Beyond this statistical challenge, the Web is changing
how social science, along with all of science, is
conducted. For example, massive records of the Web
transactions themselves are data for analysis that uses
complexity theory and network theory to understand
social and economic networks (42, 43).
18. J. Witte in Society Online: The Internet in Context,
P. N. Howard, S. Jones, Eds. (Russell Sage Foundation,
New York, 2004), p. xv.
19. M. Schonlau, A. Van Soest, A. Kapteyn, M. Couper,
J. Winter, in preparation.
20. W. S. Bainbridge in Computing in the Social Sciences,
O. V. Burton, Ed. (University of Illinois Press, Urbana, IL,
2002), pp. 51–56.
21. U.S. Census Bureau Center for Economic Studies
22. Procedures and Costs for Use of the Research Data Center
23. J. M. Abowd, J. I. Lane, Tech. Pap. No. TP-2003-10
(U.S. Census Bureau, Washington, DC, 2003).
24. D. Livermore, E. F. Moran, R. R. Rindfuss, P. C. Stern, Eds.,
People and Pixels (National Academy Press, Washington,
25. W. D. Nordhaus, Proc. Natl. Acad. Sci. U.S.A. 103, 3510
26. J. E. Cohen, C. Small, Proc. Natl. Acad. Sci. U.S.A. 95,
27. B. C. O’Neill, F. L. MacKellar, W. Lutz, Population and
Climate Change (Cambridge Univ. Press, Cambridge,
2001), pp. 114–117.
28. L. Jin et al., Proc. Natl. Acad. Sci. U.S.A. 96, 3796 (1999).
29. V. L. Bonham, E. Warshauer-Baker, F. S. Collins, Am. Psychol.
60, 9 (2005).
30. S. M. McClure, D. I. Laibson, G. Loewenstein, J. D. Cohen,
Science 306, 503 (2004).
31. G. S. Berns et al., Science 312, 754 (2006).
32. J. H. Kagel, R. C. Battalio, L. Green, Economic Choice
Theory: An Experimental Analysis of Animal Behavior
(Cambridge Univ. Press, Cambridge, 1995).
33. Chicago Core on Biomarkers in Population-Based
Aging Research (http://biomarkers.uchicago.edu/
studiescollectingbiomarkers.htm) summarizes studies
collecting biomarkers in population settings.
34. J. Henrich, Science 312, 60 (2006).
35. D. C. Dennett, Breaking the Spell: Religion as a Natural
Phenomenon (Viking Press, New York, 2006).
36. Demographic and Health Surveys (www.measuredhs.com).
37. Luxembourg Income Study (www.lisproject.org).
38. World Values Survey (www.worldvaluessurvey.org).
39. K. W. Deutsch, J. Platt, D. Senghaas, Science 171, 450
40. P. M. Smith, B. B. Torrey, Science 271, 611 (1996).
Univ. Press, Cambridge, 2004).
42. M. E. J. Newman, A. L. Barabasdi, D. J. Watts, Eds., The
Structure and Dynamics of Complex Networks (Princeton
Univ. Press, Princeton, NJ, 2003).
43. D. J. Watts, Six Degrees: The Science of a Connected Age
(Norton, New York, 2003).
44. W.P.B. is a member of the Board of Reviewing Editors of
Science and former Director of the Division of Social,
Economic, and Behavioral Research at the NSF. B.B.T. is a
Fellow of the AAAS and former Executive Director of the
Commission on Behavioral and Social Sciences and
Education at the National Research Council.
Skill Formation and the Economics of
Investing in Disadvantaged Children
James J. Heckman
This paper summarizes evidence on the effects of early environments on child, adolescent, and
adult achievement. Life cycle skill formation is a dynamic process in which early inputs strongly
affect the productivity of later inputs.
our core concepts important to devising
have emerged from decades of independent
research in economics, neuroscience, and develop-
mental psychology (1). First, the architecture of the
brain and the process of skill formation are
influenced by an interaction between genetics and
individual experience. Second, the mastery of skills
that are essential for economic success and the
development of their underlying neural pathways
follow hierarchical rules. Later attainments build
on foundations that are laid down earlier. Third,
cognitive, linguistic, social, and emotional com-
petencies are interdependent; all are shaped power-
fully by the experiences of the developing child;
and all contribute to success in the society at large.
Fourth, although adaptation continues throughout
life, human abilities are formed in a predictable
sequence of sensitive periods, during which the
development of specific neural circuits and the be-
haviors they mediate are most plastic and therefore
optimally receptive to environmental influences.
A landmark study concluded that
every aspect of early human development, from
_s evolving circuitry to the child_s ca-
pacity for empathy, is affected by the environ-
ments and experiences that are encountered in a
cumulative fashion, beginning in the prenatal pe-
riod and extending throughout the early child-
[ (2). This principle stems from two
characteristics that are intrinsic to the nature of
learning: (i) early learning confers value on ac-
quired skills, which leads to self-reinforcing moti-
vation to learn more, and (ii) early mastery of a
range of cognitive, social, and emotional compe-
tencies makes learning at later ages more efficient
and therefore easier and more likely to continue.
Early family environments are major predic-
tors of cognitive and noncognitive abilities.
Research has documented the early (by ages 4
to 6) emergence and persistence of gaps in cog-
nitive and noncognitive skills (3, 4). Environ-
ments that do not stimulate the young and fail to
cultivate these skills at early ages place children
at an early disadvantage. Disadvantage arises
more from lack of cognitive and noncognitive
stimulation given to young children than simply
from the lack of financial resources.
This is a source of concern because family
environments have deteriorated. More U.S. chil-
dren are born to teenage mothers or are living in
single parent homes compared with 40 years ago
(5). Disadvantage is associated with poor parent-
ing practices and lack of positive cognitive and
noncognitive stimulation. A child who falls be-
hind may never catch up. The track records for
criminal rehabilitation, adult literacy, and public job
training programs for disadvantaged young adults
are remarkably poor (3). Disadvantaged early en-
Department of Economics, University of Chicago, Chicago, IL
60637, USA. Department of Economics, University College
Dublin, Dublin 4, Ireland. E-mail: firstname.lastname@example.org
Fig. 1. Average percentile rank on Peabody Individual Achievement Test–Math score by age and income
quartile. Income quartiles are computed from average family income between the ages of 6 and 10.
Adapted from (3) with permission from MIT Press.
L I F E C Y C L E S
30 JUNE 2006
on a number of social and economic measures.
Many major economic and social problems
can be traced to low levels of skill and ability in the
population. The U.S. will add many fewer college
graduates to its workforce in the next 20 years than
it did in the past 20 years (6, 7). The high school
dropout rate, properly measured with inclusion of
individuals who have received general educational
development (GED) degrees, is increasing at a
time when the economic return of schooling has
increased (8). It is not solely a phenomenon of
unskilled immigrants. Over 20% of the U.S.
workforce is functionally illiterate, compared with
about 10% in Germany and Sweden (9). Violent
crime and property crime levels remain high,
despite large declines in recent years. It is
estimated that the net cost of crime
in American society is $1.3 trillion
per year, with a per capita cost of
$4818 per year (10). Recent research
documents the importance of deficits
in cognitive and noncognitive skills
in explaining these and other social
Noncognitive Skills and Examples
of Successful Early Interventions
Cognitive skills are important, but
noncognitive skills such as motiva-
tion, perseverance, and tenacity are
also important for success in life.
Much public policy, such as the No
Child Left Behind Act, focuses on
cognitive test score outcomes to
measure the success of interventions
in spite of the evidence on the im-
portance of noncognitive skills in
social success. Head Start was deemed
a failure in the 1960s because it did
not raise the intelligence quotients
(IQs) of its participants (12). Such
judgments are common but miss the
larger picture. Consider the Perry
Preschool Program (13), a 2-year
experimental intervention for disad-
vantaged African-American children
initially ages 3 to 4 that involved morning pro-
grams at school and afternoon visits by the teacher
to the child’s home. The Perry intervention group
had IQ scores no higher than the control group by
age 10. Yet, the Perry treatment children had higher
achievement test scores than the control children
because they were more motivated to learn. In
followups to age 40, the treated group had higher
rates of high school graduation, higher salaries,
higher percentages of home ownership, lower
rates of receipt of welfare assistance as adults,
fewer out-of-wedlock births, and fewer arrests
than the controls (13). The economic benefits of
the Perry Program are substantial (Table 1). Rates
of return are 15 to 17% (14). (The rate of return is
the increment in earnings and other outcomes,
suitably valued, per year for each dollar invested
in the child). The benefit-cost ratio (the ratio of
the aggregate program benefits over the life of
the child to the input costs) is over eight to one.
Perry intervened relatively late. The Abecedar-
ian program, also targeted toward disadvantaged
children, started when participants were 4 months
of age. Children in the treatment group received
child care for 6 to 8 hours per day, 5 days per week,
through kindergarten entry; nutritional supple-
ments, social work services, and medical care
were provided to control group families. The
program was found to permanently raise the IQ
and the noncognitive skills of the treatment group
over the control group. However, the Abecedarian
program was intensive, and it is not known
whether it is the age of intervention or its inten-
sity that contributed to its success in raising IQ
Reynolds et al. present a comprehensive
review of early childhood programs directed
toward disadvantaged children and their impact
(18). Similar returns are obtained for other early
intervention programs (19, 20), although more
speculation is involved in these calculations be-
cause the program participants are in the early
stages of their life cycles and do not have long
Schools and Skill Gaps
Many societies look to the schools to reduce skills
gaps across socioeconomic groups. Because of the
dynamics of human skill formation, the abilities
and motivations that children bring to school play
a far greater role in promoting their performance
in school than do the traditional inputs that receive
so much attention in public policy debates. The
Coleman Report (21) as well as recent work
(22, 23) show that families and not schools are the
major sources of inequality in student performance.
By the third grade, gaps in test scores across socio-
economic groups are stable by age, suggesting that
later schooling and variations in schooling quality
have little effect in reducing or widening the gaps
that appear before students enter school (4, 24).
Figure 1 plots gaps in math test scores by age across
family income levels. The majority of the gap at age
12 appears at the age of school enrollment. Carneiro
and Heckman performed a cost-benefit analysis of
classroom size reduction on adult earnings (3).
Although smaller classes raise the
adult earnings of students, the earn-
ings gains received by students do not
offset the costs of hiring additional
teachers. The student-teacher achieve-
ment ratio (STAR) randomized trial
of classroom size in Tennessee shows
some effect of reduced classroom size
on test scores and adult perform-
ance, but most of the effect occurs in
the earliest grades (25, 26). Schools
and school quality at current levels
of funding contribute little to the
emergence of test score gaps among
children or to the development of
Second Chance Programs
America is a second chance society.
Our educational policy is based on a
fundamental optimism about the
possibility of human change. The
dynamics of human skill formation
reveal that later compensation for de-
ficient early family environments is
very costly (4). If society waits too
long to compensate, it is economical-
ly inefficient to invest in the skills of
the disadvantaged. A serious trade-off
exists between equity and efficiency
for adolescent and young adult skill policies.
There is no such trade-off for policies targeted
toward disadvantaged young children (28).
The findings of a large literature are captured in
Fig. 2. This figure plots the rate of return, which is
the dollar flow from a unit of investment at each
age for a marginal investment in a disadvantaged
young child at current levels of expenditure. The
economic return from early interventions is high,
and the return from later interventions is lower.
Remedial programs in the adolescent and young
adult years are much more costly in producing the
same level of skill attainment in adulthood. Most
are economically inefficient. This is reflected in
Fig. 2 by the fact that a segment of the curve lies
below the opportunity cost of funds (the horizon-
Fig. 2. Rates of return to human capital investment in disadvantaged children. The
declining figure plots the payout per year per dollar invested in human capital
programs at different stages of the life cycle for the marginal participant at current
levels of spending. The opportunity cost of funds (r) is the payout per year if the
dollar is invested in financial assets (e.g., passbook savings) instead. An optimal
investment program from the point of view of economic efficiency equates returns
across all stages of the life cycle to the opportunity cost. The figure shows that, at
current levels of funding, we overinvest in most schooling and post-schooling
programs and underinvest in preschool programs for disadvantaged persons.
Adapted from (3) with permission from MIT Press.
return from funds if they were invested for
purposes unrelated to disadvantaged children.
Investing in disadvantaged young children is a
rare public policy initiative that promotes fairness
and social justice and at the same time promotes
productivity in the economy and in society at
large. Early interventions targeted toward
disadvantaged children have much higher returns
than later interventions such as reduced pupil-
teacher ratios, public job training, convict reha-
bilitation programs, tuition subsidies, or expend-
iture on police. At current levels of resources,
society overinvests in remedial skill investments
at later ages and underinvests in the early years.
Although investments in older disadvantaged
individuals realize relatively less return overall, such
investments are still clearly beneficial. Indeed, the
advantages gained from effective early inter-
ventions are sustained best when they are followed
by continued high-quality learning experiences. The
technology of skill formation shows that the returns
on school investment and postschool investment are
higher for persons with higher ability, where ability
is formed in the early years. Stated simply, early
investments must be followed by later investments
if maximum value is to be realized.
References and Notes
1. E. I. Knudsen, J. J. Heckman, J. Cameron, J. P. Shonkoff,
Proc. Natl. Acad. Sci. U.S.A., in press.
2. J. P. Shonkoff, D. Phillips, From Neurons to Neighborhoods:
The Science of Early Child Development (National
Academies Press, Washington, DC, 2000).
3. P. Carneiro, J. J. Heckman, in Inequality in America: What
Role for Human Capital Policies? J. J. Heckman, A. B.
Krueger, B. Friedman, Eds. (MIT Press, Cambridge, MA,
2003), ch. 2, pp. 77–237.
4. F. Cunha, J. J. Heckman, L. J. Lochner, D. V. Masterov,
in Handbook of the Economics of Education,
E. A. Hanushek, F. Welch, Eds. (North Holland,
Amsterdam, in press).
5. J. J. Heckman, D. V. Masterov, ‘‘The productivity argument for
investing in young children,’’ (Working Paper No. 5, Committee
on Economic Development, Washington, DC, 2004).
6. J. B. Delong, L. Katz, C. Goldin, in Agenda for the Nation,
H. Aaron, J. Lindsay, P. Nivola, Eds. (Brookings Institution
Press, Washington, DC, 2003), pp. 17–60.
7. D. T. Ellwood, in The Roaring Nineties: Can Full Employment
Be Sustained? A. Krueger, R. Solow, Eds. (Russell Sage
Foundation, New York, 2001), pp. 421–489.
8. J. J. Heckman, P. LaFontaine, J. Lab. Econ., in press.
9. International Adult Literacy Survey, 2002: User’s Guide,
Statistics Canada, Special Surveys Divison, National
Literacy Secretariat, and Human Resources Development
Canada (Statistics Canada, Ottawa, Ontario, 2002).
10. D. A. Anderson, J. Law Econ. 42, 611 (1999).
11. J. J. Heckman, J. Stixrud, S. Urzua, J. Lab. Econ., in press.
12. Westinghouse Learning Corporation and Ohio University,
The Impact of Head Start: An Evaluation of the Effects of
Head Start on Children’s Cognitive and Affective
Development, vols. 1 and 2 (Report to the Office of
Economic Opportunity, Athens, OH, 1969).
13. L. J. Schweinhart et al., Lifetime Effects: The High/Scope
Perry Preschool Study Through Age 40 (High/Scope,
Ypsilanti, MI, 2005).
14. A. Rolnick, R. Grunewald, ‘‘Early childhood development:
Economic development with a high public return’’ (Tech.
rep., Federal Reserve Bank of Minneapolis, Minneapolis,
15. C. T. Ramey, S. L. Ramey, Am. Psychol. 53, 109 (1998).
16. C. T. Ramey, S. L. Ramey, Prev. Med. 27, 224 (1998).
17. C. T. Ramey et al., Appl. Dev. Sci. 4, 2 (2000).
18. A. J. Reynolds, M. C. Wang, H. J. Walberg, Early
Childhood Programs for a New Century (Child Welfare
League of America Press, Washington, DC, 2003).
19. L. A. Karoly et al., Investing in Our Children: What We
Know and Don’t Know About the Costs and Benefits of Early
Childhood Interventions (RAND, Santa Monica, CA, 1998).
20. L. N. Masse, W. S. Barnett, A Benefit Cost Analysis of the
Abecedarian Early Childhood Intervention (Rutgers
University, National Institute for Early Education
Research, New Brunswick, NJ, 2002).
21. J. S. Coleman, Equality of Educational Opportunity (U.S.
Deparment of Health, Education, and Welfare, Office of
Education, Washington, DC, 1966).
22. S. W. Raudenbush, ‘‘Schooling, statistics and poverty:
Measuring school improvement and improving schools’’
Inaugural Lecture, Division of Social Sciences, University
of Chicago, Chicago, IL, 22 February 2006.
23. J. J. Heckman, M. I. Larenas, S. Urzua, unpublished data.
24. D. A. Neal, in Handbook of Economics of Education,
E. Hanushek, F. Welch, Eds. (Elsevier, Amsterdam, in
26. B. Krueger, D. M. Whitmore, in Bridging the Achievement
Gap, J. E. Chubb, T. Loveless, Eds. (Brookings Institution
Press, Washington, DC, 2002).
27. W. S. Barnett, Benefit-Cost Analysis of Preschool Education,
28. F. Cunha, J. J. Heckman, J. Hum. Resour., in press.
29. This paper was generously supported by NSF (grant nos.
SES-0241858 and SES-0099195), National Institute of
Child Health and Human Development (NIH grant no.
R01HD043411), funding from the Committee for Eco-
nomic Development, with a grant from the Pew
Charitable Trusts and from the Partnership for America’s
Economic Success. This research was also supported by
the Children’s Initiative project at the Pritzker Family
Foundation and a grant from the Report to the Nation of
America’s Promise. The views expressed in this paper are
those of the author and not necessarily those of the
sponsoring organizations. See our Web site (http://
jenni.uchicago.edu/econ_neurosci) for more information.
Linda M. Richter
Young people in their teens constitute the largest age group in the world, in a special stage
recognized across the globe as the link in the life cycle between childhood and adulthood.
Longitudinal studies in both developed and developing countries and better measurements of
adolescent behavior are producing new insights. The physical and psychosocial changes that occur
during puberty make manifest generational and early-childhood risks to development, in the form
of individual differences in aspects such as growth, educational attainment, self-esteem, peer
influences, and closeness to family. They also anticipate threats to adult health and well-being.
Multidisciplinary approaches, especially links between the biological and the social sciences, as well
as studies of socioeconomic and cultural diversity and determinants of positive outcomes, are
needed to advance knowledge about this stage of development.
oung people aged 10 to 19 currently
are the largest age group in the world,
making up close to 20% of the 6.5 billion world
population estimated in 2005 (1), 85% of whom
live in developing countries and account for
about one-third of those countries
_ national pop-
ulations. Adolescence has also been described as
Bdemographically dense[: a period in life during
which a large percentage of people experience a
large percentage of key life-course events (2).
These include leaving or completing school,
bearing a child, and becoming economically
productive. They also include experiences, more
common in this age group than in others, that
are capable of substantially altering life trajecto-
ries: nonconsensual sex, alcohol and drug abuse,
self-harm and interpersonal violence, and getting
into trouble with the law. Diet and activity pat-
terns, friendships, educational achievement, and
civic involvement all affect current health,
Child, Youth, Family, and Social Development, Human
Sciences Research Council, Private Bag X07, Dalbridge
4014, South Africa, and University of KwaZulu-Natal,
South Africa. E-mail: email@example.com
Table 1. Economic benefits and costs of the Perry
Preschool Program (27). All values are discounted at
3% and are in 2004 dollars. Earnings, Welfare, and
Crime refer to monetized value of adult outcomes
(higher earnings, savings in welfare, and reduced
costs of crime). K–12 refers to the savings in reme-
dial schooling. College/adult refers to tuition costs.
Net present value