10 1 1 193 6844

The Future of Children
The Future
of Children
Winter/Spring 2002
Children and Welfare Reform
Volume 12 Number 1
Volume 12 Number 1
Winter/Spring 2002
The Future
of Children
Board of Trustees
Susan Packard Orr,
Chairman
Nancy Burnett,
Vice Chairman
Jane Lubchenco
Dean O. Morton
Franklin M. Orr, Jr.
Julie Packard,
Vice Chairman
Lewis E. Platt
William K. Reilly
Allan Rosenfield
Richard T. Schlosberg, III,
President and CEO
Robert Stephens
Colburn S. Wilbur
Honorary Emeriti
Trustees
Robin Chandler Duke
Robert J. Glaser, M.D.
Frank H. Roberts, Esq.
Edwin E. van Bronkhorst
(ISSN 1054-8289) 2002
by The David and Lucile
Packard Foundation, all
rights reserved. Printed
in the United States of
America. Cover photo
Stone/Bruce Ayres
Printed on acid-free,
recycled paper with soy Ink.
(The electronic edition of this
issue can be found at
http://www.futureofchildren.org
on the World Wide Web.)
www.futureofchildren.org
Selected Bibliography
Volume 12 Number 1
Winter/Spring 2002
Published by
The David and Lucile Packard Foundation
Editor-in-Chief
Richard E. Behrman, M.D.
Editorial Advisory Board
Issue Editor
Margie K. Shields, M.P.A.
Issue Editorial Advisor
Kristin A. Moore, Ph.D.
Senior Staff Editors
Courtney Bennett, Ph.D.
Kathleen Reich, M.P.P.
Production Editors
Patricia Fewer
Roselyn Lowe-Webb
Copy Editor
Lee Engfer
Web Editor
Forrest Bryant
Design and Publishing
Barbieri & Green, Inc.
We also appreciate the contributions to this journal issue by the
Foundations research librarians.
Peter P. Budetti, M.D., J.D.

Professor and Director, Institute
for Health Services Research and
Policy Studies
Northwestern University
Felton J. Earls, M.D.
Professor of Child Psychiatry in the
Department of Psychiatry
Consolidated Psychiatry Department
Harvard University Medical School
Leon Eisenberg, M.D.
Presley Professor of Social Medicine
in the Department of Social Medicine
Deanna S. Gomby, Ph.D
Special Advisor, The David and
Lucile Packard Foundation
David E. Hayes-Bautista, Ph.D.
Professor of Medicine and Director
Center for the Study of Latino
Health and Culture
University of California, Los Angeles
Philip R. Lee, M.D.
Professor (Emeritus) of Social
Medicine, School of Medicine
Senior Scholar, Institute for Health
Policy Studies
University of California,
San Francisco
Eleanor E. Maccoby, Ph.D.
Professor (Emerita) of
Developmental Psychology
Stanford University
1. Blank, R.M., and Haskins, R., eds. The new world of welfare.
Washington, DC: Brookings Institution Press, 2001.
President and Senior Scholar
Child Trends
Paul W. Newacheck, Dr.P.H.
Professor of Health Policy
Institute for Health Policy Studies
San Francisco
Judith S. Palfrey, M.D.
Chief, Division of General Pediatrics
The Childrens Hospital
Harvard University
Nigel S. Paneth, M.D., M.P.H.
Chair, Department of Epidemiology
Professor of Pediatrics and Epidemiology
College of Human Medicine
Michigan State University
Lisa Simpson, M.B., M.P.H.
Deputy Director
Agency for Healthcare Research
and Quality
Barbara Starfield, M.D., M.P.H.
University Distinguished Professor
Professor of Health Policy and Pediatrics
Johns Hopkins University
Heather B. Weiss, Ed.D.
Director, Harvard Family
Research Project
Harvard University
Daniel Wikler, Ph.D.
Senior Staff Ethicist
World Health Organization
Rebecca A. Maynard, Ph.D.

University Trustee Professor of
Education and Social Policy and
Education
Graduate School of Education
University of Pennsylvania
The Future of Children is published twice annually by The

David and Lucile Packard Foundation, 300 Second Street,
Suite 200, Los Altos, California 94022. Fourth-class
postage paid at Los Altos, California, and at additional
mailing offices. The Future of Children is a controlled-circulation publication distributed free of charge. Opinions
expressed in The Future of Children by the editors or the
writers are their own and are not to be considered those of
The Packard Foundation. Authorization to photocopy articles for personal use is granted by The Future of Children.
Reprinting is encouraged, with the following attribution:
From The Future of Children, a publication of The David

and Lucile Packard Foundation. Correspondence should be
addressed to Richard E. Behrman, M.D., Editor-in-Chief,
The Future of Children, 300 Second Street, Suite 200, Los
Altos, California 94022. To be added to the mailing list,
write to the Circulation Department at the same address; or
to sign up for our e-newsletter, go to our Web site at
http://www.futureofchildren.org. Photographs that
appear in The Future of Children were acquired independently of articles and have no relationship to material discussed therein.
2. Center on Law and Social Policy and the Center on Budget and
Policy Priorities. The state policy documentation project.
Washington, DC: CLASP and CBPP, June, 2000. Available online
at http://www.spdp.org.
16. McGroder, S.M., Zaslow, M.J., Moore, K.A., and Le Menestrel, S.

National Evaluation of Welfare-to-Work Strategies: Impacts on
young children and their families two years after enrollment:
Findings from the Child Outcomes Study. Washington, DC: U.S.
Department of Health and Human Services and U.S. Department
of Education, 2000.
17. McLanahan, S., and Sandefur, G. Growing up with a single parent:
What hurts, what helps. Cambridge, MA: Harvard University Press,
1994.
3. Collins, A., Layzer, J., Kreader, J., et al. National study of child
care for low-income families: State and community substudy interim
report. Washington, DC: U.S. Department of Health and Human
Services, Administration for Children and Families, November
2000.
18. McLoyd, V.C. Socioeconomic disadvantage and child development. American Psychologist (1998) 53:185204.
4. Dalaker, J. Poverty in the United States: 2000. Current population

reports: Consumer income, P60-214. Washington, DC: U.S.
Census Bureau, September 2001.
19. Morris, P., Huston, A., Duncan, G., et al. How welfare and work
policies affect children: A synthesis of the research. New York:
Manpower Demonstration Research Corp., March 2001.
5. Danziger, S., Corcoran, M., Danziger, S., et al. Barriers to the

employment of welfare recipients. In Prosperity for all? The economic boom and African Americans. R. Cherry and W.M. Rodgers III,
eds. New York: Russell Sage Foundation, 2000.
20. Primus, W., Rawlings, L., Larin, K., and Porter, K. The initial
impacts of welfare reform on the incomes of single-mother families.
Washington, DC: Center on Budget and Policy Priorities, August
1999.
6. Duncan, G.J., and Brooks-Gunn, J., eds. Consequences of growing

up poor. New York: Russell Sage Foundation, 1997.
21. Shonkoff, J.P., and Phillips, D.A., eds. From neurons to neighborhoods: The science of early childhood development. Washington, DC:
National Academy Press, 2000.
7. Duncan, G.J., and Chase-Lansdale, P.L., eds. For better and for
worse: Welfare reform and the well-being of children and families.
New York: Russell Sage Foundation, January 2002.
8. Edin, K., and Lein, L. Making ends meet: How single mothers survive welfare and low-wage work. New York: Russell Sage
Foundation, 1997.
9. Fix, M., and Passel, J.S. Trends in noncitizens and citizens use of
public benefits following welfare reform: 199497. Washington, DC:
Urban Institute, March 1999.
10. Garfinkel, I., McLanahan, S., Meyer, D., and Seltzer, J., eds.
Fathers under fire: The revolution in child support enforcement. New
York: Russell Sage Foundation, 1998.
11. Gennetian, L.A., and Miller, C. Reforming welfare and work: Final
report on the Minnesota Family Investment Program. Vol. 2, Effects
on children. New York: Manpower Demonstration Research Corp.,
2000.
12. Greenberg, M., and Savner, S. The final TANF regulations: A preliminary analysis. Washington, DC: Center for Law and Social
Policy, 1999.
13. Kirby, D. Emerging answers: New research findings on programs
to reduce teen pregnancy. Washington, DC: National Campaign to
Prevent Teen Pregnancy, May 2001.
14. Knitzer, J. Promoting resilience: Helping young children and parents
affected by substance abuse, domestic violence, and depression in the
context of welfare reform. Children and welfare reform issue brief 8.
New York: National Center for Children in Poverty, 2000.
15. Loprest, P. How are families that left welfare doing? A comparison
of early and recent welfare leavers. Assessing the New Federalism
Policy Brief No. B-36. Washington, DC: Urban Institute, 2001.
22. Smith, K. Whos minding the kids? Child care arrangements.

Household economic studies, no. P70-70. Washington, DC: U.S.
Census Bureau, October 2000.
23. Sorensen, E., and Zibman, C. Child support offers some protection
against poverty. Assessing the New Federalism Policy Brief No. B10. Washington, DC: Urban Institute, 2000.
24. U.S. Department of Health and Human Services. Temporary
Assistance for Needy Families (TANF) program: Third annual
report to Congress. Washington, DC: DHHS, August 2000.
25. U.S. Department of Health and Human Services. Trends in the
well-being of Americas children and youth: 2000. Washington, DC:
U.S. Government Printing Office, 2000.
26. U.S. General Accounting Office. Education and care: Early childhood programs and services for low-income families. HEHS-00-11.
Washington, DC: GAO, 1999.
27. U.S. House of Representatives, Committee on Ways and Means.
2000 green book: Background material and data on programs within the jurisdiction of the Committee on Ways and Means.
Washington, DC: U.S. Government Printing Office, 2001.
28. Zaslow, M.J., Brooks, J.L., Moore, K.A., et al. Experimental studies of welfare reform and children. Washington, DC: Child Trends,
2001. Available online at http://www.childtrends.org.
29. Zedlewski, S. Former welfare families continue to leave the food
stamp program. Assessing the New Federalism Discussion Paper
No. 01-05. Washington, DC: Urban Institute, 2001.
30. Zimmermann, W., and Tumlin, K.C. Patchwork policies: State assistance for immigrants under welfare reform. Washington, DC:
Urban Institute, 1999.
Volume 12, Number 1
Stone/Bruce Ayres
Statement of Purpose
he primary purpose of The Future of Children

is to promote effective policies and programs
for children. The journal is intended to provide policymakers, service providers, and the
media with timely, objective information based on the
best available research regarding major issues related to
child well-being. It is designed to complement, not
duplicate, the kind of technical analyses found in academic journals and the general coverage of childrens
issues by the popular press and special interest groups.
The 1996 federal welfare reform law, the Personal
Responsibility and Work Opportunity Act, sought to
reduce the number of children growing up in poor, single-parent families by requiring mothers to move from
welfare to work and by promoting marriage. This journal issue examines how well programs implemented
under the 1996 law are accomplishing these goals, and
how they are affecting childrens development and wellbeing. Whether or not the law is having positive effects
on children is important not only for the children
involved, but for all of our citizens. Untoward effects
on child development translate into educational failure,
increased crime and violence, and reduced productivity
among adults, which are costs that impact all of society.
The articles presented here summarize the knowledge
and research about how low-income children have
been faring since passage of the welfare reform law in
1996. Overall, as of the fall of 2001, low-income children had been faring fairly well. Child poverty rates
were down, and fewer children were living in families
headed by single mothers. At the same time, many

families who had left welfare were still struggling economically, and many remaining on the rolls faced serious barriers to employment. Even among families
moving successfully from welfare to work, the effects
on children were not always positive. As our nation
enters a period of economic downturn, the future of
low-income children becomes even less certain. Reauthorization of the federal welfare reform law in 2002
offers a critical opportunity to reexamine the purpose
and goals of the law, and ensure that programs are
structured to have positive effects on children.
We welcome your comments and suggestions regarding
this issue of The Future of Children. Our intention is to
encourage informed debate about the well-being of
children under welfare reform. To this end, we invite
correspondence to the Editor-in-Chief. We would also
appreciate your comments about the approach we have
taken in presenting the focus topic and welcome your
suggestions for future topics.
Editor-in-Chief
Journal/Publications Department
300 Second Street, Suite 200
Los Altos, CA 94022
CONTENTS

Page 1
Statement of Purpose
Page 5
Children and Welfare Reform: Analysis and Recommendations

Margie K. Shields, M.P.A., and Richard E. Behrman, M.D.
An analysis of how low-income children have been faring since federal welfare reform was
enacted in 1996, and how welfare and other income support programs can be restructured
to offer the greatest promise for improving childrens chances to succeed in life.
Page 27
The 1996 Welfare Law: Key Elements and Reauthorization Issues Affecting Children
Mark H. Greenberg, J.D., Jodie Levin-Epstein, Rutledge Q. Hutson, J.D., M.P.H.,
Theodora J. Ooms, M.S.W., Rachel Schumacher, M.P.P., Vicki Turetsky, J.D.,
and David M. Engstrom, M.Sc., M.A.
A description of how the 1996 law changed the social policy landscape across a broad array
of programs and initiatives affecting children, and the issues likely to emerge before
Congress in 2002 when reauthorization of the law is debated.
Page 59
Reforms and Child Development

Aletha C. Huston, Ph.D.
A framework for understanding and assessing the impact of welfare and income support
programs on childrens healthy development through the influences these programs
have on childrens environments at home and in the community.
Page 79
Experimental Studies of Welfare Reform and Children

Martha J. Zaslow, Ph.D., Kristin A. Moore, Ph.D., Jennifer L. Brooks, Ph.D.,
Pamela A. Morris, Ph.D., Kathryn Tout, Ph.D., Zakia A. Redd, M.P.P., and Carol A. Emig, M.P.P.
A synthesis of findings from the first seven experimental evaluations of welfare programs,
distilling and presenting evidence concerning the impacts of various reform strategies
on children of different ages.
Page 97
Welfare Reform and Child Care Options for Low-Income Families

Bruce Fuller, Ph.D., Sharon L. Kagan, Ed.D., Gretchen L. Caspary,
and Christiane A. Gauthier, M.S.Ed.
An assessment of the research concerning the role child care can play in promoting
childrens development and life opportunities, and the efforts to extend access to
high-quality child care options in low-income neighborhoods.
Winter/Spring 2002
Stone/Bruce Ayres
Volume 12 Number 1
Page 121
Family Economic Resources in the Post-Reform Era

Sheila Rafferty Zedlewski, M.P.A.
A discussion of the relationship between economic resources and child well-being and how
family economic resources have changed under welfare reform, including an overview of
states innovative uses of welfare funds to expand supports for working poor families.
Page 147
Welfare Reform, Fertility, and Father Involvement

Sara S. McLanahan, Ph.D., and Marcia J. Carlson, Ph.D.
An examination of the important role that fathers play in childrens lives, how public
policies have affected childbearing and father involvement, and the need for better programs
to increase fathers financial and emotional support of children in low-income families.
Page 167
Welfare Reform and Parenting: Reasonable Expectations

P. Lindsay Chase-Lansdale, Ph.D., and Laura D. Pittman, Ph.D.
An assessment of the research concerning the impact of welfare reform on various dimensions of parenting, and the need for further research to understand how states welfare
reform programs could more effectively support positive parenting practices.
Page 187
Five Commentaries: Looking to the Future

A series of responses from experts representing various disciplines and backgrounds to the
question: How can supports and services for low-income families be improved under
welfare reform to help ensure positive outcomes for children?
Wendell E. Primus, Ph.D.
Ron Haskins, Ph.D.
Fernando A. Guerra, M.D., M.P.H.
Eloise Anderson
Barbara B. Blum
Page 208
Inside back cover
List of Acronyms
Children and Welfare Reform:

Analysis and Recommendations
hildren do best when they grow up in lowconflict families, with parents who are married to each other and who earn enough to
meet their familys needs. The evidence on
this is strong and widely accepted.1 The challenge for
our nations welfare system is to determine how best to
help children in families that do not have the support
of both parents and that do not have enough income.
About half the children born in the 1980s will spend
some time in a single-parent family before age 18.2
More than one-third will spend some of their childhood living in poverty.3
The federal welfare reform law, the Personal Responsibility and Work Opportunity Reconciliation Act of
1996, sought to reduce the number of children growing up in poor, single-parent families by promoting
marriage and requiring mothers to move from welfare
to work. This journal issue examines not only whether
the programs implemented since reform accomplished
these goals, but also whether they benefited children.
Such an examination is especially timely as Congress
begins to debate reauthorization of the federal welfare
reform law (which expires in September 2002) and as
an economic downturn changes the prospects for families striving for greater self-sufficiency.
This article reviews the main themes of the journal issue
by summarizing what the new social policy landscape
looks like for children, how low-income children are

faring in this new landscape, and how welfare programs
and related support services might be restructured to
better promote childrens well-being. The 1996 law
essentially transformed U.S. welfare programs into
employment programs. Riding the wave of a strong
economy, these new programs successfully moved many
mothers from welfare to work, and improved many
childrens lives as a resultmany, but not all. Even in
very prosperous times, some families were unable to
overcome their barriers to employment, and many families who found employment still needed additional supports to help make ends meet. As the economy
weakens, the need for supports is likely to grow.
Reauthorization of the federal welfare reform law offers
a critical opportunity to reexamine the goals of reform.
If the ultimate goal of promoting work and marriage is
not just to end families dependence on government
benefits, but also to improve disadvantaged childrens
chances for success in life, then more attention must be
paid to structuring programs for low-income families in
ways that promote positive child development and wellbeing. This is important not only for the children themselves, but for all of society, as we all pay the costs of
educational failure, increased crime and violence, and
reduced worker productivitycosts that inevitably
result when children fail to get the nurturing and supports they need to achieve their potential.
The New Social Policy Landscape

since Reform
Passage of the federal welfare reform law in 1996
brought many changes to the broad array of programs
serving low-income children and their families, as
detailed in the article by Greenberg and colleagues in
this journal issue. Prior to reform, all children in poor
families that met state eligibility criteria were entitled to
assistance under the Aid to Families with Dependent
Children (AFDC) program, although many states set
the threshold so low that only the poorest families
qualified. The federal government reimbursed states
for at least half the cost of providing this assistance,
with no cap on expenditures. AFDC receipt also
assured ready access to other benefits, such as Medicaid and food stamps. Families who participated in
AFDC-related work programs were provided with
child care assistance,4 but these programs were often
underfunded and involved only a fraction of eligible
families. Most mothers receiving AFDC payments
stayed home and cared for their children themselves.
Program structures, priorities, and funding streams all
changed dramatically with passage of the 1996 law.
The AFDC program was replaced with a block grant
to the states called Temporary Assistance for Needy
Families (TANF). Funding levels for TANF were
based on states historical expenditures under AFDC.
Families entitlement to assistance ended, and the links
between cash assistance and other benefits and services were severed. Instead, the law gave states increased
flexibility to design their own welfare programs and
support services for low-income families. At the same
time, the laws focus, as indicated by its title, was to
increase parental responsibility and work, and a major
theme was to move families off welfare and into
employment. Thus, families receiving TANF cash
assistance had to meet several important new rules,
such as more stringent work requirements, sanctions
for noncompliance, and time limits. The law also
increased funding for child care to facilitate mothers
employment, and strengthened the child support
enforcement program to help ensure that fathers
would contribute to the support of their children.
Two other key themes in the 1996 law were promoting marriage and reducing out-of-wedlock births.
Provisions of the law that addressed these goals

focused primarily on allowing more liberal eligibility
criteria for two-parent families and imposing more
stringent requirements on unmarried minor teen parents. The law also called on the U.S. Department of
Health and Human Services to establish a strategy for
preventing out-of-wedlock teen pregnancies, and provided new funding for abstinence education and
bonuses for states with the highest reductions in outof-wedlock births.
The law changed benefits and services for low-income
children in other ways as well. For example, it restructured two significant funding streams for prevention of
child maltreatment and services to reunify families split
apart by child abuse or neglect, potentially reducing
the total amount of funding for such programs. The
law also modified the definition of childhood disability, restricting childrens eligibility for assistance from
the Supplemental Security Income (SSI) program. An
estimated 100,000 children lost their eligibility under
the new definition in 1996,5 and another 70,000 adolescents lost benefits when eligibility redeterminations
were conducted between 1997 and 2000.6 Finally, the
law restricted legal immigrants eligibility for many
services and benefits, including cash assistance, food
stamps, and Medicaid. Some restrictions on legal
immigrants have since been lifted, but most remain
ineligible for food stamps, and those entering the
country after passage of the law are ineligible for nearly all federal benefits for five years.7
The changes that resulted from the 1996 law could
affect poor children in many ways, both directly and
indirectly. For the most part, however, the underlying
premise of the law was that children would benefit
from seeing their parents leave welfare and go to
work. In fact, as noted in the article by Chase-Lansdale and Pittman in this journal issue, many lawmakers expected the promotion of job preparation, work,
and marriage to improve parenting practices and child
well-being, as well as reduce dependence on government benefits. It appears, however, that the links
between reform efforts, improved parenting, and positive child outcomes are not as simple or straightforward as envisioned.
Volume 12, Number 1
How Low-Income Children Are Faring

Overall, low-income families have fared well since
reform, both economically and structurally. Poverty
rates have declined, and fewer children are being raised
in single-mother households. The primary goal of the
1996 law was to end families dependence on government benefits, and efforts to address this goal have
been very successful. Bolstered by the strong economy
of the 1990s and policies that help make work pay,
such as the earned income tax credit (EITC) and
expanded health insurance programs for low-income
children,8 many families moved off welfare and into
jobs. In August 1996, 4.4 million families were receiving cash assistance. As of September 2000, the number
of families receiving cash assistance had declined by
half, to just 2.2 million.9
The dire predictions made by some critics when the
law was passedthat the number of homeless and destitute families would skyrocket, and that there would
be a massive migration of children from welfare to foster carehave not materialized.10 Yet even in a strong
economy, many families who left welfare were struggling, and many remaining on the rolls faced serious
barriers to employment. What can be gleaned from the
research about how low-income children are faring
since welfare reform in terms of their economic wellbeing, family structure, and daily life experiences is
summarized below.
Economic Well-Being and Its Implications

Welfare reform policies clearly have helped many poor
families take advantage of economic opportunities over
the past five years and move successfully from welfare
to work. More than half the families no longer receiving cash assistance are working, and the combination
of earnings and other work supports has boosted the
incomes of most single-mother families. Poverty rates
declined from 14% in 1996 to 11% in 2000, and child
poverty rates also have fallen, from 21% to 16%.11
Recent findings from evaluations of welfare-to-work
demonstration programs point to the importance of
increasing family economic resources to improve childrens outcomes. Although these demonstrations were
implemented before reform, they included many features similar to those enacted in the 1996 law, such as
work requirements, time limits, and financial incentives

to work. Results show that, for the most part, families
participation in these programs had no widespread
impacts on children. When impacts were found, some
were positive and some were negative, but most were
weak. Even when programs resulted in more positive
impacts, the children in these families still lagged
behind national norms for positive child development.
Nevertheless, some patterns have emerged that reinforce the theory that increasing families employment
and income is good for children.
As discussed in the article by Zaslow and colleagues in
this journal issue, children in families participating in
programs that increased employment and income
tended to do better in school and have fewer behavioral problems than children in families not participating in the programs.12 Positive impacts were found
most often among school-age children. Programs with
the most positive impacts on children were those that
increased families employment and income through
earnings supplements without a mandatory work
requirement, such as in an early version of the Minnesota Family Investment Program.13
In addition to having positive impacts on children, programs that increased both employment and income
through incentives without a mandatory work requirement also had small but significant impacts on the parenting practices of mothers who participated. (See the
article by Chase-Lansdale and Pittman.) Compared
with nonparticipant mothers and those required to
work 30 hours a week, long-term recipient mothers
participating in these programs had lower levels of
depression and harsh parenting. These mothers were
also more likely to marry if single, more likely to stay
married if married, and less likely to experience domestic violence. According to Chase-Lansdale and
Pittman, the critical factor leading to these positive
impacts appears to be that the mothers could work less
than full time and still benefit from income gains compared with mothers not participating.
When programs helped families gain jobs but did not
increase their income, the evaluations of the prereform welfare-to-work demonstrations found few
impacts on children. The impacts that did occur were
mixed. Children in participant families tended to score
higher on assessments of their cognitive skills, but also

to receive more reports of behavioral problems, compared with children in families not participating in the
programs.14 When programs resulted in families making no economic progress or experiencing a setback,
the effects on children tended to be negative across all
types of outcome measures.
Thus, when families move from welfare to work without an increase in income, the impacts on children are
less likely to be positive. Unfortunately, this is likely to
be the case for many families leaving welfare since
reform. As described in the article by Zedlewski in this
journal issue, many families leaving welfare for work
enter low-paying jobs with no employer-provided benefits, and about 40% of families that have left welfare
are not working. Economic resources have declined
among many of the poorest families due to the loss of
benefits such as welfare and food stamps. Census
Bureau data show that between 1995 and 1997, total
annual resources for the poorest 10% of single-mother
families declined by $814, on average, from $5,687 to
$4,873; among the next poorest 10%, resources
declined by $319, from $11,584 to $11,265.15
The decline in economic resources among poor families is especially sobering because it occurred during a
period of sustained economic growth. Following the
attack on the World Trade Center on September 11,
2001, jobs have been disappearing rapidly in manufacturing, services, and transportationespecially lowwage jobs, the types of jobs that welfare recipients
would likely fill. During October and November 2001,
more than one million jobs were lost, and according to
the U.S. Bureau of Labor Statistics, the increase in the
jobless rate has been particularly severe for blue-collar
workers.16 Thus, until the economy rebounds, many
more poor families are likely to experience declines in
their economic resources.
Family Structure
Family structures for low-income children appear to
be changing for the better, although any link to welfare reform policies is dubious, as noted in the commentary by Haskins in this journal issue. Between
1995 and 2000, the percentage of children living with
both married parents continued to decline among
families above 200% of poverty level, but the percent-
age remained unchanged, at about 50%, among children in families below 200% of poverty.17 In addition,
the percentage of low-income children living with a
single mother declined slightly, from 34% to 33%,
whereas the percentage living with a mother and a
cohabiting male increased from 5% to 6%. Also, teen
births have continued to decline over the past decade.
Between 1990 and 1997, the teen birth rate decreased
from 60 to 52 per 1,000 women ages 15 to 19.18
To the extent that fewer children are being raised by
teen parents and single parents, child well-being is
likely to have improved. As discussed in the article by
Huston in this journal issue, both adolescent and single parenting are associated with lower educational
and occupational attainment by mothers and higher
developmental risks for children. Young and single
mothers are at high risk of poverty and generally provide less stimulating and supportive home environments than those provided by older and married
mothers. Although experimental evaluations have
found few impacts on young childrens development
linked to single parenting, other studies of older children have found that adolescents with single mothers
are at greater risk of dropping out of school than are
adolescents living with both biological parents.19
However, the data suggest that many single mothers are
not rearing their children alone. As discussed in the article by McLanahan and Carlson in this journal issue,
41% of all nonmarital births in the early 1990s occurred
to cohabiting couples.20 In what McLanahan and Carlson refer to as fragile families, many unmarried parents are working together to raise their children, either
by sharing a household or maintaining frequent contact. Such father involvement is importantboth financially and emotionallyto childrens development.
Single-parent families, especially those with no male
present, are much more likely to be poor, and the children are more likely to suffer adverse effects on their
development and well-being.21 In 1998, for example,
the poverty rate for female-headed families with children was 39%, compared with 8% for male-present
families with children.22
In addition, positive father involvement, particularly
by fathers who live with their children, has been linked
Volume 12, Number 1
to less frequent behavioral and emotional problems

among children and adolescents, including delinquency, substance use, anxiety, and depression.23 Moreover,
a recent analysis of several studies found that increased
father involvementsuch as fatherchild closeness
and authoritative parentingwas associated with significant gains in childrens academic achievement and
reductions in behavioral problems.24
Thus, evidence of increased involvement of males in
single-mother households might be a positive indicator for child well-beingwith three important
caveats. First, it is the presence of fathers, not unrelated males, that is linked to positive effects on children. For example, studies have found that childrens
school performance and behavior generally do not
improve if their mother marries someone other than
the biological father, even though family income is,
on average, substantially greater.25 Second, many lowincome families are coping with conflict and domestic violence. In one large study of welfare recipients,
for example, 28% reported having been abused by an
intimate partner in the previous year.26 To have positive impacts on children, increased male presence
must not bring increased family conflict or domestic
violence. Finally, cohabiting couples relationships are
more vulnerable than those of married couples, and if
the relationship ends, fatherchild contact is more
likely to diminish.27
Childrens Daily Lives

The movement of mothers from welfare to work has
brought significant changes to the daily lives of many
children. As increasing numbers of low-income mothers move into jobs, young children are spending more
hours in nonmaternal care, and older children are likely to be spending more time unsupervised. At the
same time, many low-income children live in families
who are receiving welfare, yet are not participating in
a welfare-to-work program. For some, the situation is
temporary, but for others, the reasons for lack of participation are likely to persist for some time. In other
cases, childrens families are not working, but they are
not receiving welfare or other benefits due to such reasons as loss of eligibility, sanctions, or time limits. Little information exists on how children in these families
are faring.
Children in Families Moving from

Welfare to Work
As discussed in Hustons article, welfare policies
designed primarily to change the economic and personal behavior of parents can impact the nature and
quality of the environments where children spend their
time, at home, at school, and in the community.
Changes in childrens environments in turn can affect
their development. Considerable evidence documents
the links between each of these factors, but the connections are complex. Policies that require low-income
mothers to work could have a number of different
impactseither positive or negativeon childrens
environments and thus their development and wellbeing. Therefore, where and how low-income children
spend their time while their mothers work warrants
greater attention.
Low-income families access to subsidies is a key factor
in their decisions about where their children spend
time during work hours. As a result of changes to child
care programs stemming from the 1996 welfare reform
law, total federal and state funding for child care for
welfare and working poor families has increased dramatically, from $2.8 billion in 1995 to $8.0 billion in
2000.28 Federal and state spending on preschools and
after-school programs has also grown.29 The movement of mothers from welfare to work, together with
this increased funding for child care, has changed the
daily environments of many low-income children.
Infants and Toddlers
Infants and toddlers need safe, high-quality child care if
they are to benefit from their mothers moving from welfare to work. Under reform, states are encouraged to
require single parents on welfare to seek work when they
have a child as young as one year old, and can require
single parents of even younger children to participate in
welfare-to-work programs. As a result, 16 states now
have work requirements for single parents with children
under age one. In 11 states, parents with children as
young as 12 weeks old are required to participate in
work activities.30 Survey data gathered in 1997 indicate
that among families with children under age three, lowincome working mothers rely mostly on the other parent or a relative for care, whereas a greater percentage of
higher-income families rely on formal center-based care,
a family child care home, or a nanny.31
The research literature is mixed about whether requiring mothers of young children to work has positive or
negative effects on the children. On the one hand,
some evidence suggests that when a familys income
increases, young children benefit, especially young children living in families below poverty level.3234 Moreover, some studies suggest that younger childrens
enrollment in full-time, high-quality child care can be
beneficial, especially for children in low-income families.35,36 On the other hand, other studies indicate that
parentchild interaction may be harmed when infants
spend extended lengths of time in child care.37
What seems to emerge from these studies is that while
increased income is likely to have positive impacts on
young children, the effects of nonmaternal child care
can be either good or bad depending on the quality of
care, the number of hours in care, and the quality of
maternal care provided in the home as the alternative.
Evaluations of prereform welfare demonstrations
found few impacts on the youngest children studied,
suggesting that infants and toddlers may not be affected by their families participation in welfare-to-work
programs as much as expected. The data are limited,
however. Further research concerning the impacts of
different types of care on young children is needed.
Preschoolers
Preschoolers, especially those from low-income families, appear to benefit from high-quality child care centersthat is, child care characterized by caregivers who
are sensitive, responsive, and talk frequently with the
children, and a setting that is well-stocked with a
breadth of learning and play materials.38 Low-income
children ages three to five who were placed in highquality centers did better on school readiness scores
than children in home-based settings.39 It has been
estimated that at least one million preschool-age children moved into new child care settings between 1996
and 1998 following changes under welfare reform.40
Studies indicate that the majority of welfare families
rely on informal arrangements when they begin to participate in work activities, yet as discussed in the article
by Fuller and colleagues in this journal issue, families
moving off welfare and into more stable, full-time
employment are more likely to choose centers and family child care homes over less formal care providers,
such as relatives or neighbors.
10
Total hours of care, stability of care, and type of care all

can affect childrens development, but the quality of
care appears to have by far the greatest influence.35 For
example, some evidence indicates that preschool children spending long hours in care may be more aggressive, assertive, and defiant than children spending less
time in care.41 Studies also show, however, that children who benefit from high-quality daily interactions
in their child care settings tend to display better cognitive and language development, school readiness, and
early school achievement, compared with children in
settings with less engaged caregivers.42 Although the
effects of child care are modest overall compared to the
stronger influence of the home environment,43 highquality care has been found to be especially effective in
improving outcomes for children growing up in poverty or facing other family-based risks.39
School-Age Children
More low-income school-age children could benefit
from participation in structured after-school activities.
Although the number of programs is increasing, many
low-income school-age children still are likely to be left
in self-care while their mothers work.
As a result of greater availability of child care subsidies
tied to welfare reform, as well as other initiatives to create more and better after-school programs for youth,
an increasing number of low-income school-age children are attending structured programs after school
while their mothers work. For example, in 2001, nearly $850 million in federal funding was provided for the
21st Century Community Learning Centers initiative,
a program designed to provide expanded learning
opportunities for school-age children in a safe, drugfree, supervised environment.44 Tens of thousands of
other after-school programs are offered by schools,
youth organizations, religious groups, and local governments throughout the country.45
Moreover, studies suggest that school-age children can
benefit from time spent in structured after-school
activities, as opposed to self-care.45,46 In one study, for
example, low-income first and third graders who spent
more time in self-care were less socially competent and
received lower academic grades in sixth grade, compared with low-income children who spent less time on
their own.47 After-school activities can provide children
Volume 12, Number 1
with opportunities to explore new skill areas, discover

their talents, and build self-esteem.48 After-school programs also can serve as a safe haven in neighborhoods
where crime rates are high and the time after school
can expose children to deviant peers, illegal activities,
and violence.49
Although data are limited, there is some evidence that
in families with mothers moving from welfare to work,
an increasing number of school-age children may be
left home alone. Prereform data indicate that working
poor families were more than twice as likely to leave
their school-age children in self-care than were nonworking poor families (15% versus 8%).50
Adolescents
Potential effects on adolescent children received virtually no attention during the debates leading to welfare
reform. The tacit assumption seemed to be that adolescents would be affected less than younger children by
what their parents might be required to do. However,
as summarized in the article by Zaslow and colleagues,
evaluations of prereform welfare-to-work demonstration programs generally found negative impacts for this
age group, even in programs that increased families
employment and income, and that had positive impacts
on younger children. (For an overview of each of the
programs evaluated, see Box 1 in the article by Zaslow
and colleagues.). For example, in the Minnesota Family Investment Program, a higher percentage of mothers
rated their adolescent childrens school performance
below average compared with ratings of adolescent children by control group mothers. In Canadas Self-Sufficiency Project, mothers enrolled in the program
reported that their children had more school behavioral
problems, and adolescents themselves reported more
smoking and drinking. And in the Florida Family Transition Program, adolescent children in the program
group were more likely than those in the control group
to have been suspended from school or to have been
rated lower on school achievement.
The reasons for these negative impacts are unclear.
Possible explanations range from an erosion in parenting quality and monitoring to an increase in adolescent
responsibilities within the household.51 However, these
findings reflect only three studies. Additional data for
this age group from other evaluations are due to be
released in early 2002, which may help to clarify how

teens are being affected by reform.
Children in Families Not Moving from
Welfare to Work
Families not participating in welfare-to-work activities
account for over two-thirds of the welfare caseload, on
average, nationwide.52 Some of these families are subject to work requirements and are not participating.
But more than half the caseload comprises two major
categories of families not subject to work requirements: families with adults who are exempt from
requirements for good cause, and families with no
adult recipients in the household, referred to as childonly cases. Because neither of these types of families is
required to participate in welfare-to-work programs,
they receive little attention from program administrators or researchers, and as a result, not much is known
about how the children in these families are faring.
Good Cause Exemptions
Among the two million adults on welfare nationwide
during fiscal year 1999 (on average, per month), 28%
were not required to participate in either work or education activities.53 Most were exempt for good cause,
such as having poor health or a disability, caring for a
household member with a disability, being advanced in
age, being the victim of domestic violence, not being
able to find suitable child care, or caring for a young
child.54 Use of exemptions varies widely among
states.30,53 For example, Arizona and Vermont exempted more than 75% of their adult recipients from work
requirements in 1999, whereas Nebraska and Oregon
exempted fewer than 20%.55
Exemptions can protect vulnerable families from sanctions and time limits, but they also may remove any
sense of urgency about providing families with the services they need to overcome their problems.56 When families are struggling with barriers that affect family
functioning as well as employability, the risk to children
can be great. In households coping with serious problems such as substance abuse and maternal depression,
some children are resilient and do fine, in general, however, children in such households are at increased risk for
harmful consequences.57 As discussed in the commentary by Guerra in this journal issue, without intervention, such families can sometimes spiral downward to
11
domestic violence, child neglect or abuse, and removal

of children from the home.58 The majority of children in
the foster care system are from families on welfare.59
Studies have found that the screening tools used by
welfare offices to identify serious family problems, and
the services provided to address these problems, are
inadequate.60 Moreover, though many families are
exempted only temporarily, others may remain so
indefinitely, awaiting services that are in short supply.61
Aside from the fact that children in these families continue to receive benefits, little is known about their
well-being.
Child-Only Cases
In 1999, about 770,000 families (29% of the total caseload) were child-only cases, down from a peak of
978,000 cases in 1996.62 Because these families receive
cash payments only for the child, not the adults in the
household, they are not subject to work requirements or
time limits. Child-only cases are created when adults are
ineligible to receive benefits, such as when children are
not living with their parents and the caretaker adults are
ineligible,63 when parents receive SSI benefits, or when
citizen children are living with noncitizen parents.64
Some researchers have suggested that child-only cases
may require intensive interventions to address families
problems, but these cases generally are not being
screened to identify such needs.65 As with children in
families with exemptions, little is known about the wellbeing of children in families receiving child-only benefits.
Children in Families Disconnected from Benefits
Welfare caseloads have declined rapidly. This would be
a cause for celebration if all the families no longer
receiving welfare were doing well. However, studies
show that poor households are not earning enough to
explain the steep drop in cash assistance and other supports. Many children still live in poverty, but without
welfare and other benefits that they likely would have
received prior to passage of the 1996 law.66 Although
little is known about the well-being of children in these
poor families, it is unlikely they are faring better as a
result of welfare reform.56
Reasons for families becoming disconnected from benefits vary, ranging from burdensome administrative pro-
12
cedures to loss of eligibility and the impact of sanctions

and time limits. For example, some families have been
discouraged from seeking benefits, or they abandoned
the effort in the face of the complex administrative procedures for enrolling and maintaining eligibility. In his
commentary, Guerra describes the difficult circumstances many poor families face, and the precariousness
of their daily lives. Time-consuming, bureaucratic procedures to obtain benefits simply add to the demands
on these families, causing many to be disconnected
from the services and supports they may be eligible to
receive and may need to cope with their problems.
Other families are disconnected from benefits because
changes to the welfare reform law made them ineligible. (See the article by Greenberg and colleagues.) In
particular, certain groups of legal immigrants are no
longer deemed qualified to receive benefits. Some
states have established state-funded substitute programs for immigrants, but these programs have not
filled the gap left from the loss of federal assistance, and
participation rates remain low. Also, many U.S. citizen
children of immigrant parents do not access benefits
they are eligible to receive. As noted in the commentary by Primus in this journal issue, food stamp participation among such children fell by 74% between 1994
and 1998, compared with a 24% decline among other
families with children.
Still other families lost benefits because they exhausted
their time limits or failed to comply with program
requirements. Data from the U.S. Department of
Health and Human Services indicate that in 1999,
about 156,000 families left welfare due to sanctions.62
Greenberg and colleagues note that families whose
cases are closed due to sanctions are likely to have low
education levels, little or no work history, and more
serious employment barriers.67
Little is known about the development and well-being
of children in poor families disconnected from benefits.
But, as Primus notes in his commentary, many families
who leave welfare without finding a job are floundering. Indeed, evidence of hardship has been found in
surveys of families who have left welfare. For example,
two waves of families leaving welfare were interviewed
in the National Survey of Americas Families, one
group in 1997 and another in 1999. About half the
Volume 12, Number 1
respondents in both groups said they have had to cut

the size of meals or skip them because they did not
have enough food in the past year. About 40% said they
were unable to pay the mortgage, rent, or utility bills
at some point in the past year.68
In sum, fewer children are living in poverty; however,
those children who are living in poverty are less likely
to be receiving the benefits they would have prior to
reform. Many poor families are disconnected not just
from welfare, but also from other supports they need
to help make ends meet. At the same time, low-income
children are more likely now to be living in a home
with a male present. If that male is the father and not
violent, a child is likely to benefit, but more information is needed to understand who the cohabiting males
are and how they interact with these children. Finally,
as more low-income mothers move into the workforce,
their children are spending more time in nonmaternal
or self-care. Time spent in structured programs can be
positive, but the quality of the program is key. Among
families still receiving welfare, most are not involved in
work activities and little is known about how their children are faring.
Restructuring the Safety Net for

Low-Income Children
Welfare programs and other supports focus primarily
on moving mothers from welfare to work. This may be
a good first step, but it does not guarantee positive
outcomes for children. Reauthorization of the federal
welfare reform law offers a critical opportunity to reexamine the purpose and goals of the law in terms of providing an effective safety net for low-income families
and their children.
Before the economic downturn took hold in the fall of
2001, a nationwide poll found that more than half of
those surveyed viewed poverty as a big problem.69 Of
those aware of the new welfare law, 61% thought it was
working well, mostly because it requires people to
work. At the same time, respondents voiced strong
support for providing more assistance to help families
trying to move off welfare into work. Large majorities
of those surveyed said they supported expanding job
training programs (94%) and public employment
(82%); improving public schools in low-income areas
(94%); increasing the minimum wage (85%) and tax

credits for low-income workers (80%); and expanding
subsidized day care (85%), medical care (83%), and
housing (75%). Especially now, as the economy slows,
having strong supports in place for the families of lowwage workers and the unemployed is crucial.
What we have learned about how low-income children
are faring in this new social policy landscape can point
to ways the safety net could be strengthened to offer
the greatest promise for improving childrens chances
to succeed in school and in life. Three key strategies
address what is most important for low-income children: helping families achieve an adequate standard of
living, helping them provide stable and supportive
homes, and helping them access quality child care and
after-school programs and activities.
Adequate Family Resources

A familys economic resources influence child well-being,
because they are necessary to meet childrens basic needs
for food and shelter. In addition, a familys economic
resources influence the quality of environments children
experience at home, in child care, at school, and in the
community. (See the articles by Huston and by ChaseLansdale and Pittman.) Children in families with adequate resources tend to be healthier and to do better in
school; they are less likely to be involved in criminal
behavior and are more likely to graduate from high
school and to earn higher incomes as adults.70
Welfare-to-work programs that increase family income
as well as employment have been shown to have the
potential to improve childrens academic progress and,
to a lesser extent, their behavior. For example, subsidy
programs that increase economic resources in lowincome working families by as little as $1,200 to
$4,000 per year have been shown to have positive
impacts on the development of preschool and elementary-school-age children.71 Important steps needed to
help low-income working families attain adequate
resources include stronger efforts to connect them to
support services and to help them find better jobs and
move up career ladders to higher wages.
Connecting Working Families to Support Services
As discussed in the article by Zedlewski, government
supports such as the EITC (which refunds a portion of
13
earnings to low-income taxpayers), food stamps, health

insurance, and subsidized health care and child care are
essential to maintaining sufficient economic resources
for families earning low wages. Studies show that participation in the EITC is generally strong, with about
85% of eligible families receiving the credit.72 But the
same is not true of other programs intended to help
make work pay.
Many families who have left welfare are not using
other, nontax supports to augment their earnings. For
example, one study found that 57% of families leaving
welfare were not receiving food stamps even though
they were eligible.73 Half the families with incomes
below 50% of poverty were not receiving food stamps.
Similar trends have been noted in Medicaid coverage
and receipt of child care subsidies. In the Urban Institute study of families leaving welfare between 1997
and 1999, only 22% reported receiving government
help in paying for child care after three months.74
As noted in Primuss commentary, when families are
not receiving welfare, they may be unaware of their eligibility for food stamps, Medicaid, child care assistance,
and other supports. Blums commentary in this journal
issue also notes that differences in income, assets, citizenship, and recertification standards create confusion
and errors in determining the eligibility of families seeking benefits across various programs. In addition, services are often provided at times and in locations that
make access difficult for low-income working families.
Without extensive outreach and advocacy to address
such problems, available benefits may never reach many
eligible families. Programs adopted following the 1996
law should be judged by their ability to restore and
increase participation among the working poor.
Recommendation
Welfare and other support programs should extend outreach efforts
to ensure that low-income families who are no longer, or who never
have been, on welfare receive the supports and services they need
until they earn sufficient income to provide an adequate standard of
living for their families.
14
Job Retention and Advancement

Getting a job and moving off welfare does not assure
that a family will have adequate economic resources.
As discussed in the article by Zedlewski, even when
working full time at a minimum wage job and receiving all available supports, families do not have adequate resources. For example, the Urban Institute
study of welfare leavers found that median monthly
wages were $1,093 in 1999.68,75 According to studies
in eight states, the average monthly earnings of former recipients during the first three months after
leaving welfare were even less, generally ranging from
about $733 to $900.71
Some studies have found that at least half of families
leaving welfare have fewer economic resources than
they would have had if they had stayed on welfare.
Many, in fact, end up returning to the rolls. In the
Urban Institute study, for example, 29% of families
leaving welfare between 1995 and 1997 returned to
welfare, as did 22% of those leaving welfare between
1997 and 1999.68
After mothers find work, they need supports to help
them keep their jobs and increase their earnings. As of
October 1999, 34 states were providing case management for at least some recipients who had found
jobs or who no longer received cash assistance.76 Generally states provide postemployment support services such as transportation aid, help in purchasing work
clothes or tools, and payment of work-related fees. It
is unclear whether these efforts have actually helped
increase steady work, however. Some evaluations have
found that they have no impact on how long families
keep jobs or how much they earn.77
Sixteen states had policies to support employer-based
education and training services to promote job
advancement. But most of these efforts served only a
small number of families. Alternatively, some states
were addressing the issue of job advancement by
increasing access to postsecondary education or training that would enable unemployed parents to find
better jobs.
The current welfare system rewards states for moving
families off the rolls, but not for ensuring that those
families gain steady employment and earn an ade-
Volume 12, Number 1
quate income. To encourage states to develop better

approaches to helping families keep their jobs and
increase their earnings, an expansion of welfares
goals is needed.
Recommendation
Welfare programs should expand their goals to include reducing
poverty and promoting family economic and child well-being. Job
retention and advancement services for low-income families should
be one of the strategies in support of this goal, and incentives
should be provided to encourage development of more effective
programs.
In particular, a greater emphasis on education and

training is warranted. Many families currently receiving welfare could benefit from further education. In
1999, more than 40% of adults on welfare had less
than a high school education.78 As noted in the commentary by Haskins, many of the mothers now working for around $7 per hour could earn more with
additional education and training.
Prior to reform, welfare programs had education
components that focused on helping recipients
enhance their basic skills, but this approach demonstrated little success in moving families into higherpaying jobs.76 Largely as a result of these findings, the
1996 law placed restrictions on education and training as a work activity, and on the length of time families could receive assistance. Under current rules, not
more than 30% of recipients being counted toward a
states work participation rate can be participating in
educational activities for no more than 12 months,
including teen parents still completing high school.
Research suggests, however, that when combined
with other activities such as job search or work itself,
education can play an important role in helping families find better jobs.76 Further, training for lowincome parents who are between jobs is a wise
investment, especially during times of high unemployment.
In addition to improving a familys future earning

potential, maternal educational progress appears to
be linked to positive child outcomes.79 As noted in
the article by Chase-Lansdale and Pittman, parents
with higher education levels tend to display more
effective parenting through a variety of behaviors
with positive implications for child outcomes.80 Preexisting differences in mothers attributes, skills, and
personalities likely influence both educational attainment and economic progress. Nevertheless, in
Zaslow and colleagues synthesis of findings from
evaluations of prereform welfare-to-work demonstrations, maternal educational progress tracked more
closely with favorable child outcomes than did economic progressthat is, increased educational attainment more consistently resulted in positive child
outcomes than did increased family employment and
income. Perhaps seeing a parent go to school provides an even better role model for children than seeing a parent go to work. Taken together, these
findings suggest that welfare programs could promote more positive outcomes for children by
enabling mothers to increase their education while
receiving temporary cash assistance within established
time limits.
Recommendation
Restrictions that discourage welfare recipients from participating in
education and training as their work activity should be liberalized.
Welfare programs should enable mothers to continue with their
schooling past high school, and to enroll in other training and education programs to advance their careers.
Stable and Supportive Homes

Although having adequate economic resources is
important, a nurturing and secure home environment, including the love and support of both biological parents, remains the most critical influence on
young childrens adjustment and well-being, even for
children who spend substantial time in child care.35
As described in the article by Huston, having the love
15
and encouragement of adult family members and

other adults who are mentally healthy, responsible,
constant, and reliable is extremely important to childrens healthy development.
Lawmakers expected programs implemented under
welfare reform to have a positive influence on parenting and the home environment, primarily as a byproduct of efforts to promote job preparation,
employment, and marriage. In addition, many states
are using welfare funds to implement initiatives targeted directly at improving parenting, such as providing home visits to new parents or requiring certain
welfare recipients to attend parenting classes.81
Although parenting behaviors are complex and difficult to change, welfare programs can help strengthen
family functioning and the home environment in various ways depending on a familys circumstances. Different approaches are required depending on whether
a family is moving from welfare to work, or is struggling with barriers to employment. In many situations,
greater involvement of fathers can be a key strategy for
strengthening families and promoting childrens wellbeing both financially and emotionally.
Flexible Work Requirements
In families with a mother moving from welfare to
work, parenting practices and the home environment
could be affected by a change in family resources,
family relationships, parents time at home, and parents personal sense of well-being. According to
Chase-Lansdale and Pittman, the limited available literature on this topic indicates that, compared with
unemployed single mothers, single mothers who
choose to work may provide a more cohesive, stimulating, and organized family environment, and may
do more to promote the value of education.82 But
other studies have found that mothers who were
required to participate in welfare-to-work programs
experienced higher levels of depression and stress
than those who were not required to participate.83 In
evaluations of prereform welfare-to-work demonstrations, few impacts on parenting were identified. One
program that did have positive impacts on parenting,
however, was a version of the Minnesota Family
Investment Program that increased families employment and income through earnings supplements
without a mandatory work requirement.84 Mothers in
16
this program generally decided to work part time or

reduce some work hours.
These findings, though limited, suggest that welfare
programs could strengthen families and promote positive outcomes for children by allowing greater flexibility concerning part-time work for single mothers,
and by stopping the time limit clock on welfare benefits while families are working at least part time, but
not earning enough to support themselves.
Recommendation
Welfare programs should provide sufficient supports to allow single
mothers greater flexibility concerning part-time work schedules,
and should stop the clock on time-limited benefits while parents are
working either full or part time.
Addressing Barriers
Many low-income families live on the edge, both economically and functionally. According to data from
the National Survey of Americas Families in 1997
and in 1999, about three-quarters of adults on welfare had at least one potential barrier to employment,
including very poor mental or physical health, limited
education, minimal work experience, or family
responsibilities that limit their ability to work, such as
caring for an infant or a disabled child.78
When families are struggling with serious barriers
such as substance abuse, maternal depression, and
domestic violencethe problems not only present
challenges in getting and keeping jobs, they also
detract from parents ability to provide stable and
supportive homes for their children. Estimates of the
percentage of welfare recipients with substance abuse
problems range from 16% to 37%.85 About 22% to
28% are estimated to be suffering from very poor
mental health,78 and 10% to 31% to be coping with
domestic violence.86
Some of these families are exempt from work require-
Volume 12, Number 1
ments, whereas others are not, since states policies

on exemptions vary widely. But simply providing welfare and exempting families from work and time limit
requirements will not enable most families to overcome their problems, enter the workforce, and provide stable and supportive homes for their children.
The underlying conditions contributing to a familys
precarious situation must be addressed.
Often these underlying conditions are rooted in the
mothers own childhoods. Many low-income mothers have grown up in families that did not provide
good models of positive family functioning. They
may have experienced sexual and physical abuse.
Among participants in a New Jersey welfare-to-work
program, for example, 20% report having been sexually molested as children.86 Other studies indicate that
the proportion of teen mothers with histories of sexual and physical abuse may be as high as 68%.87 Thus,
it should not be surprising that a large proportion of
mothers on welfare have scars from the past that contribute to their difficulties in coping with their complex lives, and may lead them to substance abuse or
involvement in high-conflict relationships.
In 1997, Congress enacted the Welfare-to-Work program to support state and local efforts to help the
most disadvantaged welfare recipients (as well as noncustodial parents) address specific problems affecting
their employment prospects and move into the labor
force.78 Some states have begun implementing programs to address families barriers to employment,
but most serve only a small percentage of those needing help and little information is available on program outcomes and effectiveness. Most caseworkers
have limited training or skills to identify families with
barriers.88
To deal more effectively with families struggling with
serious difficulties, a restructuring of services may be
needed. As Anderson suggests in her commentary in
this journal issue, families coping with mental illness
or learning disabilities, for example, should be treated and monitored by the mental health system or the
developmental disabilities system, not by welfare or
employment programs. This does not mean that such
families should not be encouraged to move from wel-
fare to work, only that specialized systems are better

equipped to provide the supports that can enable
these families to find appropriate work situations.
Such systems are also better able to help these families connect with long-term parenting supports that
may continue to be needed even after their economic situation improves.
Recommendation
Welfare programs should develop better strategies for identifying
and addressing the problems of families with serious barriers to
employment and healthy family functioning. Options that include
integrating services with mental health systems and developmental
disabilities systems should be explored.
Father Involvement
A large number of low-income children grow up in
homes with no father. In 2000, nearly half of all families living in poverty were headed by a female with no
husband present.89 Although many unmarried parents
work together to raise their children by cohabiting or
maintaining frequent contact, father involvement for
most low-income families in this situation is not stable
and tends to decline over time. Yet research in the last
decade has pointed to the range of contributions
fathers can make in their childrens lives, both financially and emotionally.90 Welfare programs and other
support services could be restructured in various ways
to promote greater inclusion and involvement of lowincome fathers in their childrens lives.
For example, as discussed in the article by McLanahan
and Carlson, most unmarried couples are closely connected to each other and invested in their new family
when a baby is born. Thus, a window of opportunity
exists around the time of birth to strengthen the connections between fathers and their children. In a study
of approximately 3,700 children born to unmarried
parents across the country in 2000, 99% of the new
fathers who were interviewed expressed a desire to be
involved in raising their children, and 93% of mothers
17
said they wanted the father to be involved.91 Targeting

fatherhood programs to unmarried fathers at the time
of a childs birth appears to offer the greatest promise
of successful involvement.
In addition, the chances for positive father involvement
would improve if a broader range of TANF services were
targeted to two-parent families and noncustodial parents, as noted in the commentary by Primus. For example, programs that provide employment opportunities
for fathers help increase their ability to pay child support
and also make them more appealing as marriage partners.92 More successful programs that teach men to be
good fathers and reduce family violence are also needed.
Finally, reducing the marriage penalties in programs
aimed at helping low-income familiesincluding welfare, food stamps, and Medicaid, as well as state and
federal EITCshas been noted as an important societal signal and support for marriage. The way these programs are currently structured, benefits generally
decrease as household income increases. Because marriage results in one combined household income, a
couple could lose thousands of dollars a year in benefits compared with what they would be eligible to
receive if they each remained single.93
Unfortunately, most poor children eligible for child

support do not receive it. Child support is collected for
only 44% of families who have left welfare and 25% of
families on welfare.96 In large part, low collection rates
are due to a lack of support orders. Unless an order is
established, it cannot be enforced. Among families
receiving welfare in 2000, fewer than half had a support order established.97 Increased efforts to establish
support orders among low-income families is an
important first step to increasing collections.
Programs serving low-income families should remove any marriage

penalties, and should provide services and benefits for fathers
without reducing services and benefits for mothers.
Low collection rates also stem from the fact that many
fathers of poor children are themselves poor and have
limited ability to pay.98 Several demonstration projects
were undertaken in the 1980s and 1990s to improve
fathers employment and earnings, but without much
success, as discussed in the article by McLanahan and
Carlson.99 More recently, nearly $2 billion in grants
were awarded in 1998 and 1999 through the Welfareto-Work program to help move noncustodial parents
(and other hard-to-serve populations) into unsubsidized jobs.100 As mentioned earlier, little information is
available on the effectiveness of these programs, but an
evaluation is under way.101
Improving the child support system is also important,

not only for increasing collections, but also for promoting more positive father involvement. Child support can
significantly augment the wages of single-parent families.
One study found that child support provided 14% of the
income for families with incomes below 150% of poverty who were not on welfare, and 35% of the income
among all families receiving some child support.94 Other
research found that $1,000 in child support was associ-
Beyond having few resources, fathers of children in

welfare families generally have little incentive to pay
child support because, for the most part, it goes to
repay welfare expenditures and not to increase family
income. Only a small amount, if any, is passed
through to the family.102 However, a demonstration
project in Wisconsin allowed all support to be passed
through to TANF families, and the project appears to
have had positive results. Noncustodial fathers participating in the program were more likely to establish
paternity and pay support to their families, and family
conflict around child support was reduced.103
Recommendation
18
ated with higher grades and fewer school problems

among children, and that child support income was
associated with a larger gain in childrens well-being than
an identical amount of income from other sources.95
The positive effects of child support income are likely
due to the fact that fathers who pay child support also
tend to be more involved with their children.
Volume 12, Number 1
Recommendation
States should increase noncustodial parents ability to pay child
support by replicating successful programs that help them, along
with custodial parents, find employment and increase earnings.
Also, to increase the incentives for paying child support, policies
should be adopted to ensure that children benefit financially when
payments are made, even if their families receive welfare.
High-Quality Child Care and After-School Options

Welfare reform brought with it a clear mandate that
low-income single mothers should work. An equally
clear mandate is needed that while their mothers are at
work, low-income children should have access to highquality child care and after-school options in their
neighborhoods. Providing quality environments for
children not only enables mothers to work, but it is also
an opportunity to enhance childrens development and
well-being. As discussed earlier, studies show that poor
children are especially likely to benefit from positive
experiences in child care and after-school programs.
The increases in funding for child care since welfare
reform have helped many low-income families afford
higher-quality options for their children. However, as
discussed in the article by Fuller and colleagues, highquality child care and after-school options for lowincome families are often still constrained, both in
terms of cost and supply.
Despite funding increases, state studies show child care
subsidies are used by less than one-quarter of all eligible
families.104 Data compiled by the U.S. Department of
Health and Human Services indicate that nationwide,
of the 14.7 million children eligible to receive subsidies
in 1999, only 1.8, or 12%, received them.105 This percentage does not include families participating in Head
Start or other state-funded preschool or child care programs, but subsidy utilization rates, especially among
working poor families, are still low. Although eligible
for assistance, working poor families tend to have less
access to subsidies than families on or just leaving welfare, and the availability of high-quality programs in
low-income communities is uneven. Fuller and col-
leagues identify several barriers to use of subsidies and

high-quality care, including state eligibility criteria,
copayment policies, reimbursement rates, burdensome
administrative processes, and lack of high-quality centers in many low-income neighborhoods.
Though not eligible for child care subsidies, teens also
benefit from participation in structured activities after
school. Several studies have shown that participation in
targeted prevention programs can reduce high-risk
behaviors among teenagers.106 For example, in four consecutive annual evaluations, participants in the Teen
Outreach Program, sponsored by the Association of
Junior Leagues International, were shown to be less likely than their nonparticipant peers to have experienced
either pregnancy or school failure.107 An evaluation of
another youth development program, the Quantum
Opportunities Program, also found significant positive
effects on economically disadvantaged high school
youths.108 A five-year study of this program at four sites
showed that participants had better high school graduation rates, higher enrollment in postsecondary education, lower teen pregnancy rates, and more community
involvement compared with those who did not participate. A recent evaluation of the Childrens Aid Society
model found similar results.109 In light of the initial negative impacts found on teens whose mothers participate
in welfare-to-work programs, additional investment in
programs for low-income youth may be warranted.
As low-income mothers move into the workforce and
their children spend more time unsupervised or in child
care and after-school programs, those experiences have
an increasingly important influence on childrens development and well-being. Low-income children of all
ages could benefit from increased access to high-quality child care and after-school programs and activities.
Recommendation
Further efforts are needed at the federal, state, and local levels to
expand child care subsidy programs for working poor families, and
to strengthen the supply of high-quality child care and after-school
options in low-income neighborhoods. Age-appropriate services are
needed for all age groups, from infants to teens.
19
Conclusion
What have we learned from welfare reform about how
best to help children in families that do not have the
support of both parents and adequate income? The evidence is clear that moving mothers into jobs is a good
first step, but further efforts are needed to promote
better outcomes for children.
First, working poor families need to stay connected to
supports, including job advancement and continuing
education and training, to ensure that children have
adequate resources. An adequate income is important
to meeting childrens basic needs and enabling families to provide healthier learning environments for
children both inside and outside the home. Welfare
and other income support programs need to continue
to provide services and supports to families after they
find jobs to help them increase their earnings and
escape poverty. Welfare programs should also allow
single mothers to work more flexible, part-time hours
and pursue more education and training. Not only
would this enhance a familys earning potential, but
maternal educational advancement also appears to foster positive impacts on childrens school performance
and behavior. Such a strategy makes sense in the current economy, as finding and keeping low-wage jobs
becomes increasingly difficult.
Second, barriers to healthy family functioning and
greater father involvement need to be identified and
addressed to ensure that children have stable, supportive homes. Many families are neither working nor participating in welfare-to-work programs for a variety of
reasons, and more families are likely to join their numbers as the economy slows. Welfare programs need to
develop better strategies for identifying families with
serious barriers to employment and healthy family
functioning, and to explore new ways of helping these
families, such as by integrating services with mental
health and developmental disabilities systems. Welfare
and other income support programs also need to
encourage greater father involvement by engaging
fathers soon after a childs birth, and by adopting more
even-handed policies providing access to services and
benefits. Fathers, as well as mothers, need programs to
help them find jobs and advance in their careers, and to
20
help them address their barriers to employment and

positive family functioning.
And third, children and teens need high-quality child
care and after-school activities in their neighborhoods
to ensure that the time they spend while their mothers
are at work plays a positive role in their lives. For lowincome children especially, such programs can provide
safe havens and a range of opportunities to enhance
their development. Welfare programs need to work
with child care agencies to expand access to child care
subsidies for working poor families, and to expand the
supply of high-quality options for children of all ages in
low-income communities.
For the nations welfare programs to improve the lives
of low-income children, child well-being needs to be a
top priority. This analysis has highlighted some of the
ways that current welfare programs could be restructured to help promote healthier child development
among low-income families. As the economy weakens
and more families are forced to go on welfare, more
funds will be needed for cash assistance and less will be
available for some of the services described here as beneficial to children, such as programs to help families
address barriers to positive family functioning as well as
employment; programs to promote father involvement; and programs to improve the quality of child
care, enrichment activities, and programs for teens.
The present level of funding for welfare and other support programs needs to be maintained, at a minimumand increased, if possibleto ensure that the
progress made since reform in restructuring programs
to help families move from welfare to work is not lost,
and that a strong safety net for poor families is in place.
The goal of welfare programs should be to reduce
poverty among families with children. Ending dependence on benefits should be the result of achieving this
goal, not be the goal itself. Otherwise, programs will
succeed only in moving families off the rolls, not in
helping them escape poverty or improve their childrensand societyschance for a brighter future.
Volume 12, Number 1
Special thanks to Kristin Moore, Mark Greenberg,

and Sheila Zedlewski for their insightful comments
and support throughout the development of this
analysis.
ENDNOTES
1. See, for example, previous issues of The Future of Children on
children and divorce (Spring 1994) 4(1); children and poverty
(Summer/Fall 1997) 7(2); and domestic violence and children
(Winter 1999) 9(3). See also various articles describing the
decade in review in Journal of Marriage and the Family (November 2000) 62(4).
2. Bumpass, L., and Raley, K. Redefining single-parent families:
Cohabitation and changing family reality. Demography (1995)
32(1):97109.
3. Duncan, G.J., and Rodgers, W.L. Longitudinal aspects of childhood poverty. Journal of Marriage and the Family (November
1988) 50:100721.
4. Under the Job Opportunities and Basic Skills Training (JOBS)
program, some families receiving welfare were required to participate in work activities, and those who participated were eligible
for child care assistance. See the article by Fuller and colleagues in
this journal issue.
5. U.S. General Accounting Office. Supplemental Security Income:
Progress made in implementing welfare reform changes; more action
needed. GAO-HEHS-99-103. Washington, DC: GAO, 1999.
6. U.S. Social Security Administration. 2001 annual report of the
Supplemental Security Income program. Baltimore, MD: SSA,

Office of the Chief Actuary, May 2001, table V.D3.
7. See the article by Greenberg and colleagues in this journal issue.
8. The expansion of government programs that provide health insurance coverage, such as the State Childrens Health Insurance Program (SCHIP), is important for low-income families, who often
work in low-paying jobs with no employer-provided benefits, and
who otherwise might stay on welfare to maintain their health coverage through Medicaid.
9. U.S. Department of Health and Human Services. ACF data and
statistics: U.S. welfare caseloads information. Washington, DC:
DHHS, August 2001. Available online at
http://www.acf.dhhs.gov/news/stats.
10. Geen, R., Fender, L., Leos-Urbel, J., and Markowitz, T. Welfare
reforms effect on child welfare caseloads. Assessing the New Federalism Discussion Paper No. 01-04. Washington, DC: Urban Institute, February 2001.
21
12. See also Morris, P., Huston, A., Duncan, G., et al. How welfare
and work policies affect children: A synthesis of research. New York:
13. Gennetian, L.A., and Miller, C. Reforming welfare and work:
Final report on the Minnesota Family Investment Program. Vol. 2,
Effects on children. New York: Manpower Demonstration
Research Corp., 2000.
14. The Teenage Parent Demonstration program in Chicago, the
JOBS program in all sites studied, and the Minnesota Family
Investment Program for recent applicants all resulted in greater
employment for participant families, but not increased income.
(See the Appendix following the article by Zaslow and colleagues
in this journal issue.)
1999, p. 16.
16. U.S. Bureau of Labor Statistics. The employment situation:
November 2001. Press release. Washington, DC: U.S. Department of Labor, December 7, 2001; and U.S. Bureau of Labor
Statistics. Statement of Lois Orr, Acting Commissioner, before
the Joint Economic Committee, U.S. Congress. November 2,
2001.
17. Dupree, A., and Primus, W. Declining share of children lived with
single mothers in the late 1990s. Washington, DC: Center on Budget and Policy Priorities, June 2001.
18. See Figure 3 in the article by McLanahan and Carlson in this
journal issue.
29. See fact sheets available on the U.S. Department of Health and
Human Services Head Start Bureau Web site at http://
www2.acf.dhhs.gov/programs/hsb/; and on the U.S. Department of Education Web site at http://www.ed.gov/21stcclc.
Also see Cauthen, N.K, Knitzer, J., and Ripple, C.H. Map and
track: State initiatives for young children and families, 2000 edition. New York: National Center for Children in Poverty, Columbia University, 2001, p. 5.
30. State Policy Documentation Project. Work requirements: Exemptions, as of October 1999. Washington, DC: Center for Law and
Social Policy and Center for Budget and Policy Priorities, July
2000. Available online at http://www.spdp.org.
31. Ehrle, J., Adams, G., and Tout, K. Whos caring for our youngest
children? Child care patterns of infants and toddlers. Washington,
DC: Urban Institute, January 2001.
32. Duncan, G.J., and Brooks-Gunn, J. Income effects across the life
span: Integration and interpretation. In Consequences of growing
up poor. G.J. Duncan and J. Brooks-Gunn, eds. New York: Russell Sage Foundation, 1997, pp. 596610.
33. Baydar, N., and Brooks-Gunn, J. Effects of maternal employment
and child-care arrangements on preschoolers cognitive and
behavioral outcomes: Evidence from the children of the National
Longitudinal Survey of Youth. Developmental Psychology (1991)
27:93245.
19. McLanahan, S., and Sandefur, G. Growing up with a single parent: What hurts, what helps. Cambridge, MA: Harvard University
Press, 1994.
34. See Bradley, R.H., Caldwell, B.M., and Rock, S.L. Home environment and school performance: A ten-year follow-up and
examination of three models of environmental action. Child
Development (1988) 59:85267.
20. Bumpass, L., and Lu, H.-H. Trends in cohabitation and implications for childrens family contexts. Population Studies (2000)
54:2941.
35. Phillips, D., and Adams, G. Child care and our youngest children. The Future of Children (Spring/Summer 2001)
11(1):3551.

36. NICHD Early Child Care Research Network. Motherchild

interaction and cognitive outcomes associated with early child
care: Results of the NICHD Study of Early Child Care. Paper
represented at the Biennial Meeting of the Society for Research
in Child Development. Washington, DC. April 36, 1997.

2000 green book: Background material and data on programs
within the jurisdiction of the Committee on Ways and Means.
Washington, DC: U.S. Government Printing Office, 2001,
p. 1289, table H-6.
23. See, for example, Amato, P.R., and Rivera, F. Paternal involvement and childrens behavior problems. Journal of Marriage and
the Family (1999) 61:37284; and Harris, K.M., Furstenberg Jr.,
F.F., and Marmer, J.K. Paternal involvement with adolescents in
intact families: The influence of fathers over the life course.
Demography (1998) 35(2):20116.
24. Amato, P.R., and Gilbreth, J.G. Nonresident fathers and childrens well-being: A meta-analysis. Journal of Marriage and the
Family (1999) 61:55773.
25. See the article by Huston in this journal issue.
26. Miller, C., Knox, V., Gennetian, L.A., et al. Reforming welfare
and work: Final report on the Minnesota Family Investment Program. Vol. 1, Effects on adults. New York: Manpower Demonstration Research Corp., 2000.
22
and Means, pp. 62527; and data charts available on the U.S.
Department of Health and Human Services Child Care Bureau
Web site at http://www.acf.dhhs.gov/programs/ccb/research.
37. NICHD Early Child Care Research Network. The effects of

infant child care on infant-mother attachment security: Results of
the NICHD Study of Early Child Care. Child Development
(1997) 68(5):86079.
38. For more details, see the article by Fuller and colleagues in this
journal issue.
39. Zaslow, M., McGroder, S., Moore, K., and LeMenestrel, S.
Behavior problems and cognitive school readiness among children in families with a history of welfare receipt. Paper presented
at the Biennial Meeting of the Society for Research in Child
Development. Albuquerque, NM. April 1518, 1999. See also
Zaslow, M., Oldham, E., Moore, K., and Magenheim, E. Welfare
families use of early childhood care and education programs, and
the implications for their childrens development. Early Childhood Research Quarterly (1998) 13:53563.
27. For more details, see the article by McLanahan and Carlson in
this journal issue.
40. See Growing Up in Poverty Project. Remember the children:

Mothers balance work and child care under welfare reform. Berkeley: University of California, Policy Analysis for California Education, 2000, technical supplement.
28. See note 22, U.S. House of Representatives, Committee on Ways
41. NICHD Early Child Care Research Network. Quality of child
Volume 12, Number 1
care and child outcomes. Paper presented at the Biennial Meeting

of the Society for Research in Child Development. Minneapolis,
MN. April 19, 2001.
43. NICHD Early Child Care Research Network. The relation of
child care to cognitive and language development. Child Development (2000) 71(4):96080.
44. See the U.S. Department of Education Web site at
http://www.ed.gov/21stcclc.
45. See Vandell, D.L., and Shumow, L. After-school child care programs. The Future of Children (Fall 1999) 9(2):6480.
46. Vandell, D.L., and Posner, J. Conceptualization and measurement of childrens after-school environments. In Assessment of the
environment across the lifespan. S.L. Friedman and T.D. Wachs,
eds. Washington, DC: American Psychological Association Press,
1999, pp. 16797.
47. Pettit, B.S., Laird, R.D., Bates, J.E., and Dodge, K.A. Patterns of
after-school care in middle childhood: Risk factors and developmental outcomes. Merrill-Palmer Quarterly (1997) 43:51538.
48. Larner, M.B., Zippiroli, L., and Behrman, R.E. When school is
out: Analysis and recommendations. The Future of Children (Fall
1999) 9(2):420.
49. Posner, J.K., and Vandell, D.L. Low-income childrens afterschool care: Are there beneficial effects of after-school programs?
Child Development (1994) 65:44056.
Census Bureau, October 2000, table 11.
51. Brooks, J., Hair, E., and Zaslow, M. Welfare reforms impact on
adolescents: Early warning signs. Research brief. Washington, DC:
Child Trends, July 2001.
and Means, p. 424. See also note 9, U.S. Department of Health
and Human Services.
53. U.S. Department of Health and Human Services. Characteristics
and financial circumstances of TANF recipients, fiscal year 1999.
Washington, DC: DHHS, 2000, table 19.
54. Some states allow good cause exemptions for other reasons, such
as not having transportation, living in a remote area, or caring for
a child when not the parent. See note 30, State Policy Documentation Project.
55. States with limited or no exemptions generally allow a broader
array of activities to qualify as meeting work requirements, such
as attending parenting classes or participating in a substance
abuse program, even though these activities are not defined as
work activities in law and do not count toward a states mandated
participation rate. See the Personal Responsibility and Work
Opportunity Reconciliation Act of 1996. Public Law 104-193,
section 103, 110 Stat. 2133, August 22, 1996.
56. Pavetti, L., and Bloom, D. State sanctions and time limits. In The
new world of welfare. R.M. Blank and R. Haskins, eds. Washington, DC: Brookings Institution Press, 2001, pp. 24564.
57. Knitzer, J. Promoting resilience: Helping young children and parents affected by substance abuse, domestic violence, and depression in
the context of welfare reform. Children and welfare reform issue
brief 8. New York: National Center for Children in Poverty, 2000.
58. The 1996 welfare reform law included specific provisions with
respect to families coping with domestic violence. The Family
Violence Option, if adopted, allows states to identify and exempt
domestic violence victims from program requirements. In turn,
states are not penalized for failure to meet work or time limit
requirements if due to domestic violence exemptions, but the
exemptions must be temporary and services must be provided.
U.S. Department of Health and Human Services. Work not welfare: Clinton administration issues new proposed welfare regulations. Press release. Washington, DC: DHHS, November 17,
1997.
59. However, it should be noted that most families receiving welfare
are not involved with the child welfare system, and that a majority of children in the foster care system are there due to neglect,
not abuse. See Chapin Hall Center for Children at the University
of Chicago. Dynamics of childrens movement among the AFDC,
Medicaid and foster care programs prior to welfare reform:
19951996. Washington, DC: U.S. Department of Health and
Human Services, March 2000. See also Sedlak, A.J., and Broadhurst, M.L.A. Executive summary of the third national incidence
study of child abuse and neglect. Washington, DC: Department of
Health and Human Services, National Center on Child Abuse
and Neglect, September 1996; and U.S. Department of Health
and Human Services, Administration on Children, Youth and
Families. Child maltreatment 1999. Washington, DC: U.S. Government Printing Office, 2001.
60. U.S General Accounting Office. Welfare reform: Moving hard-toemploy recipients into the workforce. GAO-01-368. Washington,
DC: GAO, March 2001. See also Johnson, A., and Meckstroth,
A. Ancillary services to support welfare to work: Domestic violence.
Washington, DC: U.S. Department of Health and Human Services, June 1998.
61. As the number of families with exempt status grows, however,
states may be at risk for failing to meet mandated work participation rates or for exceeding the 20% limit on caseloads that can be
exempted from time limits based on hardship.
62. See note 53, U.S. Department of Health and Human Services,
August 27, 2000, update.
63. Even if eligible, caretaker relatives can opt to receive child-only
benefits to avoid being subject to work and time limit requirements.
64. In addition, families that receive a partial sanction for failure to
comply with a program requirement also may receive only the
childs portion of benefits, but such cases can be designated child
only (and thereby excluded from the calculation of a states mandated participation rate) for no more than three months, according to federal rules. See the final regulations for TANF,
Child-only cases. Federal Register (April 12, 1999)
64(69):17739.
65. Blum, B.B. Welfare research perspectives: Past, present, and future,
2001 edition. New York: National Center for Children in Poverty,
August 2001.
66. See note 15, Primus, et al.
67. Goldberg, H., and Schott, E. A compliance-oriented approach to
sanctions in state and county TANF programs. Washington, DC:
Center on Budget and Policy Priorities, October 2000.
69. Based on a nationwide telephone survey conducted as part of an
ongoing project of National Public Radio, the Henry J. Kaiser
23
Family Foundation, and Harvard Universitys Kennedy School of

Government. The survey was conducted in English and Spanish
between January 4 and February 7, 2001, among a random representative sample of 1,952 adults (age 18 and older). Survey
results are available online at http://www.npr.org/programs/
specials/poll/poverty/summary.html.
70. Mayer, S. Trends in the economic well-being and life chances of
Americas children. In Consequences of growing up poor. G. Duncan and J. Brooks-Gunn, eds. New York: Russell Sage Foundation, 1997, pp. 4969.
71. See the article by Zedlewski in this journal issue.
72. Phillips, K. Who knows about the earned income tax credit? Policy
brief no. B-27. Washington, DC: Urban Institute, January 2001.
74. Loprest, P. How are families who left welfare doing across time?
A comparison of two cohorts of welfare leavers. In Economic policy review. New York: The New York Federal Reserve, 2001.
pp. 919.
75. Loprest, P. Families who left welfare: Who are they and how are
they doing? Assessing the New Federalism Discussion Paper No.
99-02. Washington, DC: Urban Institute, 1999.
76. Strawn, J., Greenberg, M., and Savner, S. Improving employment outcomes under TANF. In The new world of welfare. R.M.
Blank and R. Haskins, eds. Washington, DC: Brookings Institution Press, 2001, pp. 22344.
77. Rangaran, A., and Novak, T. The struggle to sustain employment:
The effectiveness of the postemployment services demonstration.
MPR reference 8194-620. Princeton, NJ: Mathematica Policy
Research, 2000; and Strawn, J., and Martinson, K. Steady work
and better jobs: How to help low-income parents sustain employment and advance in the workforce. New York: Manpower
Demonstration Research Corp., 2000.
78. Zedlewski, S., and Loprest, P. Will TANF work for the most disadvantaged families? In The new world of welfare. R.M. Blank and
R. Haskins, eds. Washington, DC: Brookings Institution Press,
2001. pp. 31134.
79. See the Appendix (subgroup MFIP/LT/I) following the article
by Zaslow and colleagues in this journal issue.
80. Kelley, M.L., Power, T.G., and Wimbush, D.D. Determinants of
disciplinary practices in low-income black mothers. Child Development (1992) 63:57382. See also Collins, A., Layzer, J.,
Kreader, J., et al. National study of child care for low-income families: State and community substudy interim report. Washington,
DC: U.S. Department of Health and Human Services, Administration for Children and Families, November 2000.
24
Project at 36 months: Effects on children of a program that

increased parental employment and income. Ottawa, Ontario:
Social Research and Demonstration Corp., 2000.
84. Another program found to have positive impacts on parenting
was the New Hope program, perhaps due in part to the intensive
case management provided. For further discussion of both the
Minnesota Family Investment Program and the New Hope Project and their effects on parenting, see the article by Chase-Lansdale and Pittman in this journal issue.
85. Center on Addiction and Substance Abuse and American Public
Human Services Association. Building bridges: States respond to
substance abuse and welfare reform. New York: Columbia University, Center on Addiction and Substance Abuse, 1999.
86. See Danziger, S., Corcoran, M., Danziger, S., et al. Barriers to
the employment of welfare recipients. In Prosperity for all? The
economic boom and African Americans. R. Cherry and W.M.
Rodgers III, eds. New York: Russell Sage Foundation, 2000,
pp. 24578.
87. Boyer, D., and Fine, D. Sexual abuse as a factor in adolescent
pregnancy and child maltreatment. Family Planning Perspectives
(January/February 1992) 24(1):411, 19.
88. The capacity to screen families for barriers is needed not just to
assess families after they have been unable to move quickly into
employment, but as soon as families apply for assistance. In their
synthesis of welfare-to-work evaluations, Zaslow and colleagues
foundunexpectedlythat children in families that were new to
welfare tended to experience more negative outcomes compared
with children in families that had been receiving welfare for more
than two years. They hypothesized that perhaps these negative
outcomes occurred because recent applicant families are experiencing acute rather than chronic stress. For more details, see the
article by Zaslow and colleagues in this journal issue.
89. See note 11, Dalaker, table A-3.
90. See, for example, Federal Interagency Forum on Child and Family Statistics. Nurturing fatherhood: Improving data and research
on male fertility, family formation and fatherhood. Washington,
DC: FIFCFS, 1998.
91. See the article by McLanahan and Carlson in this journal issue
for further details of these findings from the Fragile Families and
Child Wellbeing Study.
92. Offner, P. Reducing non-marital births. Welfare Reform &
Beyond policy brief no. 5. Washington, DC: The Brookings
Institution, August 2001.
93. See Horn, W.F., and Sawhill, I.V. Fathers, marriage, and welfare
reform. In The new world of welfare. R.M. Blank and R. Haskins,
eds. Washington, DC: Brookings Institution Press, 2001,
pp. 42141.
81. For a summary of states use of TANF funds for initiatives to

improve parenting, see the Welfare Information Network Web
site at http://www.welfareinfo.org/FamilySupportServices.htm.
against poverty. Assessing the New Federalism Policy Brief No.
B-10. Washington, DC: Urban Institute, 2000.
82. Ricciuti, H.N. Single parenthood and school readiness in white,

black, and Hispanic 6- and 7-year-olds. Journal of Family Psychology (1999) 13:45065. See also Vandell, D.L., and Ramanan,
J. Effects of early and recent maternal employment on children
from low-income families. Child Development (1992) 63:93849.
95. McLanahan, S., Seltzer, J., Hanson, T., and Thomson, E. Child
support enforcement and child well-being: Greater security or
greater conflict? In Child support and child well-being.
I. Garfinkel, S.S. McLanahan, and P.K. Robins, eds. Washington,
DC: Urban Institute Press, 1994.
83. Quint, J.C., Bos, J.M., and Polit, D.F. New Chance: Final report
on a comprehensive program for young mothers in poverty and their
children. New York: Manpower Demonstration Research Corp.,
1997; and Morris, P., and Michalopoulos, C. The Self-Sufficiency
96. U.S. Department of Health and Human Services, Office of Child

Support Enforcement. Child support enforcement: FY 2000 preliminary data preview report. Washington, DC: DHHS, July
2001.
Volume 12, Number 1
97. See note 96, DHHS, figure 3.4.
104. See note 80, Collins, et al., p. 36.
98. Sorensen, E. Noncustodial fathers: Can they afford to pay more

child support? Washington, DC: Urban Institute, 1995.
105. See Appendix 2 in the article by Fuller and colleagues in this

journal issue for a state-by-state breakdown of percentages.
99. Johnson, E., Levine, A., and Doolittle, F. Fathers fair share:
Helping poor men manage child support and fatherhood. New
106. Quinn, J. Where need meets opportunity: Youth development

programs for early teens. The Future of Children (Fall 1999)
9(2):96116.
100. For more information about the Welfare-to-Work program, see

the U.S. Department of Labor Web site at
http://wtw.doleta.gov/resources/fact-noncustodial.asp.
107. Association of Junior Leagues International. Teen Outreach Program: A three-year proposal for replication/institutionalization.
New York: AJLI, 1991.

and Means, pp. 44243.
108. Hahn, A., Leavitt, T., and Aaron, P. Evaluation of the Quantum
Opportunities Program (QOP): Did the program work? Waltham,
MA: Center for Human Resources, Heller Graduate School,
Brandeis University, 1994.
102. Under AFDC, states were required to pass through the first
$50 per month in child support to mothers. This requirement
was eliminated with the 1996 welfare reform law, but nearly half
the states have chosen to continue at least a modest pass-through
with state monies. See Turetsky, V. Realistic child support policies
for low income fathers. Kellogg Devolution Initiative Paper. Washington, DC: Center for Law and Social Policy, March 2000.
109. Philliber, S., Kaye, J., and Herrling, S. The national evaluation of
the Childrens Aid Society Carrera-Model Program to prevent teen
pregnancy. Accord, NY: Philliber Research Associates, May 2001.
103. Meyer, D., and Cancian, M. W-2 child support demonstration

evaluation, phase 1: Final report. Madison: University of WisconsinMadison, Institute for Research on Poverty, April 2001.
25
The 1996 Welfare Law:

Key Elements and Reauthorization
Issues Affecting Children
Mark H. Greenberg, Jodie Levin-Epstein, Rutledge Q. Hutson, Theodora J. Ooms,
Rachel Schumacher, Vicki Turetsky, and David M. Engstrom
SUMMARY
The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 changed the
social policy landscape for children in many
ways. It replaced the prior welfare program with
block grants to the states entitled Temporary
Assistance for Needy Families, and modified a
broad array of other programs and initiatives
affecting low-income children. This article
describes the key themes dominating the debate
over welfare reform in 1996, specifically:
Increased state discretion in program design,
leading to more variability in states eligibility
requirements and services provided to lowincome families;
More stringent work requirements even for
parents of very young children;
Time limits on the use of federal funds for
cash assistance, and a strong focus on caseload
reduction;
Increased emphasis on parental responsibility,
with stronger child support requirements; and
Increased emphasis on reducing out-of-wedlock births, including bonuses to states with
the largest reductions, and special requirements for unmarried teen parents who seek
welfare.
Although child well-being received little attention during the congressional debates in 1996,
the authors conclude with the hope that
improving child outcomes and child well-being
will emerge as a key theme when the law is reauthorized in 2002.
Mark H. Greenberg, J.D., is senior staff attorney at
the Center for Law and Social Policy (CLASP), a
nonprofit organization engaged in research, policy
analysis, technical assistance, and advocacy on issues
affecting low-income families.
Jodie Levin-Epstein is senior policy analyst at CLASP.
Rutledge Q. Hutson, J.D., M.P.H., is senior policy
analyst at CLASP.
Theodora J. Ooms, M.S.W., is senior policy analyst and
director of the Couples and Marriage Policy Resource
Center at CLASP.
Rachel Schumacher, M.P.P., is a policy analyst at
CLASP.
Vicki Turetsky, J.D., is senior staff attorney at CLASP.
David M. Engstrom, M.Sc., M.A., is a doctor of
jurisprudence candidate at Stanford Law School in
Stanford, California, and was formerly an intern at
CLASP.
27
Greenberg, Levin-Epstein, Hutson, et al.
he enactment of the Personal Responsibility

and Work Opportunity Reconciliation Act
of 1996 marked an extraordinary turning
point in U.S. social policy. The legislation is
probably best known for having repealed the Aid to
Families with Dependent Children program and for
providing states with block grants to design workfocused, time-limited welfare programs. However, the
scope of the 1996 law was much more extensive. The
law made major changes affecting a broad array of programs that provide services to low-income children,
including child support enforcement, child care, Medicaid, food stamps, child welfare, and disability benefits.
The law restricted services to immigrants and generally reduced federal protections for individuals while
expanding state discretion and flexibility in numerous
aspects of social policy. Also, the law prompted new
and intensified discussions about out-of-wedlock
births, fathers, and marriage and family formation.
Opinions about the motivating factors for the 1996
law differ sharply. Some emphasize the antagonism
toward poor families, minority women, and immigrants, or credit presidential politics and the need to
do something about welfare before the 1996 elections. In general, however, the debates did not focus
on how best to reduce child poverty. Instead, the
debates centered on:
Promoting devolution;
Reducing government spending;
Requiring work and imposing time limits for families
receiving welfare;
Promoting parental responsibility;
Restricting public benefits eligibility for legal immigrants; and
Addressing out-of-wedlock births.
In 2002, these themes may be revisited as Congress,
the states, and the public take stock of what has and has
not been accomplished since 1996. In 2002, Congress
must reauthorize the block grants to states for Temporary Assistance for Needy Families (TANF), the Child
Care and Development Block Grant, and the food
stamp program, and the debates will likely extend to
other programs and policies affecting low-income fam-
28
ilies as well. This article seeks to contribute to the discussions by describing how the 1996 law changed the
social policy landscape across a broad array of programs
and initiatives affecting children (see Figure 1). For
each program or initiative, key developments since
enactment of the law are described based on evolving
state practices, changing economic conditions, and
new research findings. In addition, key issues likely to
be before Congress in 2002 are highlighted. The policies that emerge from reauthorization will likely have a
major impact on the well-being of low-income childrenindeed, on all childrenfor many years ahead.
Temporary Assistance for Needy Families

The 1996 law repealed Aid to Families with Dependent Children (AFDC), the principal program providing cash assistance to families with children, as well as
several related programs, and replaced them with
TANF block grants.1 Under AFDC, states were mandated to provide assistance to all eligible poor families,
but had broad discretion in setting benefit levels. The
federal government paid half or more of all program
costs on an open-ended basis (that is, federal funding
rose and fell with caseload levels). States also were
required to provide work-related services and requirements for AFDC families, but these programs typically
were not a central focus in state and local welfare
administration. Such programs were often underfunded and affected a limited share of eligible families.
Under TANF, each state receives a block grant and has
broad discretion in using the funds for programs that
provide cash assistance for needy families, as well as for an
array of other benefits and services that accomplish the
purposes of the law.2 The laws four purposes are to:
Provide assistance to needy families so that children
may be cared for in their own homes or in the homes
of relatives;
End needy parents dependence on government benefits by promoting job preparation, work, and marriage;
Prevent and reduce the incidence of out-of-wedlock
pregnancies; and
Encourage the formation and maintenance of twoparent families.
Volume 12, Number 1
The 1996 Welfare Law
Figure 1
Federal Welfare Reforms Principal Changes Affecting Children
Placed more stringent work
requirements and time limits on
federal TANFa cash assistance
Included a new emphasis on
family formation
Restricted immigrant eligibility

for many services and benefits
Established stronger child support

enforcement provisions
1996
Welfare
Law
Modified the definition of childhood

disability and eligibility for continuing
assistance from SSIb
Changed the structure of two

significant funding streams for
child welfare services
Increased funding for child care
Delinked Medicaid from

welfare receipt
Restricted eligibility for food stamps
among some groups
a TANF = Temporary Assistance for Needy Families

b SSI = Supplemental Security Income
Key features of TANF block grants are summarized in

Box 1. In implementing TANF, states have generally
developed programs that provide time-limited assistance and place a strong emphasis on work.3 In most
states, parents receiving TANFincluding parents of
very young childrenare required to participate in
work-related activities, and all cash assistance can be
terminated when parents violate a program rule or
reach a time limit of 60 months or less. State programs generally emphasize rapid entry to the workforce and restrict education and training activities.
Liberalized eligibility rules adopted by many states
have expanded the availability of cash assistance for
two-parent families and those entering employment.
However, new restrictions were placed on federal

assistance to certain groups, such as immigrants and
teen parents.
As states implemented the 1996 law, there was an
unprecedented decline in the nations welfare caseload. Implementation of TANF coincided with
expanded employment opportunities in an extraordinarily strong economy and a number of policy
changes that greatly increased returns from low-wage
work (that is, an increase in the minimum wage,
expansion of the earned income tax credit, and
increases in child care spending and public health care
coverage). The decline in the welfare caseload began
before TANF was enacted, but accelerated afterward.
29
Box 1
Key Features of TANF Block Grants
Essentially fixed funding levels from 1997 through 2002. Federal block grant levels were set to reflect federal spending during the early 1990s for the programs that were repealed when
TANF was enacted (referred to as antecedent programs). In
addition, a maintenance-of-effort (MOE) requirement stipulates
that, to avoid a fiscal penalty, a state must spend nonfederal
dollars representing at least 80% (or, if the state meets TANF
work participation rates, 75%) of the amount the state spent in
1994 on a set of AFDC-related programs.
tance; a higher rate applies to two-parent families. To count

toward participation rates, an individual must be involved in
one or more of the listed work-related activities for a specified
number of hours each week throughout the month. Education
and training count toward the rates only to a very limited
extent. A states participation rate requirement can be reduced
if the states caseload has declined since 1995 for reasons
other than changes in eligibility rules; this caseload reduction
credit creates a strong incentive to reduce caseloads.
Broad state discretion in use of TANF funds. Unless otherwise

prohibited, a state can spend TANF funds in any way reasonably calculated to accomplish any of the purposes of the law
(or in any way permissible under the antecedent programs). A
state may also transfer up to 30% of its TANF funds to programs under the child care and social services block grants.
A time limit on federally funded TANF assistance. States are

prohibited from using federal TANF funds to provide assistance
to a family for more than 60 months. However, a state may
allow exceptions for up to 20% of its caseload, and may use
MOE or other state funds to continue to provide assistance to
others. In addition, the time limit does not apply to benefits and
services that are not defined as assistance, as noted below.
Elimination of federal entitlements to assistance. States

determine which families are eligible for assistance, and are not
required to provide assistance to any family or group of families
(although federal law prohibits use of federal TANF funds to
assist some categories of people, such as certain immigrants).
Work participation rate requirements. A state risks a fiscal
penalty unless a specified percentage of families receiving
TANF assistance are participating in work or work-related
activities. One rate is calculated for all families receiving assis-
A distinction between assistance and nonassistance.

Key requirements (such as time limits, work requirements, and
child support cooperation) apply only to families receiving
assistance. Assistance includes benefits designed to meet
ongoing basic needs and support services for families who are
not employed. Many important benefits provided to employed
families, such as child care, transportation, refundable earned
income tax credits, and work expense allowances, are not considered assistance.a
a For a discussion of the significance of the assistance/nonassistance distinction in designing state policies and programs, see Greenberg, M., and Savner, S. The final TANF
regulations: A preliminary analysis. Washington, DC: Center for Law and Social Policy, 1999.
In early 1994, 5.1 million families were receiving

AFDC assistance. By the time TANF was enacted, in
August 1996, the number had fallen to 4.4 million
families. By September 2000, only 2.2 million families were receiving ongoing assistance.4 Child poverty
also fell during this period, from 22% in 1994 to 17%
in 1999. However, the welfare caseload dropped
much faster, and the share of poor children receiving
assistance fell from 62% in 1994 to 40% in 1999.5
30
Many families leaving welfare are still living in poverty.

Studies consistently have found that about 60% of families leaving welfare are working, but often in low-paying
jobs with no employer-provided benefits, and often
without continued food stamp benefits or Medicaid coverage even when still eligible. (See the article by
Zedlewski in this journal issue.) The other 40% of families leaving welfare are not working. Based on the limited information available, it appears that somethough
Volume 12, Number 1
Box 2
Potential TANF-Related Reauthorization Issues
(1) Should TANFs purposes be modified? A major question will
be whether to include poverty reduction among the purposes of TANF. There may be proposals to modify TANF purposes
to include providing supports and promoting employment
retention and advancement for the working poor. Some policymakers may also urge that the law place a stronger emphasis
on addressing fathers involvement and promoting marriage.
(2) Should federal funding and state maintenance-of-effort
(MOE) levels be changed? Some policymakers will likely
argue that federal funding is too high in light of caseload
declines since 1994. Those seeking to defend or expand block
grant funding will argue that existing funding levels should be
sustained or increased because states have broadened their
focus beyond welfare and now serve low-income families
more generally, and because there is still no experience concerning the adequacy of block grant funds in a recession.
(3) Should federal time limit rules be modified? How should
federal law address the families still receiving assistance?
Caseload declines have meant that 20% of the caseload (the
percentage allowed for exceptions to the time limit) is a far
smaller number of families than contemplated in 1996. Some
groups will likely argue that there should be no federal time
limit, or that the 20% cap on exceptions should be reconsidered. There will be proposals to allow exemptions or extensions for particular types of families, such as working families,
families with disabled members, or families with infants.
Moreover, now that the caseload is smaller, the discussion of
time limits is likely to lead to a broader discussion of how to
not mostof these families are residing with partners or

other adults, and many face multiple obstacles to
employment. In some states, a significant share of case
closures are due to sanctions or noncompliance with
program requirements, such as failure to meet a work
requirement, attend a meeting, or respond to a notice.6
Families whose cases are closed because of sanctions are
likely to have a low education level, little or no work history, and more serious employment barriers.7
better address the difficult circumstances faced by many of

the remaining familiesincluding illness, disability, mental
health problems, substance abuse, language barriers, and
domestic violence.
(4) How should states performance be measured? For many,
caseload decline was a principal measure of success under
TANF. In 2002, some lawmakers will likely advocate that state
success in reducing child poverty or improving family economic well-being be treated as a key measure of performance.
An increased focus on out-of-wedlock births and other family
formation issues is likely as well. The basic issue of state
accountability for results is likely to be central in these
debates.
(5) Do states sanction policies require more federal safeguards? States have implemented varying levels of protection
before reducing or terminating assistance to a family for violating a program rule. State discretion in sanction policy is
believed by some to have contributed to the numbers of families leaving welfare without work, and to the deepening poverty of the poorest female-headed families. As a result, critics of
this policy may urge that before states can terminate assistance, they should be required to strengthen their efforts to
identify and address problems faced by families being sanctioned. Many states, however, emphasize that their flexibility in
program administration under TANF has been key in promoting
employment and reducing caseloads, and may oppose provisions that could be seen as restricting their discretion.
Concerns about families leaving welfare without

employment have been heightened by findings that
the poorest families have experienced a drop in
income in recent years. Between 1995 and 1998,
female-headed families generally experienced increases in disposable income, but the poorest 20% suffered
a loss over this period, principally because of the sharp
drop in receipt of various welfare-related benefits.8
31
Before the 1996 law was enacted, no federal program expressly

sought to promote marriage, but the issues surrounding family
formation have gained increasing public attention.
Among families still receiving TANF assistance, many
have serious barriers to employment, such as no
recent work history and no high school diploma.
About half report poor physical or mental health.9
Other barriers include illiteracy, substance abuse,
domestic violence, and ill or disabled family members.
Nationally, as the welfare caseload declined and cash
assistance expenditures fell, more funds became available for programs to help families address barriers to
employment and for other state welfare-related activities.10 States were able to use TANF funds to serve
poor families that left TANF or that never received
TANF cash assistance, as well as those still on the
rolls, and many states redirected TANF funds to child
care and child welfare services.11 Some states used
TANF funds to substitute for prior state spending,
giving rise to disputes about the magnitude and significance of such supplantation.12
Some of the key questions likely to emerge during
TANF reauthorization are summarized in Box 2.
Family Formation
Before the 1996 law was enacted, no federal program
expressly sought to promote marriage, but the issues
surrounding family formation have gained increasing
public attention as the incidence of children living in
divorced, never-married, and teen parent households
has grown. The proportion of children living with only
one parent has more than doubled in the past 30 years,
from 12% in 1970 to 27% in 1998.13 In the 1960s and
1970s, most of the growth in single-parent families was
caused by increases in divorce, but in the next two
decades nearly all the growth was driven by increases in
out-of-wedlock childbearing.14 Although fewer than
4% of all births were to unmarried women and adolescents in 1940, 33% of all births were outside of marriage by 1999.15 (See the article by McLanahan and
Carlson in this journal issue.)
Marital status and teen pregnancy became key issues in
the debates that led to the 1996 law. Inspired by con-
32
servative authors such as Charles Murray, who

described illegitimacy as the single worst social problem of our time,16 some welfare critics viewed cash
assistance as enabling poor women to have children
out of wedlock and sought to limit or eliminate assistance as a means of reducing nonmarital births. Proposals were introduced to prohibit states from
providing assistance for children born out of wedlock
to teen parents and for children born to families receiving welfare. Although these proposals were eventually
dropped, states were free to adopt such policies under
the block grant structure,17 and several other provisions
related to family formation were successfully incorporated into the law.18 In fact, three of the four purposes
of TANF refer to family formation: promoting marriage, reducing out-of-wedlock pregnancies, and
encouraging the formation and maintenance of twoparent families. Further, states were given broad flexibility to determine program rules and the range of
TANF services to be provided. As a further incentive
for states to develop effective programs in these areas,
the law established state bonuses for reductions in outof-wedlock births, and mandated restrictions for minor
parents receiving TANF.
Since 1996, the majority of states have made some
effort to pursue these family formation goals.19 Many
states are using TANF funds to prevent out-of-wedlock
births, focusing primarily on teen pregnancy prevention, and most states have taken policy measures to
strengthen two-parent families. Highlights of activities
and progress toward implementing these provisions of
the law are summarized below.
Promoting Marriage
Although marriage rates have declined and divorce
rates have increased dramatically over the last 40 years,
polls show that most Americans continue to prize and
value marriage as an important life goal.20 Yet a great
deal of uncertainty surrounds the appropriate role of
government with respect to marriage.21 To date, states
have taken few steps to address this purpose of TANF
directly, although various strategies have been suggest-
Volume 12, Number 1
ed.22 As of mid-2000, only two states had adopted specific strategies to strengthen marriage: Oklahoma had
launched a multisector initiative using $10 million in
TANF funds to strengthen marriages,23 and Arizona
had committed $1.6 million for marriage education
and other activities related to abstinence promotion.
No research is available to date concerning the relative
efficacy of these efforts. At the same time, new studies
suggest that well-established programs whose primary
purposes are to enhance economic security or to provide other kinds of family supportsuch as child support enforcement, family planning, and expanded
Medicaidmay effectively, although indirectly, promote marriage and reduce nonmarital childbearing.24
Nevertheless, congressional interest in enacting legislation specifically focused on promoting marriage
appears to be growing.25
Reducing Out-of-Wedlock Pregnancies
Various provisions were included in the 1996 law to
support states efforts to reduce out-of-wedlock pregnancies, especially among teens. First, the law required
the U.S. Department of Health and Human Services
(DHHS) to establish a strategy for preventing nonmarital teen pregnancies and to assure that at least 25%
of U.S. communities had teen pregnancy prevention
programs in place by January 1, 1997.26 States could
spend TANF funds on teen pregnancy prevention and
family planning services, with few restrictions. As of
2000, at least 34 states had tapped some TANF funds
for such projects, ranging from after-school programs
to peer education to media campaigns.27
Second, the law allocated nearly $440 million in combined federal and state funds over five years for a specific type of abstinence education. The funds are available
for curricula that teach, in part, that sexual activity outside the context of marriage is likely to have harmful
psychological effects, and that abstinence is the only
appropriate option outside of marriage, regardless of
age.28 Programs are not to provide information on how
to use contraception. As of 1999, at least five states
passed laws applying the abstinence-unless-married criteria to all sexuality education programs in the state.29
In 2000, Congress authorized additional funds of $50
million, and in 2001, added another $10 million, for
implementation of such programs.30
Third, several provisions in TANF focused specifically

on unmarried minor parents, even though the number
of teen parents in the TANF caseload is quite small.31
With limited exceptions, the law prohibits states from
providing federal TANF assistance to unmarried
minor, custodial parents who do not participate in
school or training and who do not live in an adultsupervised setting.32 As states have implemented these
provisions, anecdotal evidence from one local study
suggests that confusion has been widespread and that
local caseworkers have inappropriately diverted teen
parentsminors as well as those who are olderfrom
receiving TANF.33
Some recent trends are encouraging. For example, the
teen birth rate has declined significantly since 1991.
However, much of this decline occurred before welfare
reform. Relatively little research has examined the
impact of efforts implemented since reform to reduce
out-of-wedlock pregnancies. Rigorous evaluation studies of programs that teach only abstinence are scarce,
and it is not yet known whether and which 1996 abstinence education programs might be effective.34 A federal evaluation of the 1996 abstinence programs is
under way, with interim results due in early 2003. With
respect to provisions involving minor teen parents,
some studies indicate that, under some circumstances,
the school requirement (often called Learnfare) can
improve a teen parents enrollment, grade completion,
and chances of receiving a high school equivalency
diploma,35 but that the requirement appears to have no
impact on teen parent fertility.36 There has been virtually no research on the impact of the minor parent livingarrangement rule.37
Strengthening Two-Parent Families
Most states used the new flexibility under TANF to
drop the stricter eligibility requirements for two-parent
families that had existed under AFDC. As of 1999, 33
states were treating two-parent (married or unmarried)
families the same as single-parent families when determining eligibility. At the same time, under the 1996
law, states are required to meet a higher work participation rate for two-parent families receiving assistance
than for single-parent families. Many states viewed the
higher rate as a disincentive to assisting two-parent
families in their TANF programs, and at least 14 states
established state-funded programs for two-parent fam-
33
Box 3
Potential Family FormationRelated Reauthorization Issues
(1) Is new legislation needed to promote and support healthy
marriages? Some policymakers are likely to argue that states
have not put sufficient effort into pursuing the TANF family
formation goals and that further measures are needed. Proposals may include requiring states to set aside a certain percentage of their block grants for marriage and family
formation activities; to establish financial incentives for lowincome parents to marry; or to provide relationship education courses for high school students and low-income
parents. Alternative suggestions include investing in promising yet indirect approaches to enhancing marriage (such as
child support and Medicaid programs) and in carefully evaluated demonstration programs designed to strengthen marriage to learn what works before implementation becomes
widespread. At the same time, concerns have been raised
about whether TANF is the appropriate vehicle for implementing a marriage agenda, and about the danger of enacting policies that may perpetuate abusive or otherwise harmful
marriages.
(2) Should the provisions regarding teen parents be changed?
Reauthorization provides an opportunity to consider a range of
approaches to the teen parent provisions. Should states be
given more flexibility in designing programs for minor parents? Are better assessments of individual and community
capacity to serve teen parents needed? Should the current

competition between teen parents and adults for limited education slots be eliminated?a Should special incentives and
performance standards (or a competitive grants program) be
developed to encourage and reward expanded investments,
improved outcomes (including completion of education), and
effective coordination of youth services?
(3) Should the work participation rate for two-parent families
be changed? The requirement that states engage 90% of twoparent families in a work activity is expected to be revisited.
Many observers contend that a separate, higher participation
rate is not needed and only acts as a disincentive to serve twoparent families.
(4) Should the bonus for out-of-wedlock births be changed?
Congress will likely consider whether the illegitimacy bonus
affects state and family behavior, and whether the funds are
effectively targeted. There may be proposals to expand, contract, or eliminate the bonus, or to condition the bonus on new
investments designed to reduce nonmarital births, such as
family planning initiatives or couples counseling. The American Public Human Services Association, for example, has suggested that the bonus focus more on teen pregnancy
prevention.
a Under current TANF rules, no more than 30% of adult recipients counted toward a states participation rate may be included either on the basis of participation in voca-
tional educational training (for no more than 12 months) or, if under age 20 and the head of household or a married parent, on the basis of engagement in education or
school completion.
ilies. Two-parent families remain a very small proportion of the caseload (around 5%),38 and eligible twoparent families have very low participation rates in
TANF, food stamps, and Medicaid.
State Bonuses
Two bonuses have been adopted to promote family
formation goals. An award of $100 million annually is
available for up to five states that have the highest
reduction in their out-of-wedlock birth ratio and that
34
also have reduced their abortion rate. The formula is

based on the share of nonmarital births among women
of all incomes and ages within the state (not just TANF
recipients). In addition, up to $200 million annually is
available to states to reward high performance in several categories, including, beginning in 2002, up to $10
million a year to be divided between 10 states in the
new category of family formation and stability.39 The
bonus will be awarded based on increases in the per-
Volume 12, Number 1
Welfare reform has affected the child support program in both

direct and indirect ways. ... Between 1995 and 2000, collection
rates increased from 19% to 42%.
centage of all children (not just low-income children)
living in married-couple families.
It is unclear, however, whether the TANF bonus structure has motivated states to address family formation
issues. In three of the five states awarded the illegitimacy bonus, state officials indicated that they undertook no special activity to win the bonus.40 The bonus
category of family formation was added only recently,
so it is too early to tell if it will motivate states to focus
on marriage promotion.
Among those concerned about the impact and effectiveness of family formation initiatives, key questions
likely to be raised during reauthorization will focus on
marriage, minor parents, the two-parent family participation rate, and state bonuses, as summarized in Box 3.
Child Support
Child support can provide a significant income source
for low-income families and reduce child poverty.41 It
also can help increase a single mothers labor force participation, stabilize and supplement low-wage earnings,
link families to private and public health care coverage,
and reinforce paternal involvement.42,43 However, many
TANF families cannot count on child support as a steady
source of income when their welfare benefits end.
Child support is collected for only 44% of families that
have left welfare, and for just 25% of families that are on
welfare.44 In part, this is because most states do not provide adequate child support services to the families in
their caseloads, despite marked improvements (discussed
further below). It is also because TANF families often
involve never-married parents, who are the most difficult
to serve. Thus, poor children whose parents never married are the least likely to receive child support.45 Finally,
it is because many fathers of poor children are themselves
poor and have limited ability to pay.46 (See the article by
McLanahan and Carlson in this journal issue.)
Nearly two-thirds of all child support cases in the country are processed through the public child support pro-
gram, one of the largest human services programs

reaching low-income mothers, fathers, and children.47
Changes to the child support program under the 1996
law, and subsequently under the Child Support Performance Incentive Act of 1998, were intended to increase
mothers and fathers cooperation in establishing paternity and support orders, and increase child support collections.48 Major provisions under these laws:
Linked federal and state databases to improve matching of child support orders with information on
newly hired employees and quarterly wages, and
other data sources;
Expanded states authority to order genetic tests,
subpoena information, adjust orders, order income
withholding, suspend licenses, secure and seize
assets, report to credit bureaus, conduct quarterly
bank matches, and enforce interstate cases;
Required that birth certificates include the fathers
name only if paternity is formally acknowledged
under hospital-based procedures;
Tightened cooperation rules to require at least a 25%
sanction (up to 100% by state option) on TANF assistance for failure to comply with child support
enforcement efforts, and extended cooperation rules
to Medicaid and foster care programs (and food
stamp programs by state option);
Changed support assignment and distribution rules
to allow families that have left welfare to keep more
child support, but eliminated the requirement that
gave the first $50 of support collected to current
TANF families; and
Established new performance incentives and report49
ing requirements.
Welfare reform has affected the child support program
in both direct and indirect ways. Increased automation,
expanded paternity establishment procedures, new
database matching, and expanded federal tax offsets
have helped states improve program performance
markedly since reform. Between 1995 and 2000, collection rates increased from 19% to 42%.50 At the same
35
time, welfare reform indirectly impacts the child support program in a number of important ways due to
the programs fiscal structure, the complex interface
with TANF, a new focus on low-income fathers, and
concerns about domestic violence.
The Fiscal Structure of the Program
The child support program originally was designed to
reimburse federal and state welfare costs. When families apply for welfare assistance, they are required to
cooperate with the child support program and assign
(or relinquish) their rights to child support to the state.
For the most part, any assigned child support collections that are received are shared between the state and
the federal government as reimbursement for welfare
benefits, and are used by virtually all states to help pay
for child support or TANF program costs. But today,
only about 20% of the child support caseload is comprised of families receiving TANF. As welfare caseloads
have declined, most families being served by the child
support program are working families with incomes
below 250% of poverty who have left welfare or never
received welfare. This change has prompted a reexamination of the child support programs dual and often
conflicting goals of promoting family self-sufficiency
and of recovering welfare costs, and of the difficulties,
for both families and states, stemming from existing
assignment and distribution policies.
From the familys perspective, once child support rights
have been relinquished, it makes little financial difference
to the family whether the noncustodial parent pays child
support because the government retains nearly all the
payment.51 Not surprisingly, therefore, less child support
is collected for families receiving welfare than for equally poor families not receiving welfare.52 To encourage
increased child support collections for families receiving
welfare and to strengthen child support as a long-term
income supplement for working families, the rules were
revised in 1996, allowing families to keep more child
support once they leave welfare. However, local child
support programs have found the distribution rules to
be complicated and costly to administer.53 A demonstration project in Wisconsin that allowed all support collected to be passed through to TANF families has shown
some positive results, however.54 Early results indicate
that noncustodial parents in TANF families are more
likely to establish paternity and pay support under these
36
circumstances, and families participating in the project

experience several other positive outcomes as well.
From the states perspective, the decline in TANF caseloads has resulted in fewer child support collections for
families receiving TANF and, thus, declining government revenues to support program operations.55
About two-thirds of states use their share of welfare
collections to help meet their TANF/maintenance-ofeffort (MOE) obligation (see Box 1), but the remaining one-third use it to pay their share of child support
program costs.56 In response to shrinking revenues, a
number of states have begun to consider how to refinance their programs, and DHHS has engaged stakeholders in a national consultation process to examine
how welfare collections and program costs should be
distributed among families, the federal government,
and states. Whether states should be allowed to use
TANF funds to help pay for child supportrelated
activities is also under consideration.
Complex Interface with TANF
Welfare reform also has affected the child support programs intake mechanisms. As TANF-funded components increasingly are operated by multiple state and
local agencies and private organizations, the TANF and
child support interface has become more complex.
Moreover, an increasing number of families are entering
the child support system through programs such as
Medicaid, food stamps, and foster care, instead of
TANF. Discussion has emerged about whether the child
support program should replace TANF as a hub,
interacting more closely with family members, developing case management capacity, and developing links with
other public means-tested and community-based programs, including private and public health care coverage
for children who would otherwise be uninsured.57
New Focus on Low-Income Fathers
Welfare reform has prompted a new focus within the
child support program on low-income fathers. For
example, some states have implemented initiatives to
enhance low-income noncustodial fathers ability to
pay support by allowing them to participate in welfareto-work programs. Working in partnership with local
child support agencies, these programs generally provide a range of services to help low-income fathers,
including: (1) job readiness, job search, and postem-
Volume 12, Number 1
Box 4
Potential Child SupportRelated Reauthorization Issues
(1) Should the fiscal structure of the child support enforcement
program be changed? One set of issues involves whether and
how the child support program can shift from a fiscal structure that supports cost recovery to one that supports family
self-sufficiency. A second set of issues concerns whether and
how to decouple the existing funding links between the child
support and TANF programs. These links include the revenuesharing formula for retained welfare collections, and the use
of retained welfare collections to fund TANF-related MOE
expenditures. There is increasing political support to eliminate
the TANF assignment requirement and distribute all collected
support to families. In addition, there may be interest in reexamining the federal funding structure.
ployment services to help them find and keep employment; (2) case management services to help them
negotiate the child support system; and (3) curriculum-based peer support groups and mediation to
improve their parenting and relationship skills.58 In
addition, discussion about how to improve the child
support system has been broadened to include impacts
on paternal involvement and child well-being.59 In particular, more attention has been focused on how federal and state child support policies might affect
marriage, cohabitation, and fragile families. (For further discussion of this topic, see the article by McLanahan and Carlson in this journal issue.)
Concerns about Domestic Violence
Concerns have been raised about whether tightened
cooperation requirements, coupled with expanded
databases and more aggressive paternity establishment
and enforcement procedures, will increase the risk of
domestic violence for women in the child support caseload.60 Although recent research indicates that most
domestic violence victims decide to actively pursue child
support if safety and confidentiality issues are addressed,
concerns remain about the number of TANF families
(2) How should the child support program relate to the broader
working families agenda? The increasingly complicated
ways families enter the child support system, and the difficulty in linking nonwelfare families to child support services (as
well as other public supports), are focusing more attention on
child support outreach and intake procedures and interagency
coordination. Should the child support program develop as a
hub, linking families with food stamp programs, employment programs, responsible fatherhood programs, domestic
violence programs, and even health insurance programs?
How should child support relate to low-income fathers and
fragile families? Should the program assume an explicit
role in supporting paternal involvement and family formation?
sanctioned for child support noncooperation,6 and the

extension of child support cooperation requirements to
other programs and benefits. At the same time, early
findings from the Wisconsin demonstration suggest
that passing through support payments to families may
reduce family conflict around child support.54
In sum, the 1996 law had important direct and indirect
impacts on the child support program and the families
it serves. Policy issues related to the fiscal interaction
between the child support program and TANF, as well
as the role of child support in the broader agenda for
working families, are likely to be a key part of the reauthorization discussions, as summarized in Box 4.
Child Care
Before the 1996 welfare law, multiple federal funding
streams for child care existed, each with its own policies
and procedures. Three federalstate matching child care
funding streams provided child care for AFDC families
in work programs or approved education and training
programs, for families leaving AFDC due to employment, and for working families considered at risk of
relying on AFDC without help in paying child care
37
Provisions included in the final welfare reform law consolidated

child care funding into a single new funding stream, ... increased
funding levels, ... [and] provided for state flexibility to determine
most matters affecting families access to subsidies.
costs. In addition, the Child Care and Development
Block Grant provided federal funds to states, with no
state matching requirements, for child care for lowincome families in work or education and training programs, and for initiatives to enhance child care quality.
During the 1996 debates, interest arose in reducing
the fragmentation and complexity that resulted from
these four separate funding streams. In addition, concerns were raised about the potential impact of TANF
work requirements on the need for child care, the
funding level that would be necessary to assure
access, and the tension between quantity of children
served and quality of care. Provisions included in the
final welfare reform law:
Consolidated child care funding into a single new
funding stream, referred to as the Child Care and
Development Fund (CCDF) in federal regulations,
to provide a basic funding level to all states and a
capped amount of additional matching funds to
states that maintained their prior spending levels;
Increased funding levels above anticipated spending
under prior law, and allowed states to spend TANF
funds directly for child care and to transfer up to
30% of TANF funds to the CCDF;
Repealed guarantees to child care that had been
provided under AFDC for families in approved
activities and families in their first year leaving
AFDC due to employment, but allowed states the
option of providing such guarantees and setting
maximum income eligibility levels up to 85% of a
61
states median income;
Provided for state flexibility to determine most matters affecting families access to child care subsidies,
such as how much to pay child care providers, how
much to ask parents to pay toward the cost of child
care, and how stringent the health and safety standards should be for providers who may care for subsidized children;62
38
Prohibited states from sanctioning single custodial

TANF parents with children under age six when
child care was unavailable (although this provision
does not stop the 60-month federal time limit); and
Required states to spend at least 4% of their CCDF
funds on measures to improve child care quality.
Since 1996, use of federal and state funds for child care
has increased significantly, but with large variations in
child care policies among the states on key issues affecting access to child assistance. According to the most
recent data from the U.S. Department of Health and
Human Services Child Care Bureau, federal and state
spending on child care totaled $8.0 billion in 2000.63
Spending under CCDF (including funds from prioryear CCDF allocations and those transferred from
TANF) totaled $7.0 billion, more than double 1996
expenditures under the previous child care funding
streams.64 In addition, the Child Care Bureau reported
direct TANF spending on child care of $1.0 billion in
2000, but other research indicates that states more
likely spent about $1.5 billion in direct TANF funds,
plus another $873 million in state TANF MOE funds,
on child care.65 As a result of TANF transfers to CCDF
and direct TANF expenditures on child care, states
reliance on TANF funds for child care has increased.
Through transfers and direct expenditures combined,
states redirected TANF funds totaling $3.9 billion to
child care in 2000, more than the entire federal portion
of the CCDF allocation. Moreover, states spending on
initiatives to improve the quality of child care also has
increased beyond the required 4% setaside, reaching
6.1% of total federal and state CCDF expenditures in
fiscal year 2000.63 To increase the supply of high-quality infants and toddler child care specifically, Congress
has earmarked $50 million to $100 million in CCDF
funds annually beginning in 1998.
The increased funding for child care has resulted in
more families and children receiving child care subsidies. The number of children receiving CCDF subsi-
Volume 12, Number 1
Box 5
Potential Child CareRelated Reauthorization Issues
(1) Should there be a significant increase in CCDF funding?
Demand for child care assistance has grown, as approximately one million additional single women have entered the workforce. Although federal and state spending on child care has
expanded significantly, many states still have waiting lists for
subsidies, and state policy choices are constrained by
resource limitations. In considering this issue, Congress is
likely to review current law regarding use of TANF for child
care.
care for infants and toddlers, sick and disabled children, children in low-income areas, and children whose parents have
nontraditional work schedules.
(4) Are current provisions on payments to providers adequate to
attract enough qualified child care workers? As turnover
increases in the child care field, and child care providers face
growing difficulty in hiring and retaining staff, discussion may
arise concerning the need to increase the compensation rates
for child care workers. Current guidelines that rely on the market to set rates are likely to be reconsidered, along with other
strategies used by some states to increase child care worker
compensation directly.
(2) Are families aware of their eligibility for subsidies, and are
they able to access the subsidy system? Some attention may
focus on whether families who are working and qualify for
CCDF assistance are aware of their eligibility. Of particular
concern are families who have moved into the labor force from
TANF, and other low-income working families. Questions have
also been raised about how administrative rules and procedures, including eligibility certification periods, may limit
access to, and maintenance of, child care subsidies.
(5) Should promotion of child development be a central goal of

the CCDF? Concerns about school readiness among young
children in child care have led to questions about whether the
current quality set-aside is sufficient and whether current
strategies to enhance child care quality are adequate.
(3) Is the supply of child care providers sufficient to support

parent choice of care, especially for specific populations?
One of the purposes of the CCDF is to enable families to exercise choice in care, but this requires a full array of options.
Concerns may be raised about the supply of high-quality child
(6) Do protections for TANF families need to be strengthened?

Questions may be raised about whether single-parent TANF
recipients with children under age six are aware of the protections provided by TANF, and whether families who lack necessary child care are being sanctioned in violation of federal law.
dies increased from an estimated 1.0 million in 1996 to

1.8 million in 1999.64 However, many eligible families
still are not receiving child care subsidies,66 and at least
17 states have waiting lists.67 One review found that in
most of the states studied, less than one-third of the
families that left welfare and were working were receiving child care subsidies.68 Other research has found
that administrative hurdles make accessing and maintaining subsidy assistance a challenge for families that
are also trying to maintain employment.69 (See the article by Fuller and colleagues in this journal issue.)
State child care subsidy policies vary according to state
priorities and resources.70 Although most states have
made investments in child care subsidy programs since

1996, a comparison of state child care policies in 1995
and 2000 suggests some areas of concern.67 For example, researchers found that although most states have
increased their income eligibility limits, these limits
have declined as a percentage of state median income.
In addition, though some states lowered the copayment charged to families with a child in care, the
copayment amount rose as a percentage of income for
some families. Also, most states still prioritize TANF
and post-TANF families over other low-income working families, although the overall proportion of families receiving child care subsidies and not receiving
39
TANF has grown.71 Finally, more than half the states

pay child care providers at rates that are lower than the
rate federal regulations suggest would be considered
sufficient to provide equal access to child care.72
Child care policy and funding issues are likely to be
important in the reauthorization debate. As outlined in
Box 5, key issues include funding levels, program goals,
parent awareness, and choice of child care providers.
Medicaid
Medicaid is the principal federalstate program providing health care coverage for low-income children and
their parents. Most Medicaid expenditures are for elderly and disabled individuals, but about two-thirds of
Medicaid recipients are children or parents of children.
Although states are not required to participate in Medicaid, all states do participate because the federal government pays half or more of the costs of Medicaid
benefits for eligible persons.73
During the debates about the 1996 welfare law, the
Republican leadership originally proposed that Medicaid be converted into a block grant to states. The proposal faced strong opposition and was ultimately
dropped, but the final legislation did make several significant changes to the Medicaid program. First, the
1996 law banned most future legal immigrants from
receiving federally funded Medicaid assistance for five
years after coming to the United States, with an exception for emergency services. (For a more detailed discussion of these provisions, see the Immigrants section
later in this article.) Second, the law delinked Medicaid
from welfare cash assistance.74 AFDC recipients were
automatically eligible for Medicaid, and families leaving
AFDC due to employment could qualify for up to a
year of transitional Medicaid coverage. But in shifting
from AFDC to a block grant, policymakers recognized
that such automatic eligibility would be problematic
under the new law. Because states would have the discretion to determine who is eligible for TANF and for
how long, tying Medicaid eligibility to TANF receipt
could result in inappropriate contractions or unintended expansions of Medicaid coverage. To resolve the
issue, Congress created a new Medicaid eligibility category known as Section 1931.
40
Under Section 1931, family members may qualify for

Medicaid by satisfying the income, resource, and family
composition rules that applied in the states AFDC program on July 16, 1996, regardless of whether the family is receiving TANF assistance. A family losing Medicaid
eligibility under Section 1931 due to employment may
qualify for up to a year of transitional Medicaid assistance. Section 1931 also gives states the option of making their income and resource eligibility rules more
liberal and extending Section 1931 coverage to lowincome two-parent families.75
Despite these provisions, Medicaid enrollment for parents and children declined during initial implementation of the 1996 law. Between 1995 and 1997,
enrollment of nondisabled, nonelderly adults dropped
by 11%, and enrollment of nondisabled children
dropped by 3%.76 One key factor contributing to the
drop in Medicaid enrollment was confusion among
both families and welfare workers about the new eligibility rules. A Kaiser Commission study found that misperceptions and confusion about Medicaid eligibility
kept many families from enrolling.77 Other researchers
have documented failures to accurately apply Medicaid
eligibility rules for new applicants and families leaving
welfare.78,79 As a result, some families erroneously concluded that they were ineligible, failed to complete the
application process, had their applications mistakenly
denied, or had their assistance wrongly terminated.80
Beginning in June 1998, however, Medicaid enrollment began to climb in some states, particularly among
families, children, and pregnant women.81,82 This was
due, in part, to improved practices for linking families
with Medicaid in some states, and to increased oversight at the federal level.83 Enactment of the State Childrens Health Insurance Plan (SCHIP) in 1997 also
contributed to increased Medicaid enrollment. SCHIP
provides federal funds for state health coverage for children in families with incomes above the Medicaid cutoff, either through new programs or expanded
Medicaid eligibility.84 In addition to expanding eligibility, SCHIP prompted many states to simplify the application process for childrens Medicaid coverage, which
also contributed to higher participation rates.
Although the overall number of children with health
insurance grew between 1998 and 1999, poor chil-
Volume 11, Number 2
Box 6
Potential Medicaid-Related Reauthorization Issues
(1) How can the interaction between TANF and Medicaid be
improved? Proposals are likely to be introduced to help
ensure that families seeking TANF assistance are made aware
of Medicaid and are able to apply for coverage, regardless of
the disposition of their request for TANF. Discussions will also
likely focus on the decline in Medicaid coverage when families leave TANF and how to increase continuing coverage for
eligible families.
(2) How should transitional Medicaid be structured? Discussions are likely to focus on addressing the complexity of tran-
dren and their parents continue to be uninsured at

much higher rates than are middle- or high-income
families. Overall enrollment figures mask the extreme
variation among states, however.85 In some states,
poor children are more likely to be uninsured now
than they were before the 1996 law was enacted,86
and the low rates of Medicaid enrollment have raised
concerns. But in several other states, Section 1931
has led to expanded Medicaid coverage for lowincome working families. The Medicaid-related issues
likely to arise in connection with TANF reauthorization are summarized in Box 6.
Food Stamps
The food stamp program is the principal federal food
assistance program, and provides a critical supplement
to family income for poor households. Most food
stamp recipients are children, and nearly 80% of food
stamp benefits go to households with children.87 In
2000, the average household with children that
received food stamps had a gross monthly income averaging $727, and received a monthly food stamp benefit averaging $234.88 The federal government pays the
full cost of the benefits and shares program administrative costs with the states.89
sitional Medicaid, the reporting requirements for participating

families, and the restrictive eligibility conditions.
(3) Should family Medicaid coverage be expanded? Recent
expansions in childrens health insurance coverage have
meant that children often qualify for Medicaid when their parents do not. States can opt to expand parental coverage, but
lawmakers will likely also look at whether additional federal
options, incentives, or mandates should be implemented to
promote expanded family coverage.
During congressional debates on the 1996 welfare law,

the Republican leadership proposed that the food
stamp program and a set of other federal nutrition programs be repealed, and that their funding be consolidated into food assistance block grants to states. As was
the case with Medicaid, however, the proposal faced
strong opposition and was dropped. Nevertheless, curtailments in food stamp program eligibility and benefits ultimately represented about half the spending
reductions originally projected under the 1996 law.90
Key changes in the food stamp program:
Restricted food stamp eligibility for able-bodied adults
without children to no more than 3 months in a 36month period, with exceptions for individuals working
or in a work activity for at least 20 hours a week;
Made most legal immigrants ineligible for food
stamps (as discussed in greater detail in the later section on Immigrants);
Restricted eligibility and reduced benefit levels by
changing eligibility rules related to household deductions and income, and by removing provisions that had
adjusted food stamp benefits to reflect inflation; and
Provided that a households food stamp benefits
could not increase if a familys TANF assistance was
41
Box 7
Potential Food StampRelated Reauthorization Issues
(1) Should more be done to ensure access to food stamps for
families leaving welfare and for working families? Because
food stamps are an important support for low-earning families, proposals will likely focus on ways to foster greater participation in the food stamp program by families applying for
and leaving TANF assistance, and by low-income working
families more broadly.
stamp program is too complex and needs to be simplified. But
(2) How can the program be simplified without compromising

other goals? Broad agreement has emerged that the food
increase program complexity in order to reduce errors.
reduced due to a sanction, and allowed states to

impose food stamp sanctions against individuals who
violated TANF rules.
Similar to trends in welfare receipt, food stamp program participation began to decline in 1994, and the
drop accelerated significantly after enactment of the
1996 law. Between March 1994 and March 2001, participation dropped from 28 million to about 17 million
persons, a decline of 39%.91 The decline in childrens
participation was especially sharp, falling from 86% of
eligible children in 1994 to 69% in 1998, with over
three-quarters of the decline coming after 1996.92 As
noted in a 1999 report from the U.S. General
Accounting Office, there is a growing gap between
the number of children living in povertyand the
number of children receiving food stamp assistance.93
Moreover, an analysis by the U.S. Department of Agriculture concluded that most of the decline in the food
stamp caseload between 1994 and 1999 was because of
a drop in participation by eligible adults and children,
not because fewer families were eligible.94 Only a small
portion (8%) was due to provisions of the 1996 law
that restricted food stamp eligibility. About one-third
of the decline (35%) occurred because household
income and assets grew during the economic expansion. The bulk of the decline (56%) was due to fewer
42
various provisions of the law concerning food stamps address

diverse policy goals, ranging from special provisions for the
elderly to child support payments and housing costs. Efforts
to simplify the food stamp program may require trade-offs
with these other goals. Also, many observers believe that the
current quality control system puts pressure on states to
eligible individuals participating in the program. A

sharp drop in participation occurred among eligible
citizen children residing with noncitizens (from 80% to
46%). Participation also dropped among families leaving welfare, even though most of these families continued to be eligible financially,95 and many still reported
food insecurity and shortages.96 Participation dropped
for individuals with earnings and those without earnings. Program administrative practices, confusion, lack
of awareness, and individual choices all appeared to
play a role in declining participation rates.97,98
The food stamp program is scheduled for reauthorization in 2002, as is TANF. Food stamp reauthorization
discussions have mainly focused on ways to increase
participation among eligible households, and on the
need for simplification and for improvements in quality control procedures, as summarized in Box 7. Another concern is whether to restore food stamp benefits for
legal immigrants, an issue discussed in greater detail in
the later section on Immigrants.
Child Welfare
The nations child welfare system is comprised of 30 to
40 separate programs designed to protect children from
abuse and neglect. The major programs are described in
Box 8. Taken together, these programs provide services
Volume 12, Number 1
Box 8
The Array of Major Child Welfare Programs
The Child Welfare Services program and the Promoting Safe
and Stable Families program provide a broad range of services, including prevention of abuse and neglect, family support,
time-limited reunification, and adoption support. Funds for the
Child Welfare Services program are discretionary, whereas
funds for the Promoting Safe and Stable Families program are
an entitlement, but the funding streams for both programs are
capped, generally with a federal matching rate of 75%.
The Foster Care Maintenance program reimburses states for a
portion of the foster care costs of certain children. Children are
entitled to federal foster care maintenance payments if they
meet the prior AFDC eligibility criteria at the time they are
removed from their homes either by a voluntary placement
agreement or by court order, and if they are under the care and
responsibility of a state welfare agency. Federal reimbursement
for the program is open-ended, with states receiving their Medicaid matching rate for each eligible child. The federal government also reimburses 75% of training expenditures and 50% of
administrative costs.
The Adoption Assistance program provides financial assistance
to adoptive parents on behalf of certain children with special
to prevent abuse and neglect; investigate reports of maltreatment and remove children from abusive or neglectful homes when necessary; provide supports to
families so that children may remain in or return to their
homes safely; and secure alternative homes for children
who cannot be returned. Federal funding for foster care
costs is open-ended, whereas federal funding for prevention and reunification services is limited.
Historically, the child welfare system connected with
the AFDC program in several ways.99 First, research
indicates that child maltreatment is highly correlated
with poverty, and the majority of children entering the
foster care system have come from families receiving
welfare assistance.100 Second, families in the two systems often face many of the same challenges: substance
needs, defined as a child whose condition makes it unlikely that

he or she will be placed for adoption without financial assistance. The adoptive parents are entitled to payments for certain
nonrecurrent adoption expenses, and, if the child meets the
1996 AFDC or current Supplemental Security Income criteria, the
state may provide the parents with ongoing payments. As with
foster care, the federal government reimburses states at their
Medicaid matching rate for each eligible child, as well as 75% of
training expenditures and 50% of administrative costs.
The Independent Living program helps adolescents transition
from foster care to living on their own. The program is a capped
entitlement to states, with a federal matching rate of 80%.
The Child Abuse Prevention and Treatment Act provides federal funding and guidelines for reporting and investigating abuse
and neglect. The act also funds innovative research and demonstration projects and community-based family resource centers.
Other federal programs supporting various child welfare services include the Social Services Block Grant (commonly referred
to as Title XX), Medicaid, and the TANF block grants.
abuse, mental or physical health problems, and domestic violence, as well as poverty. Finally, both the child
welfare system and the cash welfare assistance program
support many children who are being cared for by
grandparents and other relatives.
During the welfare reform debates in 1996, one controversial proposal called for the consolidation of several major child welfare programs into a block grant.
The proposal was defeated, however, and the legislation made few direct changes to child welfare programs. The most significant provisions that affected
child welfare programs:
Required the continuation of state foster care maintenance and adoption assistance programs;
43
Many aspects of TANF could indirectly impact the child welfare

system by affecting the risk of maltreatment, the funding
available for services, and the supports provided to kin caregivers.
Tied eligibility criteria to AFDC standards in effect
on July 16, 1996;101
Required states to consider giving preference to kin
when placing a child outside the home, provided the
relative meets state child protection standards; and
Permitted states to use TANF funds for a broad
range of child welfare services.
At the same time, however, many aspects of TANF
could indirectly impact the child welfare system by
affecting the risk of maltreatment, the funding available
for services, and the supports provided to kin caregivers. Each of these areas is discussed below.
Risk of Maltreatment
TANF could increase or decrease the risk of child
maltreatment. The more stringent work, sanction,
and time limit requirements under TANF, as well as
the decline in disposable income some families may
experience as they move from welfare to work, could
cause material hardship and stress, leading to an
increased risk of abuse or neglect.102 In addition, lack
of adequate child care for families required to work
may result in children being left home alone,103
which could be defined as neglect, depending on the
age of the children. Alternatively, TANF may result
in increased employment, leading to increased
income and self-esteem, and in turn decreased risk of
child maltreatment.
Currently, the data are inconclusive about which of
these possible influences may result from the implementation of TANF. On the one hand, in 1999 the
incidence of substantiated child maltreatment
decreased for the sixth year in a row.104 On the other
hand, the estimated number of children in foster care
grew from 483,000 in 1995 to 568,000 in 1999,105
and several research studies raise concerns that TANF
could be detrimental to child welfare. Two studies
looking at AFDC caseload data found that sanctions
and work combined in ways that were associated with
increased involvement with child protective services
and longer stays in foster care.106 Another study
44
examining state-level data found that as cash assistance benefit levels decreased, neglect and out-ofhome care increased.107 The study also found that as
the share of working single mothers increased, so did
the rates of neglect. In an evaluation of Delawares
AFDC waiver program, families that participated in
the program (with features similar to TANF requirements) were found to have higher rates of child neglect than did those in the control group.108 Finally,
though not identifying overall impacts on maltreatment rates, case studies of child welfare agencies in 12
states noted that reports of lack of supervision had
increased under TANF, as had the number of families
surrendering their children to child welfare agencies
or delaying reunification of children already in care.109
The findings from these studies do not provide conclusive evidence that TANF provisions are leading to
increased child maltreatment, but they do raise red
flags about potential problems for some families.
Funding for Services
TANF may also impact child welfare by increasing or
decreasing the funding available for services. The
1996 law changed the structure of two significant
funding streams, potentially reducing the amount of
federal dollars spent on child welfare. The Emergency
Assistance program, which accounted for 13% of
1996 federal child welfare dollars, was repealed and
consolidated into the TANF block grant; and funding
for the Social Services Block Grant (Title XX), which
had accounted for 16% of 1996 federal child welfare
dollars, was reduced.110 At the same time, however,
the 1996 law allows states to use TANF funds for certain child welfare services.111
Many states are spending their TANF and maintenance-of-effort (MOE) funds to provide services that
could be characterized as child welfare services, such
as nonmedical substance abuse treatment, home visiting programs, parenting education classes, and subsidized guardianship.112 Sometimes these services are
offered to families receiving TANF cash assistance,
sometimes to families in the child welfare system, and
Volume 12, Number 1
sometimes more broadly to families in need. One

illustration of a restructuring of funding and services
can be found in El Paso County, Colorado, where the
local government decided to unite its TANF and
child welfare programs, treating TANF as the primary
prevention program for child welfare and treating
child welfare as an antipoverty program.113 (See the
appendix following the article by Chase-Lansdale and
Pittman in this journal issue for a more detailed
description of the El Paso County program.)
Some policymakers are concerned that states may be
spending too many TANF dollars on child welfare
services and not enough on other supports for lowincome families. Others are concerned that TANF
child welfare spending may be merely replacing state
funds with federal funds. In addition, some child
advocates are concerned that states may be using
TANF and MOE funds for child welfare services
without adhering to procedural safeguards mandated
by Congress for child welfare programs. At this point,

state-reported TANF financial data indicate only
broad categories of spending and do not specify how
much states are spending on child welfare, or the
types of services provided.
Supports for Kin Caregivers
It is also unclear whether changes under TANF are
increasing or decreasing the supports available for the
growing number of children living with grandparents or
other relatives. In 2000, more than 2.1 million children
were living with relatives with no parent present.114
Many relatives care for these children without specific
federal or state assistance, but significant numbers of kin
caregivers rely on TANF or child welfare services.
In 1999, approximately 500,000 children who
received TANF lived in households headed by relatives. Recognizing the unique needs of these kin caregivers, some states have developed special provisions
or programs for them.115 In most states TANF pro-
Box 9
Potential Child WelfareRelated Reauthorization Issues
(1) Should the child welfare system be more integrated with
TANF? Many families being served by child welfare and TANF
have similar needs, and often move between the two systems.
Recognizing these connections, policymakers are likely to
consider whether and how to better coordinate and integrate
services to these families. For example, what are the best
ways to address substance abuse, domestic violence, mental
and physical disorders, literacy deficits, and other barriers to
both employment and adequate parenting? What structures
are needed to facilitate better cooperation between the two
systems? What, if any, safeguards are needed to protect families when these two systems join forces?
(2) Should the fiscal structure of the child welfare system be
changed? Are more incentives needed for providing prevention and reunification services? With states spending TANF and
MOE funds on a variety of child welfare services, questions
about the best way to finance the child welfare system will
likely arise during TANF reauthorization. Many child welfare
specialists are concerned that federal funding for foster care
is open-ended, whereas federal funding for prevention and
reunification services is limited. They argue that these funding
policies encourage out-of-home placements and discourage
states from developing the capacity to provide services that
might make many out-of-home placements unnecessary.
(3) To what extent should federal assistance be used to support
children living with kin? Currently, kin may be able to receive
support when caring for relative children through TANF or the
child welfare system, depending on the circumstances. Each
system has different requirements and different levels of
assistance. What level of support is appropriate? What types of
supports do kin caregivers need? What types of supports do
the children need?
45
Under the new definition of childhood disability, an estimated

100,000 children lost their eligibility for SSI.
grams, however, relatives must adhere to the same
requirements as parents. Relative caregivers who meet
the states income criteria may choose to receive a
TANF grant for themselves and the children, but they
are then subject to the federal time limit and to work
and participation requirements. These requirements
do not apply if relatives receive a child-only TANF
grant, which is based solely on the childs needs and
income.116 However, such grants are generally smaller than those received when the relative is included,
and may be insufficient to allow the relative to provide adequate care for the child. Nonetheless, more
than four-fifths of the children living with kin who
received any TANF assistance in 1999 received childonly grants.
Relative caregivers can also seek assistance from the
formal child welfare system and receive a foster care
maintenance payment.117 Many advocates argue that
such a result is not good for children in some cases,
because the child welfare system can be intrusive, and
the level of supervision and oversight the system provides is unnecessary for many kin caregivers. In addition, some relatives may not be able to turn to the
child welfare system. This option is available only
when children are at risk of abuse or neglect, and not
all relatives are caring for kin under these circumstances. Moreover, some relatives will not meet state
licensing standards for a foster care placement, such
as the square footage requirements for their homes.
As the 107th Congress considers TANF reauthorization, child welfare issues are likely to be considered as
well. Questions about how to finance the child welfare system and improve coordination with TANF will
likely be discussed, as summarized in Box 9.
Supplemental Security Income

The Supplemental Security Income (SSI) program,
administered by the Social Security Administration
(SSA), provides income assistance to low-income
people who are elderly, blind, or disabled, including
children. In 1990, a Supreme Court decision (Sullivan v. Zebley) held that SSAs test for determining
46
whether children were disabled (and thereby entitled to SSI assistance) was unlawfully restrictive. Following this decision, the number of children in the
program grew dramatically, from approximately
300,000 in 1989 to about 1 million in 1996.118 As
the numbers swelled, the program came under attack.
A series of press and television stories alleged that
children were being coached to misbehave in order
to get crazy checks.119
Although subsequent investigation revealed no evidence of widespread abuse of the program,120 Congress took up the issue during the welfare reform
debates in 1996. Disagreements arose concerning the
criteria for determining childrens eligibility for SSI,
the meaning of disabled, and the degree of impairment necessary to justify assistance. Consequently,
the 1996 law made a number of changes to SSI.
Specifically, the law:
Modified the definition of childhood disability,
requiring that a child have an impairment that
results in marked and severe functional limitations in order to be eligible for SSI benefits;
Required that SSA use the new definition to redetermine the eligibility of children already receiving benefits, if it might lead to termination of their benefits;
Eliminated the medical improvement test for 18year-olds (which had allowed children to continue
to receive benefits after age 18 unless their condition had medically improved to the extent that they
were no longer disabled), and instead required that
SSA use the adult criteria for disability to redetermine eligibility for children turning 18.
Under the new definition of childhood disability, an
estimated 100,000 children lost their eligibility for
SSI.121 In 1997, SSA issued interim regulations interpreting the phrase marked and severe functional limitations.122 Many critics contended that this
interpretation was unduly restrictive. Indeed, when
SSA issued the regulations, several U.S. senators who
were instrumental in drafting the SSI provisions of the
1996 law sent a letter to the president claiming that
this interpretation was inconsistent with their intent.123
Volume 12, Number 1
In September 2000, SSA issued revised final regulations.124 Although the basic interpretation of marked
and severe functional limitations did not change,
some advocates saw the regulations as an improvement, because they simplified and clarified the
process for determining childhood disability.125 Also,
some are encouraged that SSA has undertaken a project with the American Association of University-Affiliated Programs to examine ways they might improve
the evaluation of childhood disability claims. Advocates are hopeful that the process of determining disability in children will continue to improve as the
project progresses.
In addition to questions about the definition of childhood disability, another concern is that adolescents
who should continue to receive benefits are losing
them. SSA data indicate that between 1997 and 2000,
just over 90,000 adolescents (nearly half the SSI recipients who went through eligibility redeterminations at
age 18) lost their eligibility for SSI benefits.126 However, the number of adolescents who are expected to
remain ineligible once all appeals processes are

exhausted is expected to drop to about 70,000.127
Also, the proportion of adolescents losing their benefits appears to have decreased each year since 1997.
As adolescents lose SSI eligibility, many also lose access
to medical care through Medicaid.128 Advocates for disabled teens argue that applying adult eligibility requirements to children turning 18 creates an unfair burden
because the adult criteria are based partially on work
histories, which most teen SSI recipients do not have.
Advocates are also concerned that eliminating the medical improvement test for 18-year-olds treats them like
new applicants, with no transition period as provided in
other programs.129
Although the 1996 law amended the SSI program, it is
not clear whether SSI will be part of the discussions
about TANF reauthorization. No part of the SSI program requires reauthorization. Nonetheless, key issues
concerning children and SSI that may surface as TANF
legislation is considered are summarized in Box 10.
Box 10
Potential SSI-Related Reauthorization Issues
(1) Are poor children with severe disabilities receiving SSI benefits as intended? Although advocates consider the revised
regulations to be improved, some believe the Social Security
Administrations interpretation of the definition of childhood
disability is too restrictive. As a result, the 2002 discussion
could include an assessment of whether the current test of
childhood disability is too stringent, precluding benefits for
children Congress intended to cover.
(2) Should the medical improvement test for 18-year-olds be
reinstated? Some advocates argue that adolescents should
lose coverage only when their condition has improved to the
extent that they are no longer disabled. As a result, the 2002
discussions could include attempts to reinstate the medical
improvement test for 18-year-olds.
(3) Should 18-year-olds be provided transitional assistance? If

the 2002 debates focus on issues related to helping people
with various challenges move toward greater independence,
discussion may arise about the need to provide disabled children with some type of transitional assistance if they no longer
qualify under the adult criteria now imposed when they turn 18.
(4) Should Medicaid coverage be extended to all children with
impairments, regardless of eligibility for SSI? Although many
children with impairments may qualify for Medicaid based on
their families incomes, a number will be barred from coverage
because their families incomes exceed the states income
requirements. Expanding coverage to all children with disabilites would provide increased access to the services and
treatment children may need to maximize their medical and
functional improvement.
47
Immigrant eligibility for public assistance programs as laid out

in the 1996 law is a complicated patchwork of federal eligibility
rules, state discretionary choices, and statutory exceptions.
Immigrants
Before the 1996 welfare reform law was enacted, legal
immigrants and their children were generally eligible
for public benefits under the same terms as citizens,
and states did not have discretion to develop their own
rules for determining immigrants eligibility for public
assistance. During the 1996 debates, however, immigrant provisions became a focus of discussion, driven
by both policy and fiscal concerns. The policy dispute
centered on whether immigrants access to public benefits should be curtailed in order to discourage people
from immigrating to the United States just to gain
access to public benefits.130 The fiscal issue was driven
largely by congressional interest in reducing government spending. Nearly half of the laws projected savings were attributable to the provisions making most
legal immigrants ineligible for public benefits.131
Immigrant eligibility for public assistance programs as
laid out in the 1996 law is a complicated patchwork of
federal eligibility rules, state discretionary choices, and
statutory exceptions.132 Key provisions of the law:
Distinguished between qualified and not qualified
(though often legal) immigrants, and between persons
who entered the United States before and after enactment of the 1996 law, in determining eligibility;133
Banned not qualified immigrants from TANF
and Medicaid assistance (except for emergency
Medicaid services), and generally made qualified
immigrants ineligible for these programs for five
years after coming to the United States, after which
eligibility is a state option;134
Made most legal immigrantsboth qualified and
not qualifiedineligible for food stamps and SSI
until they attained citizenship;
Made undocumented immigrants and other not
qualified immigrants ineligible for most state and
local public benefits, but allowed states to develop
their own policies concerning the eligibility of
qualified immigrants; and
48
Allowed exceptions to these requirements only for

refugees, asylees, persons granted withholding of
removal during their first five years (subsequently
extended to seven years) in the United States,
immigrants who meet the 40-quarter work history
test, and current and former military personnel and
their spouses and dependents.
Current Provisions
Since 1996, Congress has restored eligibility for limited
categories of immigrants,135 though most legal immigrants remain ineligible for food stamps and almost all
immigrants entering the country after enactment of the
1996 law are ineligible for a wide array of federal benefits during their first five years in the United States.
TANF and Medicaid
Nearly every state opted to provide Medicaid and
TANF assistance to all preenactment qualified legal
immigrants.136 However, with limited exceptions, legal
immigrants entering the United States on or after the
date the law took effect are ineligible for federally funded TANF assistance and Medicaid for five years. States
can choose to make these new immigrants ineligible for
TANF and Medicaid beyond the first five years, but
most states have not elected to do so.
Food Stamps
Of the 1.4 million legal immigrants receiving food
stamps in 1996, an estimated 940,000 recipients lost
eligibility when the law was implemented. In 1998,
Congress enacted a limited restoration primarily benefiting immigrant children, the elderly, and people
with disabilities.137 This restoration affected only
about 250,000 recipients. Thus, among immigrants
living in the United States when the law was enacted,
more than two-thirds of those who lost eligibility
remain ineligible. Among those entering the United
States after the law was enacted, nearly all are ineligible for food stamps until they attain citizenship.
SSI
The 1996 law, as enacted, would have resulted in an estimated 580,000 elderly and disabled immigrants losing
Volume 12, Number 1
their eligibility for SSI in 1997. Before the terminations

were due to take effect, Congress acted to retain benefits for most elderly and disabled immigrants who were
residing in the United States when the law was enacted,138 reversing about half of the projected savings from
restriction of immigrant benefits.139 With limited exceptions, however, immigrants entering the United States
after the 1996 law was enacted are ineligible for SSI.
States Response
Some states have responded to federal restrictions on
immigrant benefits by establishing state-funded substitute programs, but these programs have not filled the
gap left from the loss of federal assistance. An Urban
Institute study found that, as of May 1999, more than
half (28) of all states had created at least one substitute
program for immigrants who lost their eligibility for
federal assistance under TANF, Medicaid, food stamps,
or SSI.140 Many states with substitute programs did not
extend benefits to all legal immigrants who lost federal
eligibility, however, or to postenactment immigrants
during their five-year federal ineligibility period.141
Moreover, participation rates in these substitute programs remain low.
Low Participation by Eligible Immigrant Children

The receipt of public benefits by immigrant children
was low even before welfare reform,142 and has fallen
even lower as a result of restrictions on benefits and
the decline in participation among immigrant families. In fact, since 1994, participation rates among eligible immigrants have declined across all benefit
programs.143 After the 1996 law was enacted, participation rates continued to drop, even though the need
among the children of immigrants is well documented, and most of these children are themselves U.S.
citizens. For example, a recent study found that 37%
of children of immigrants live in families that worry
about or have difficulty obtaining food, compared
with 27% of children of nonimmigrants, and that 22%
of children of immigrants do not have health insurance, compared with 10% of children of nonimmigrants.144
The evidence suggests that the decline in participation among eligible immigrants reflects a variety of
fears. For example, many eligible immigrants fear that
if they receive benefits, they may be considered a
public charge, which disqualifies them from spon-
Box 11
Potential Immigrant-Related Reauthorization Issues
(1) Should legal immigrants have their eligibility for public benefits restored? Efforts to restore immigrant eligibility for public benefits have continued in each Congress since enactment
of the 1996 law, and no doubt further efforts toward this goal
will be made.
(2) Should special efforts be made to increase access for citizen children in mixed-status households? There may be
greater attention paid to the areas of federal policy in which
ambiguity may be having an unintended chilling effect on
receipt of public benefits by eligible children who live with
ineligible parents and guardians.
(3) Are new policies needed to address language barriers for

immigrants and other limited-English speaking families?
Concerns have been raised about the adequacy of programs to
improve the foreign language skills of service workers, and
depending on local progress, the issue may be revisited during reauthorization. Also, within the TANF program, some
caseworkers have raised concerns about restrictions on
immigrants participation in English as a Second Language
(ESL) programs as a countable activity toward a states
required work participation rate. Some argue that to help nonEnglish-speaking families move toward self-sufficiency,
greater efforts must be made to assess English literacy needs
and provide access to programs that address those needs.
49
soring relatives who may want to immigrate, or that

they may be investigated by the Immigration and
Naturalization Service (INS).145 In mid-1999, INS
clarified that getting Medicaid would not affect public charge status under most circumstances,146 but it is
not yet clear whether this clarification has improved
program participation.
Another key reason for low participation rates is that
many immigrant households have mixed statusthat
is, they include members who are citizens and noncitizens, qualified and not qualified, eligible and
ineligible.147 As a result, a substantial number of children who are citizens and fully eligible for federal and
state public assistance may not be receiving needed
benefits because they live with a noncitizen parent or
grandparent who is ineligible for various assistance
programs under the 1996 law. According to the U.S.
Department of Agriculture, for example, participation in the food stamp program by citizen children in
families headed by a noncitizen dropped by 75%
between 1994 and 1998.148
Language barriers can make it even more difficult to
understand the complex rules of the welfare system,
further exacerbating immigrant families confusion
about eligibility.149 Although an executive order
issued in August 2000 sought to address concerns
about linguistic barriers to programs for non-English
speakers,150 it remains to be seen whether the guidance is implemented locally in ways that ensure meaningful access to programs. A number of organizations
whose members would implement the guidance, such
as the American Medical Association, opposed the
order because of its purported cost, and a bill that
would prohibit its implementation was introduced in
Congress in March 2001.151
In 2002, lawmakers will likely examine current policy
concerning public benefits for immigrants. Key questions that may be raised are summarized in Box 11.
Conclusion
The landscape of social policies and programs serving
low-income children has changed dramatically since
enactment of the 1996 law. The dominant themes in
1996 included reducing spending, promoting devo-
50
lution, discouraging receipt of cash assistance, promoting work, discouraging out-of-wedlock births,
and restricting assistance to immigrants. Since 1996,
cash assistance spending has fallen, and states have
used their flexibility under the law to create a wide
variety of programs tailored to reflect state choices.
Caseloads have declined dramatically, and employment has increased among female-headed families.
Fewer children are being born to unmarried teen
mothers, and the share of children born out of wedlock appears to have leveled off. Meanwhile, immigrant participation in public benefits programs has
decreased markedly. Although not all of these
changes are directly or solely attributable to the 1996
law, notable progress has been achieved in addressing
many goals of the legislation.
At the same time, welfare participation has fallen
much more rapidly than has child poverty. Most families leaving welfare are either entering low-paying
jobs or are not working. For those entering low-wage
jobs, the need for support services remains high, but
significant concerns have been raised about the difficulties in ensuring that low-earning families have
access to food stamps, Medicaid, and child care.
Often gains in income from employment are largely
or completely offset by losses in public benefits. Many
of the families remaining on the welfare rolls face difficult barriers to employment, similar to the problems
faced by many families in the child welfare system,
such as substance abuse, mental illness, and domestic
violence. Moreover, many of the families with the
most serious barriers to employment have become
ineligible for assistance or have left welfare without
work and have disappeared from public systems.
The 2002 congressional debates surrounding the
reauthorization of federal welfare reform are likely to
be an extraordinarily important time for discussions
of national poverty policy and family policy. It will be
a time to acknowledge where progress has and has
not been made, and to look for improvements for the
future. For each specific program, discussions will
center on cost, performance, effectiveness, and incentives for governments and individuals. Across programs, lawmakers will focus on how government can
better assist working poor families; what the next
steps should be in the national dialogue about mar-
Volume 12, Number 1
riage, out-of-wedlock births, fathers, and family formation; and how and where to strike the balance
between state discretion and federal responsibility. We
also hope that in 2002, child outcomes and wellbeing will be given much more explicit consideration
in the discussion about what has and has not changed
since 1996, and about how best to structure the next
chapter in social policy affecting low-income families.
We greatly appreciate comments and assistance provided by Anya Arax, Joshua Bernstein, Marty Ford, Sean
Hartigan, Cindy Mann, Dorothy Rosenbaum, David
Super, Eileen Sweeney, and Dinah Wiley. The authors
are solely responsible for the text.
ENDNOTES
1. The programs that were repealed and whose funding was folded
into the TANF block grant were the AFDC program, the Job
Opportunities and Basic Skills Training (JOBS) program, and the
Emergency Assistance program.
2. For an overview of allowable spending under TANF, see U.S.
Department of Health and Human Services. Helping families
achieve self-sufficiency: A guide on funding services for children
and families through the TANF program. Washington, DC:
DHHS, 1999.
3. For detailed information about state policies, see U.S. Department of Health and Human Services. Temporary Assistance for
Needy Families (TANF) program: Third annual report to Congress. Washington, DC: DHHS, August 2000. See also the Web
site of the State Policy Documentation Project (SPDP), a joint
project of the Center for Law and Social Policy and the Center
on Budget and Policy Priorities, at http://www.spdp.org.
4. Caseload data is available online at
http://www.acf.dhhs.gov/news/tables.htm.
5. See U.S. Department of Health and Human Services. Indicators
of welfare dependence: Annual report to Congress. Washington,
DC: DHHS, 2001, table TANF 2.
6. U.S. General Accounting Office. Welfare reform: State sanction
policies and number of families affected. GAO/HEHS-00-44.
7. Goldberg, H., and Schott, E. A compliance-oriented approach to
sanctions in state and county TANF programs. Washington, DC:
Center on Budget and Policy Priorities, October 2000.
Washington, DC: Center on Budget and Policy Priorities, 1999.
9. In 1997, some 48% of families receiving assistance indicated that

either their general health or mental health was poor. See
Loprest, P., and Zedlewski, S. Current and former welfare recipients: How do they differ? Washington, DC: Urban Institute, 1999.
10. Cash assistance expenditures fell from nearly $23 billion in 1994
to $11.5 billion in 2000.
11. For further discussion of this topic, see Greenberg, M. Beyond
welfare: New opportunities to use TANF to help low-income working families. Washington, DC: Center for Law and Social Policy,
1999. See also Lazere, E. Unspent TANF funds at the end of federal fiscal year 2000. Washington, DC: Center on Budget and
Policy Priorities, 2001.
12. U.S. General Accounting Office. Welfare reform: Challenges in
maintaining a federal-state fiscal partnership. GAO-01-828.
Washington, DC: GAO, August 2001.
13. In 1998, 55% of black children, 31% of Hispanic children, and
21% of white children lived with only one parent. See U.S. Census Bureau. Marital status and living arrangements: March 1996.
Washington, DC: U.S. Census Bureau, March 1998. See also
updated Census Bureau data. Federal Interagency Forum on
Child and Family Statistics. Americas children: Key national
indicators of well-being. Washington, DC: U.S. Government
Printing Office, 1999, tables, pp. 6970.
14. Sawhill, I. Welfare reform and reducing teen pregnancy. The Public Interest (Winter 2000) 138:4051.
15. Ventura, S.J., and Bachrach, C.A. Nonmarital childbearing in the
United States: 194099. National Vital Statistics Reports. Vol.
48, no. 16. Hyattsville, MD: National Center for Health Statistics, October 18, 2000, pp. 26.
16. Murray, C. The coming white underclass. Wall Street Journal.
October 29, 1993, at A14.
51
17. About 20 states established a family cap under an AFDC waiver

prior to 1996; 3 more did so by 2000. See Stark, S., and LevinEpstein, J. Excluded children: Family cap in a new era. Washington, DC: Center for Law and Social Policy, February 1999, p. 1.
18. Levin-Epstein, J. Teen parent provisions in the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. Washington, DC: Center for Law and Social Policy, 1996.
19. Ooms, T. Testimony before the Subcommittee on Human
Resources, Committee on Ways and Means, U.S. House of Representatives, Hearing on Welfare and Marriage Issues. Washington, DC. May 22, 2001.
20. The National Marriage Project. The state of our unions: The social
health of marriage in America, 1999. Piscataway, NJ: Rutgers, the
State University of New Jersey, 1999.
21. Ooms, T. Toward more perfect unions: Putting marriage on the
public agenda. Report presented at the Family Impact Seminar.
Washington, DC. July 1998. Available from tooms@clasp.org.
22. See Rector, R. Welfare: Broadening the reform. In Issues 2000:
Candidates briefing book. S.M. Butler and K.R. Holmes, eds.
Washington, DC: Heritage Foundation, 2000; Horn, W.F., and
Sawhill, I.V. Fathers, Marriage and Welfare Reform. In The new
world of welfare. R.M. Blank and R. Haskins, eds. Washington,
DC: Brookings Institution Press, 2001, pp. 42141.
23. Oklahomas $10 million is available for a range of activities
including public education, training of state employees to offer
relationship skills workshops, improving data and research, and
working with faith-based groups and community leaders.
24. See note 19, Ooms. See also Rubiner, L. Testimony before the
Subcommittee on Human Resources, Committee on Ways and
Means, U.S. House of Representatives, Hearing on Welfare and
Marriage Issues. Washington, DC. May 22, 2001.
25. For example, the Fathers Count bill, introduced in 1999, awarded grants to community-based organizations that, among other
activities, promoted marriage as a means of ensuring responsible,
involved fatherhood. This bill passed the House overwhelmingly
in 2000, but stalled in the Senate and was reintroduced in 2001.
26. Personal Responsibility and Work Opportunity Reconciliation
Act of 1996 (PRWORA). Public Law 104-193, title IX, section
905. The law allows spending of federal TANF funds on family
planning, but not on abortion and other medical services.
27. Hutson, R.Q., and Levin-Epstein, J. Linking family planning
with other social services: The perspectives of state family planning
administrators. Washington, DC: Center for Law and Social Policy, 2000; and Cohen, M. Tapping TANF: When and how welfare
funds can support reproductive health or teen parent initiatives.
Washington, DC: Center for Law and Social Policy, April 1999.
28. See note 26, PRWORA, title IX, section 912. The program is
administered through the Maternal Child Health Services Block
Grant.
29. Daley, D., and Wong, V.C. Between the lines: States implementation of the federal governments section 510(b) abstinence education
program in fiscal year 1998. New York: Sexuality Information and
Education Council of the United States, 1999.
30. Congress approved $20 million for fiscal year 2001 and a total of
$40 million for fiscal year 2002, under Special Projects of
Regional and National Significance; Community-Based Abstinence Education Project Grants.
31. In 1999, teen parents accounted for about 4% of all adults in
the caseload, and another 0.4% of all children. See note 3. The
52
teen parent status is unknown for a significant percentage of

the caseload5.2% of the children and 0.5% of adults, or nearly
287,000 teen recipients.
32. See note 26, PRWORA, title I, section 103, 110 Stat. 213537.
The law also called for the U.S. Department of Justice to study
the link between statutory rape and teen pregnancy, and to educate state and local criminal law enforcement officials on statutory rape prevention and prosecution (title IX, section 906).
Because these provisions did not come with funding, however,
communities have done little to implement such programs.
33. Matthews, M., and Shelley, S. Turned away, misinformed,
deniedTeen parents experiences in welfare offices. Youth Law
News (JulyAugust 1999) 20(4):7,12.
34. Kirby, D. Emerging answers: Research findings on programs to
reduce teen pregnancy. Washington, DC: National Campaign to
Prevent Teen Pregnancy, 2001. See also Jemmott, J., Jemmott,
L., and Fong, G. Abstinence and safer sex HIV risk-reduction
interventions for African American adolescents: A randomized
controlled trial. Journal of the American Medical Association
(May 20, 1998) 279:152936.
35. See Bos, J.M., and Fellerath, V. LEAP: Final report on Ohios
welfare initiative to improve school attendance among teenage parents. New York: Manpower Demonstration Research Corp.,
1997; Wisconsin Legislative Audit Bureau. Wisconsins Learnfare
program. Madison: Wisconsin Legislative Audit Bureau, 1996;
Mauldon, J., Malvin, J., Stiles, J., et al. Impact of Californias
Cal-Learn program demonstration project: Final report. Berkeley:
UC Data, University of California, Berkeley, 2000.
36. The evaluation of Californias learnfare program suggests that
teen mothers who are married or have steady boyfriends are at
greater risk of a subsequent birth. See note 35, Mauldon, et al.
37. Levin-Epstein, J. Seeking supervision: State policy choices in implementing the TANF minor parent living arrangement rule. Washington, DC: Center for Law and Social Policy, 1999.
38. The percentage is not known exactly, because 15 states have separate two-parent programs and do not report their data to the
federal government. See note 31, U.S. Department of Health
and Human Services, table 10:7.
39. U.S. Department of Health and Human Services. HHS awards
welfare high performance bonuses: Higher increases in job placements and earnings reported. Press release. Washington, DC:
DHHS, December 16, 2000.
40. The 1999 illegitimacy awards went to Alabama, California, the
District of Columbia, Michigan, and Massachusetts. See Romero,
D., Chavkin, W., Wise, P.H., et al. State welfare reform policies
and maternal and child health services: A national study. Maternal and Child Health Journal (2001) 5(3):199206.
41. One study estimated that child support reduces poverty among
children by 5%. See Sorensen, E., and Zibman, C. Child support
offers some protection against poverty. Washington, DC: Urban
Institute, 2000; and Garfinkel, I., and Heinz, T. Child support
enforcement: Incentives and well-being. Paper presented at the
Conference on Incentive Effects of Tax and Transfer Policies.
Washington, DC. December 8, 2000.
42. Meyer, D. Child support and welfare dynamics: Evidence from
Wisconsin. Demography (1993) 30:4562; Seltzer, J. Child support
and child access: Experiences of divorced and nonmarital families.
Focus (2000) 21:5457; and Doolittle, F., Knox, V., Miller, C.,
and Rowser, S. Building opportunities, enforcing obligations. New
York: Manpower Demonstration Research Corp., 1998.
Volume 12, Number 1
43. For poor, single female-headed families that receive child support, it is the second largest component of family income after
earnings, amounting to 26% of the familys budget, or $2,000
per year. (Other components include earnings at 38%, cash assistance at 20%, and other income at 16%.) See Sorensen, E., and
Zibman, C. To what extent do children benefit from child support? New information from the National Survey of Americas
Families. Focus (2000) 21:3437.
44. U.S. Department of Health and Human Services. Child support
enforcement: FY 2000 preliminary data preview report. Washington, DC: DHHS, July 2001. Available online at
http://www.acf.dhhs.gov/programs/cse/prgrpt.htm. See also
U.S. General Accounting Office. Welfare reform: Child support
an uncertain income supplement for families leaving welfare.
GAO/HEHS-98-168. Washington, DC: GAO, 1998.
45. Poor families eligible for child support are significantly more likely to receive it if they have participated in the child support program. Sorensen E., and Halpern A. Child support enforcement is
working better than we think. Washington, DC: Urban Institute,
2000. See also note 41, Garfinkel and Heinz.
46. Although 40% of children with a nonresident parent are living
below the poverty level, 23% of nonresident fathers are poor. See
Sorensen, E., and Zibman, C. A look at poor dads who dont pay
child support. Washington, DC: Urban Institute, 2000.
47. See Lyon, M. Characteristics of families using Title IV-D services
in 1995. Washington, DC: U.S. Department of Health and
Human Services, 1999. See also note 44, U.S. Department of
Health and Human Services.
48. See the following papers available on the Center for Law and
Social Policy Web site at www.clasp.org: Roberts, P. Family law
issues and the Personal Responsibility and Work Opportunity Reconciliation Act of 1996; Roberts, P. A guide to establishing paternity for nonmarital children: Implementing the provisions of the
Personal Responsibility and Work Opportunity Reconciliation Act
of 1996; and Turetsky, V. Child support administrative processes: A
summary of requirements in the Personal Responsibility and Work
Opportunity Reconciliation Act of 1996 (1997).
49. States must meet five performance measures and have reliable
data to qualify for federal incentive payments, and these payments
must be reinvested in child supportrelated activities. New penalty, audit, and reporting requirements also were enacted. See
Roberts, P., and Turetsky, V. New federal child support legislation
on computer penalties, incentive payments, medical support, and
other topics. Washington, DC: Center for Law and Social Policy,
July 28, 1998.
50. Among families with a support order already in place, collection
rates increased from 34% to 68% between 1995 and 2000. See
note 44, U.S. Department of Health and Human Services, as
well as preceding annual reports.
51. Under AFDC, states were required to pass through the first
$50 of support to families receiving assistance, and the costs were
shared between federal and state governments. When TANF was
enacted, the pass-through requirement was repealed. States were
allowed to continue the practice, but with no federal cost sharing, and less than half the states have opted to do so. See Turetsky, V. What if all the money came home? Washington, DC: Center
for Law and Social Policy, June 2000; and Roberts, P. State policy
re: pass-through and disregard of current months child support collected for families receiving TANF-funded cash assistance. Washington, DC: Center for Law and Social Policy, January 1999.
52. See note 46, Sorensen and Zibman, p. 3; and Sorensen, E. Lowincome families and child support: Latest evidence from the
National Survey of Americas Families. Washington, DC: Urban

Institute, 2000, figure 1.
53. Program administrators say the rules have contributed to computer systems delays, increased staff and training costs, difficulties
in redirecting support payments to former TANF families, confusion among parents, and audit problems.
54. Meyer, D., and Cancian, M. W-2 child support demonstration
evaluation, phase 1: Final report. Madison: University of WisconsinMadison, Institute for Research on Poverty, April 2001.
55. See Fishman, M., Tapogna, J., Dybdal, K., and Laud, S. Preliminary assessment of the association between state child support
enforcement performance and financing structure. Washington,
DC: U.S. Department of Health and Human Services, August
2000. See also Freeman, R., and Waldfogel, J. Dunning delinquent dads: Child support enforcement policy and never-married
women. Focus (Spring 2000) 21:2730; and Turetsky, V. You get
what you pay for: How federal and state investment decisions affect
child support performance. Washington, DC: Center for Law and
Social Policy, December 1998.
56. The federal government reimburses states 66% of their program
costs through an open-ended entitlement funding stream, and
states fund their matching share of costs with some combination
of state general or special funds, assigned welfare collections, and
federal incentive payments. See Fishman, M., Dybdal, K., and
Tapogna, J. State financing of child support enforcement programs.
Final report to Washington, DC: U.S. Department of Health and
Human Services, September 1999.
57. U.S. Department of Health and Human Services. 21 million childrens health: Our shared responsibilityThe medical child support
working groups report to Congress. Washington, DC: DHHS,
August 2000.
58. Martinson, K. Serving non-custodial parents through welfare-towork grants: Labor market characteristics, employment barriers, and
service strategies. Washington, DC: Urban Institute, October 1998.
59. For example, attention has been focused on informal support,
imputing income when setting orders, arrearages, and welfare
cost recovery policies. For a discussion of these issues, see Turetsky, V. Realistic child support policies for low income fathers. Washington, DC: Center for Law and Social Policy, March 2000.
60. Pearson, J., Thoennes, N., and Griswold, E. Child support and
domestic violence: The victims speak out. Violence Against
Women (1999) 5:42748; Brandwein, R., ed. Battered women,
children, and welfare reform: The ties that bind. New York: Sage
Publications, 1999; and Turetsky, V., and Notar, S. Models for
safe child support enforcement. Washington, DC: Center for Law
and Social Policy, October 1999.
61. The law also permits states to set lower state income eligibility
levels and to make age and income exceptions for children in
protective services.
62. For details, see Blank, H. Helping parents work and children succeed: A guide to child care and the 1996 welfare act. Washington,
DC: Childrens Defense Fund, 1997; and Greenberg, M. A summary of key child care provisions of H.R. 3734. Washington, DC:
Center for Law and Social Policy, 1996.
63. U.S. Department of Health and Human Services. Fiscal year
2000 state spending under the Child Care and Development Fund.
Washington, DC: DHHS, Child Care Bureau, December 17,
2001 update. Available online at http://www.acf.dhhs.gov/
programs/ccb/research/00acf696/overview.htm.
64. Calculations based on figures from U.S. Department of Health
53
and Human Services. New statistics show only small percentage of

eligible families receive child care help. Press release. Washington,
DC: DHHS, December 6, 2000; and U.S. Department of Health
and Human Services. State spending under the Child Care Block
Grant. Fact sheet. Washington, DC: DHHS, November 12,
1998.
65. See Schumacher, R., Greenberg, M., and Duffy, J. The impact of
TANF funding on state child care subsidy programs. Washington,
DC: Center for Law and Social Policy, September 2001.
66. See U.S. Department of Health and Human Services, Administration for Children and Families. Access to child care for lowincome working families. Washington, DC: DHHS, October
1999. See also Collins, A., Layzer, J., Kreader, J., et al. National
study of child care for low-income families: State and community
substudy. Interim report. Prepared by the National Center for
Children in Poverty and Abt Associates for the U.S. Department
of Health and Human Services. Washington, DC: DHHS,
November 2000, p. 36.
67. The Childrens Defense Fund. A fragile foundation: State child
care assistance policies. Washington, DC: CDF, 2001.
68. Schumacher, R., and Greenberg, M. Child care after leaving welfare: Early evidence from state studies. Washington, DC: Center
for Law and Social Policy, October 1999.
69. Adams, G., Snyder, K., and Sandfort, J. Getting and retaining
child care subsidies: How policy and practice influence parents
experiences. Washington, DC: Urban Institute, forthcoming.
70. See Gish, M. State programs under the Child Care and Development Fund. Washington, DC: Congressional Research Service,
March 2001. See also the State Policy Documentation Project
Web site at http://www.spdp.org, Child care provision after leaving cash assistance; American Public Human Services Association.
Child care at three: Survey of state program changes. Washington,
DC: APHSA, 1999, p. 2; and Blank, H., Behr, A., and Schulman,
K. State developments in child care, early education and school-age
care 2000. Washington, DC: Childrens Defense Fund, 2001.
72. To encourage certain types of providers to participate in the system, however, more than 20 states have implemented differential
reimbursement rates. For example, some states pay higher rates
for nationally accredited programs or programs that serve children during nontraditional hours or that serve children with special needs. See note 70, Blank, Behr, and Schulman.
73. For a brief overview of the Medicaid program and populations
served, see http://hcfa.hhs.gov/pubforms/actuary/ormedmed/
DEFAULT4.htm; and http://hcfa.hhs.gov/stats/hstats99/
blusta99.htm, tables 11, 26, and 34.
74. For a summary of other Medicaid-related changes under the
1996 law, see Center for Budget and Policy Priorities. An analysis of the AFDC-related Medicaid provisions in the new welfare law.
Washington, DC: CBPP, 1996.
75. Guyer, J., and Mann, C. Taking the next step: States can now
expand health coverage to low-income working parents through
Medicaid. Washington, DC: Center on Budget and Policy Priorities, 1998.
76. See Ku, L., and Bruen, B. The continuing decline in Medicaid coverage. New Federalism: Issues and Options for States, Brief no. A37. Washington, DC: Urban Institute, 1999, tables 1 and 2. See
also U.S. General Accounting Office. Medicaid enrollment: Amid
declines, state efforts to ensure coverage after welfare reform vary.
GAO/HEHS-99-163. Washington, DC: GAO, September 1999,
54
pp. 6, 1112; Dion, R.M., and Pavetti, L. Access to and participation in Medicaid and the food stamp program: A review of the
recent literature. Washington, DC: Mathematica Policy Research,
March 2000, pp. 1113, and Appendix B; and Families USA
Foundation. Losing health insurance: The unintended consequences
of welfare reform. Washington, DC: Families USA, 1999, p. 16.
77. For example, nearly 80% of parents of eligible uninsured children
mistakenly believed that Medicaid receipt is subject to time limits.
Perry, M., Kannel, S., Burciaga Valdez, R., and Chang, C. Medicaid and children: Overcoming barriers to enrollment: Findings
from a national survey. Washington, DC: Kaiser Commission on
Medicaid and the Uninsured, January 2000, p. 9. See also Stuber, J., Maloy, K., Rosenbaum, S., and Jones, K. Beyond stigma:
What barriers actually affect the decisions of low-income families to
enroll in Medicaid? Issue Brief. Washington, DC: George Washington University Medical Center, Center for Health Services
Research and Policy, June 2000.
78. See note 77, Stuber, et al.; Ellwood, M. The Medicaid eligibility
maze: Coverage expands, but enrollment problems persist. Washington, DC: Kaiser Commission on Medicaid and the Uninsured,
September 1999; Smith, V., Lovell, R., Peterson, K., and
OBrien, M. The dynamics of current Medicaid enrollment
changes: Insights from focus groups of state human service administrators, Medicaid eligibility specialists, and welfare agency analysts.
Washington, DC: Kaiser Commission on Medicaid and the Uninsured, 1998; and Maloy, K., Pavetti, L., Shin, P., and Darnell, J.
A description and assessment of state approaches to diversion programs and activities under welfare reform. George Washington
University Medical Center, Center for Health Services Research
and Policy, August 1998. See also note 76, U.S. General
Accounting Office, pp. 1824; and Dion and Pavetti, pp. 1315.
79. A review of state leavers studies found considerable differences
among states, but typically at least 20% of children and most
adults were no longer receiving Medicaid after leaving TANF. See
Greenberg, M. Participation in welfare and Medicaid enrollment.
Washington, DC: Kaiser Commission on Medicaid and the Uninsured, 1998. See also Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services.
Leavers and diversion studies: Summary of research on welfare
outcomes funded by ASPE. Washington, DC: DHHS, 2000; and
Loprest, P. Families who left welfare: Who are they and how are
they doing? Washington, DC: Urban Institute, 1999, chart 11.
80. See note 77, Perry, et al., p. 6.
81. Ellis, E., and Smith, V. Medicaid enrollment in 50 states: June
1997 to December 1999. Washington, DC: Kaiser Commission on
Medicaid and the Uninsured, October 2000, p. 18.
82. See Holahan, J. Why did the number of uninsured fall in 1999?
Policy brief. Washington, DC: Kaiser Commission on Medicaid
and the Uninsured, January 2001; and Guyer, J. Uninsured rate
of poor children declines, but remains above pre-welfare reform levels. Washington, DC: Center on Budget and Policy Priorities,
September 2000.
83. In April 2000, based on evidence that a number of states had
failed to correctly determine continued Medicaid eligibility for
families leaving TANF, the Health Care Financing Administration
(HCFA) directed all states to review their case closures and reinstate families whose Medicaid assistance had been terminated
erroneously. In addition, because children who receive SSI are
generally eligible for Medicaid, HCFA also directed all states to
obtain a list of children who had lost SSI and to ensure that they
were enrolled in Medicaid. See State Medicaid Directors Letter,
April 7, 2000. Available online at http://wwwqa.hcfa.gov/
medicaid/smd40700.htm.
Volume 12, Number 1
84. For a brief overview of the SCHIP program, see

http://www.hhs.gov/news/press/2001pres/01fsschip.html.
85. See note 81, Ellis and Smith, p. 6.
86. See note 82, Guyer, pp. 12.
87. U.S. Department of Agriculture. Characteristics of food stamp
households: Fiscal year 2000 (Advance Report). Alexandria, VA:
USDA, Food and Nutrition Service, Office of Analysis, Nutrition, and Evaluation, June 2001.
88. See note 87, U.S. Department of Agriculture, table 1.
89. For a general overview of food stamp program structure and rules,
see http://www.fns.usda.gov/fsp/MENU/about/about.htm.
For a more detailed discussion, see Food Research and Action
Center. FRACS guide to the Food Stamp program. 10th ed.
Washington, DC: FRAC, 1999.
90. The original Congressional Budget Office projections estimated
that $27.7 billion, or about half the original projected spending
reductions from the 1996 law, would be attributable to reductions in food stamp program eligibility and benefits, including
restrictions on immigrant eligibility. Center on Budget and Policy Priorities. The depth of the food stamp cuts in the final welfare
bill. Washington, DC: CBPP, August 14, 1996.
100. Although the connection between poverty and maltreatment is

not fully understood, the risk of abuse or neglect is 22 times
greater for children living in families with annual incomes below
$15,000 than for children living in families with incomes greater
than $30,000. Sedlak, A.J., and Broadhurst, D.D. Third national incidence study of child abuse and neglect: Final report. Washington, DC: U.S. Department of Health and Human Services,
1996, pp. 5-1 to 5-8. Also, though less than 3% of children
receiving AFDC during the study period moved into foster care,
nearly 60% of children in foster care came from families that
were, or recently had been, receiving AFDC. See Chapin Hall
Center for Children, University of Chicago. Dynamics of childrens movement among the AFDC, Medicaid and foster care programs prior to welfare reform: 19951996. Washington, DC: U.S.
Department of Health and Human Services, March 2000.
101. As per the technical corrections in the Balanced Budget Act of
1997, Public Law 105133.
102. Since the law was enacted, the poorest single-mother families
have experienced a decrease in disposable income. See note 8,
Primus, et al.
92. Castner, L. Trends in FSP participation rates: Focus on 1994 to

1998. Alexandria, VA: U.S. Department of Agriculture, Food
and Nutrition Service, November 2000, pp. 16, 23.
103. The 1996 law prohibits states from sanctioning single custodial
parents for failure to comply with work requirements if the parent demonstrates that she or he is unable to obtain needed child
care for a child under age six, but it is not clear if TANF recipients are aware of this exception. They may believe that they risk
losing cash assistance if they do not comply with the work
requirements. In addition, the exception does not apply to lack
of child care for children age 6 or olderyet a number of states
include failure to supervise children up to the age of 10 or 12
within the definition of neglect.
93. U.S. General Accounting Office. Food stamp program: Various

factors have led to declining participation. GAO/RCED-99-185.
Washington, DC: GAO, July 1999, p. 10.
104. U.S. Department of Health and Human Services. Child maltreatment 1999: Ten years of reporting. Washington, DC: DHHS,
2001.
94. U.S. Department of Agriculture. The decline in food stamp participation: A report to Congress. Alexandria, VA: USDA, Food
and Nutrition Service, Office of Analysis, Nutrition, and Evaluation, July 2001.

Washington, DC: U.S. Government Printing Office, 2001, p.
720. This apparent paradox may be due to children remaining in
foster care for longer periods of time. See Wulczyn, F.H., Brunner, K., and Goerge, R.M. An update from the multistate foster
care data archive: Foster care dynamics 19831997. Chicago:
Chapin Hall Center for Children, University of Chicago, 1999.
91. Based on preliminary data from the U.S. Department of Agriculture. For a summary of monthly participation numbers, see
http://www.fns.usda.gov/pd/fsmonthly.htm.
95. It has been estimated that 65% of former welfare families that left
the food stamp program still had incomes below food stamp eligibility standards. See Zedlewski, S.R., and Brauner, S. Declines
in food stamp and welfare participation: Is there a connection?
Assessing the New Federalism, Discussion paper 99-13. Washington, DC: Urban Institute, 1999, pp. 2021. See also note 79,
Loprest, chart 11 and table 4.
96. Rangarajan, A., and Gleason, P.M. Food stamp leavers in Illinois:
How are they doing two years later? Washington, DC: U.S.
Department of Agriculture, Economic Research Service, January
2001, pp. 6870.
97. Zedlewski, S., and Gruber, A. Former welfare families continue to
leave the food stamp program. Washington, DC: Urban Institute,
2001. See also note 93, U.S. General Accounting Office,
pp. 1314.
98. In early 1999, the Department of Agriculture wrote to all states
stressing the importance of complying with food stamp protections in determining food stamp eligibility and benefits when
evaluating TANF applications and closing TANF cases. See
http://www.fns.usda.gov/fsp.
99. For an overview of the evolution of the foster care program from
AFDC, see Frame L. Suitable homes revisited: An historical look
at child protection and welfare reform. Children and Youth Services Review (1999) 21(910):71954.
106. See Shook, K. Does the loss of welfare income increase the risk
of involvement with the child welfare system? Children and
Youth Services Review (1999) 21(910):781814. See also Wells,
K., and Guo, S. The impact of welfare reform on foster care and
child welfare: A case study: Reunification of foster children in the
first entry cohort. Paper presented at the 40th Annual Workshop
of the National Association for Welfare Research and Statistics.
Scottsdale, AZ. July 29August 2, 2000.
107. Paxson, C., and Waldfogel, J. Welfare reforms, family resources,
and child maltreatment. In The incentives of government programs
and the well-being of families. B. Meyer and G. Duncan, eds.
Chicago: Joint Center for Poverty Research, 2001, pp. 147.
108. Fein, D., and Lee, W. The ABC evaluation: Impacts of welfare
reform on child maltreatment. Bethesda, MD: Abt Associates,
2000.
109. Geen, R., Fender, L., Leos-Urbel, J., and Markowitz, T. Welfare
reforms effect on child welfare caseloads. Washington, DC: Urban
Institute, February 2001.
110. Geen, R., Waters Boots, S., and Tumlin, K. The cost of protecting
55
vulnerable children: Understanding federal, state, and local child

welfare spending. Washington, DC: Urban Institute, 1999.
111. For example, under the 1996 law, states may transfer part of
their TANF funds to Title XX, subject to certain limits. Specifically, no more than 10% can be transferred to Title XX for services to children and their families whose income is below 200% of
poverty.
112. See note 27, Cohen; and Christian, S., and Ekman, L. A place to
call home: Adoption and guardianship for children in foster care.
Denver, CO: National Conference of State Legislatures, March
2000.
113. Berns, D., and Drake, B. Uniting child welfare and welfare
reform. Childrens Voice (Fall 1999) 8(5):2023; and Berns, D.,
and Drake, B.J. Combining child welfare and welfare reform at a
local level. Policy and practice (March 1999) 57(1):2634.
114. U.S. Census Bureau. Americas families and living arrangements.
March 2000 Current Population Survey Report. Washington,
DC: U.S. Census Bureau, June 2001, detailed table no. C2.
115. See note 3, SPDP Web site. Separate state programs and stateonly TANF funds (June 2000). See also note 112, Christian and
Ekman; and Generations United. Grandparents and other relatives raising children fact sheet: Subsidized guardianship programs.
Washington, DC: Generations United, June 2001.
116. Although child-only cases are a steadily increasing share of the
TANF caseload, representing nearly 30% of the TANF caseload
in 1999, this is largely because other types of cases are declining
more rapidly. In addition, children living with relatives make up
only a small fraction of the child-only cases. The majority of
these cases involve children living with parents who, for a variety
of reasons, are ineligible for TANF assistance. See note 3, U.S.
Department of Health and Human Services.
117. Nearly 30% of children in foster care live with relatives. U.S.
Department of Health and Human Services, Administration for
Children and Families, Childrens Bureau. Report to Congress on
kinship foster care. Washington, DC: DHHS, June 2000, p. 7.
Washington, DC: U.S. Government Printing Office, 1998, p.
298. In addition to a new assessment tool (the Individualized
Functional Assessment, or IFA) to determine childhood disability under the Zebley decision, other factors contributed significantly to the increased caseload, including: (1) creation of an
outreach program; (2) expansion of the category of mental disorders considered disabling; and (3) the recession of the early
1990s, which is thought to have resulted in more children meeting the income criteria of the SSI program. See Mashaw, J.L.,
Perrin, J.M., and Reno, V., eds. Restructuring the SSI disability
program for children and adolescents: Report of the Committee on
Childhood Disability of the Disability Policy Panel. Washington,
DC: National Academy of Social Insurance, 1996.
119. For an overview of this coverage, see Georges, C. A media crusade gone haywire. Forbes MediaCritic (September 1995) 3:1,
6671.
120. See note 118, Mashaw, Perrin, and Reno, p. 1.
121. See U.S. General Accounting Office. Supplemental Security
Income: Progress made in implementing welfare reform changes;
more action needed. GAO-HEHS-99-103. Washington, DC:
GAO, 1999.
122. SSAs interpretation defined childhood disability as an impair-
56
ment that causes marked limitations in two broad areas of functioning or extreme limitations in one such area. See Part IV in
Social Security Administration. Federal Register (February 11,
1997) 62(28):640832, and section 416.925, in particular. See
also U.S. General Accounting Office. Supplemental Security
Income: SSA needs a uniform standard for assessing childhood disability. GAO/HEHS-98-123. Washington, DC: GAO, May
1998.
123. The letter says, ...a large percentage of [the children expected
to] lose assistance based on the SSAs definition of disability will
be disabled children who are truly in need of assistance.The
SSA is proposing to define marked and severe as meaning listings levels severity or any equivalent level of severity. Congress
never intended and did not require this level of severity. Senators Kent Conrad, Edward Kennedy, John D. Rockefeller IV,
Max Baucus, Christopher Dodd, John Chafee, Tom Harkin,
James Jeffords, Patrick Leahy, and Tom Daschle. Letter to President William J. Clinton, dated April 14, 1997.
124. See the Code of Federal Regulations, Title 20: Employees benefits, Chapter 11, Social Security Administration. Part 404: Federal old-age, survivors and disability insurance; and 416:
Supplemental security income for the aged, blind, and disabled.
125. For a discussion of the differences in the final regulations, see
Yates, T. New rules, same standard: The Social Security Administration adopts new rules for evaluating SSI childhood disability.
Journal of Poverty, Law, and Policy (MayJune 2001), 35(1/2):
5978.
126. U.S. Social Security Administration. 2001 annual report of the Supplemental Security Income Program. Baltimore, MD: SSA, Office
of the Chief Actuary, May 2001, table V.D3. Available online at
http://www.ssa.gov/OACT/SSIR/SSI01/Redet_CDRdata.html.
127. The total number of redeterminations for the years 1997
through 2000 was 191,049. Of those, 90,899 (48%) were initially ceased. On reconsideration or appeal a number of these cessations were continued so that, by the close of 2000, the number
of cessations was 74,066 or 39% of the initial redeterminations.
However 6,653 of the cases were still pending, so the final cessation rates could be lower, as low as 35%. The office of Chief
Actuary at SSA estimates, based on past experience and trends,
that the final cessation rate for these years will likely be about
37%. Authors personal conversation.
128. About 80% of SSI recipients are automatically eligible for Medicaid because of their eligibility for SSI. See note 105, U.S. House
of Representatives, Committee on Ways and Means, p. 225.
Thus, unless these teens meet other Medicaid eligibility criteria,
they will lose their coverage.
129. For example, under the Individuals with Disabilities Education
Act (IDEA), adolescents are permitted to receive special education services until age 22. In addition, under the federal foster
care program, Congress recently made changes to help children
transition to independent living, rather than simply cutting off
support at age 18.
130. Much of the congressional concern stemmed from the fact that
noncitizens use public benefits at higher levels than citizens. For
further discussion of this topic, see note 118, U.S. House of
Representatives, Committee on Ways and Means, Appendix J.
Recent analysis notes that noncitizen households tend to have
lower household income and more children than citizen households have. When controlling for poverty and number of children, however, noncitizens actually use public benefits at levels
equal to or lower than those of citizens. See Fix, M., and Passel,
J.S. Trends in noncitizens and citizens use of public benefits fol-
Volume 12, Number 1
lowing welfare reform: 199497. Washington, DC: Urban Institute, March 1999.
131. The Congressional Budget Office estimated that the immigrant
eligibility changes in the welfare law would save almost $23.7
billion over 6 years, accounting for about 44% of the $54.1 billion savings initially projected for the law. See Congressional
Budget Office. Federal budgetary implications of the Personal
Responsibility and Work Opportunity Reconciliation Act of 1996.
CBO Memorandum. Washington, DC: CBO, December 1996,
p. 27. The savings were significantly reduced when subsequent
legislation in 1997 and 1998 modified the immigrant provisions
of the 1996 law.
132. Zimmermann, W., and Tumlin, K.C. Patchwork policies: State
assistance for immigrants under welfare reform. Washington, DC:
Urban Institute, 1999, p. 15, figure 1.
133. Generally, qualified immigrants include persons admitted for
legal permanent residence, refugees, immigrants paroled into the
United States for at least one year, and immigrants granted asylum or related relief. See note 26, PRWORA, title IV, section
431. The 1996 immigration law added certain abused spouses
and children as another class of qualified immigrants, and the
1997 Balanced Budget Act added Cuban/Haitian entrants.
134. State options were subject to requirements to deem the
income of sponsors, however, whereby the income and resources
of an immigrants sponsor are counted as the immigrants when
making eligibility determinations. See note 132, Zimmermann
and Tumlin, p. 27. The sponsors affidavit of support is legally
enforceable and remains in effect until the sponsored immigrant
attains citizenship or accumulates 40 qualifying quarters of work.
See U.S. General Accounting Office. Welfare reform: Many states
continue some federal or state benefits for immigrants.
GAO/HEHS-98-132. Washington, DC: GAO, 1998, pp. 67.
135. The principal amendments to the original 1996 law include the
Illegal Immigration Reform and Immigrant Responsibility Act,
the Balanced Budget Act of 1997, the Agricultural Research,
Extension and Education Reform Act of 1998, and the Noncitizen Benefit Clarification and Other Technical Amendments Act
of 1998.
136. Only Alabama opted not to provide TANF and only Wyoming
elected to deny nonemergency Medicaid to preenactment immigrants. See note 132, Zimmermann and Tumlin, p. 60, table 25.
137. For a detailed discussion of the 1998 modifications, see Carmody, K., and Dean, S. New federal food stamp restoration for
legal immigrants: Implications and implementation issues. Washington, DC: Center on Budget and Policy Priorities, 1998.
138. Provisions were enacted to retain benefits for qualified immigrants who were receiving SSI when the 1996 law was enacted,
and to allow benefits for new applicants who were residing in the
United States when the law was enacted and who were (or subsequently became) disabled. A further modification continued
SSI for not qualified immigrants who were residing in the
United States and receiving SSI at the time of enactment, but
unlike qualified immigrants, not qualified immigrants who
subsequently become elderly or disabled will not qualify for SSI,
even if they were residing in the U.S. when the law was enacted.
139. Overall, the modifications under the Balanced Budget Act of

1997 were estimated to reverse $11.4 billion of the $23.8 billion
in savings from restrictions on immigrant benefits under the
1996 law. See National Conference of State Legislatures. Supplemental Security Income. Immigrant policy issue brief. Denver:
NCSL, 1998.
140. See note 132, Zimmermann and Tumlin, pp. 2223.
141. See note 121, U.S. General Accounting Office, p. 16.
142. See U.S. General Accounting Office. Medicaid: Demographics of
nonenrolled children suggest state outreach strategies.
GAO/HEHS-98-93. Washington, DC: GAO, 1998; and Schlosberg, C. Beyond outreach and informing: Getting serious about
covering uninsured children in immigrant families. Policy and
Practice (December 1998) 56(3):5463.
143. See note 130, Fix and Passel, p. 2. See also Fix, M. Comment on
welfare reform and immigration by George Borjas. Washington,
DC: Urban Institute, 2001, figure 1.
144. Capps, R. Hardship among children of immigrants: Findings from
the 1999 National Survey of Americas Families. Washington,
DC: Urban Institute, 2001, table 1 and figure 2.
145. See, for example, Zimmermann, W., and Fix, M. Declining
immigrant applications for Medi-Cal and welfare benefits in Los
Angeles County. Washington, DC: Urban Institute, 1998, pp. 2,
4, and table 1.
146. See, for example, U.S. Immigration and Naturalization Service.
A quick guide to public charge and receipt of public benefits:
Summary. Washington, DC: U.S. Department of Justice, INS,
October 18, 1999.
147. See note 132, Zimmermann and Tumlin, p. 26.
148. Cited in note 143, Fix and Passel, p. 2.
149. See, for example, Gordon, R. How welfare reform punishes poor
people. Oakland, CA: Applied Research Center, 2001.
150. Under Executive Order 13166, Improving Access to Services
for Persons with Limited English Proficiency, issued August 11,
2000, all federal agencies were directed to take reasonable steps
to ensure meaningful access to their programs and activities by
[limited English proficient] persons, based on the fact that not
serving this population because of their inability to speak English
may be a violation of Title VI of the Civil Rights Act. In
response, the Department of Health and Human Services (and
other federal agencies) issued guidance to help address the hurdles people not fluent in English were encountering when trying
to obtain services. Such help may include providing interpreters,
translations, and other assistance as needed. See Federal Register
(August 30, 2000) 65(169):52,76274.
151. The bill, H.R. 969, was introduced on March 8, 2001, in the
U.S. House of Representatives to nullify Executive Order 13166
and to prohibit use of funds under any provision of law to promulgate or enforce any executive order that creates an entitlement
to services provided in any language other than English. The bill
was referred to a House subcommittee on March 15, 2001, and
as of December 2001 no futher action had been taken.
57
Reforms and
Child Development
Aletha C. Huston
SUMMARY
Since their inception in the 1930s, U.S. welfare and income support programs have
played an important role in providing benefits to children. Unlike programs directly targeted to children, however, welfare
programs are designed to produce economic and employment impacts on adults, so
that any effects on children would be indirect. This article explores the influence of
such programs on childrens well-being and
development, first by defining goals for childrens healthy development, then by proposing a framework for understanding the
impact of welfare policies on children. A
review of the literature within each component of this framework reveals the following:
Measures of childrens well-being should
encompass physical, intellectual, social,
and emotional development.
Policies designed to increase maternal

employment, reduce welfare use, and
strengthen families do not necessarily lead
to more positive environmental contexts
and increased child well-being.
The author concludes that to ensure positive
impacts on children, welfare and income
support policies must move beyond their
exclusive emphasis on adults and include
goals that focus on improving childrens
social and physical environments at home
and in the community.
Aletha C. Huston, Ph.D., is Priscilla Pond Flawn

Regents Professor of Child Development in the
Department of Human Ecology at the University of
Texas, Austin.
The physical and material environment,

family environment, and social and community environment can all affect a childs
healthy development.
59
Huston
hen Aid to Dependent Children was

established in the 1930s, its principal
goal was to promote the welfare of
children. Under the assumptions of
the day, that meant assuring not only that basic physical needs were met (orphanages could do that), but
that children could be cared for by their mothers. Since
then, children have been major beneficiaries of U.S.
welfare programs for low-income families. In 1995, for
example, approximately two-thirds of recipients in the
Aid to Families with Dependent Children (AFDC)
program were childrenmany under age three.1 Childrens health and development is still an implicit goal
of welfare programs, and policymakers often assert that
program changes will produce benefits for children.
For example, one common argument is that employed
parents provide positive models of productive, self-sufficient citizenship.2 At the same time, critics have worried that policy changes made in the Personal
Responsibility and Work Opportunity Reconciliation
Act of 1996 will diminish families ability to provide
both material and personal resources to their children.
Current federal welfare policy embodied in the new
program, Temporary Assistance for Needy Families
(TANF), was designed to address the following three
major goals: (1) to increase paid employment among
parent recipients; (2) to reduce welfare dependence
that is, to reduce the number of families receiving
welfare over a long period of time; and (3) to influence
family structure by encouraging marriage and two-parent families, reducing the incidence of out-of-wedlock
childbearing, and requiring that both parents provide
for the child. The framers of welfare reform legislation
did not make poverty reduction a goal, but in recognition that merely trading welfare for work often does
not improve a familys economic situation, states were
given the option to include various financial incentives
to encourage recipients to work.
The 1990s also brought major changes in policies
designed to help working poor families outside the
welfare system, contributing to increased employment
and other outcomes observed since welfare reform. For
example, the maximum earned income tax credit
(EITC) benefit for a family with two children increased
from $1,511 in 1993 to $3,888 in 2000. Similarly,
60
both child care assistance3 and health insurance for

children4 have been extended to more nonwelfare
working poor families. These changes are designed to
reward employment, but unlike TANF, they also are
explicitly intended to reduce poverty among working
families with children.5
This article begins by examining the goals of welfare
and income support policies aimed at the working poor
in the context of more general goals for enhancing
childrens well-being. A framework is proposed for
examining how policies for poor families may influence
childrens well-being and development, and the relevant literature is briefly summarized within each component of that framework. A final section discusses the
limits of current knowledge and poses questions for
future research.
Goals for Enhancing Childrens

Development
Economic and employment policies define their goals
for children somewhat differently than do policies
focused on early childhood education, which are
intended to improve childrens lot directly. Welfare and
income support policies are based in economic thinking, whereas many child policies are rooted in child
development, education, and related fields. The following analysis explores the issues, tensions, and common threads in these different approaches to defining
goals for children and suggests a more inclusive set of
goals that may be useful in considering existing
research and provoking new questions.
Because most welfare policies entail spending public
funds, economic considerations are understandably
important in evaluating all policies affecting children.
The economic perspective has led to the idea that public spending on children is an investment in their future
economic productivity. For example, early interventions are evaluated by their long-term effects on adult
educational attainment, employment, and income.
Although costs and benefits are legitimate issues, there
is a danger in requiring that all investments in children
be justified by demonstrable profits to society in the
form of later economic productivity. It puts an enormous burden on any policy to require that it show economic benefits 10 to 20 years later.
Volume 12, Number 1
Welfare and income support policies are based in economic

thinking, whereas many child policies are rooted in child
development, education, and related fields.
In contrast, those who specialize in child development
regard quality of life during childhood as a legitimate
goal in its own right, even though they also stress childrens achievement of developmental aims as preparation for later life. In the United Kingdom and parts of
Europe, scholars are bringing a human rights perspective into policy discussions, framing the discourse
around childrens rights as embodied in the United
Nations Convention on the Rights of the Child.6
Several issues underlie these discussions about what
outcomes for children should be sought in welfare
policies. First, tension stems from the difference
between considering a childs worth in the present and
a childs worth as a future adult. To what extent should
a childs well-being be valued for its own sake? To what
extent should policies for children be judged by their
contribution to the adult that child may become? One
way of crystallizing this issue is to ask: How much public investment would be worthwhile to care for a terminally ill child who would not live beyond age 12?
Second, debate continues about whether public spending on children should be justified on the basis of
moral or human rights, or on the basis of present and
future productivity. To what extent do children have
the right to certain types of protection or benefits from
society? To what extent should public resources be
concentrated on outcomes for children that are associated with short-term or long-term productivity?
Should assistance to poor families be justified, for
example, by demonstrating that it enhances childrens
well-being by reducing their exposure to violence or
low-quality schools? Or should it be justified as a
means of increasing the probability that children will
become economically productive citizens?
Third, differences exist among those who emphasize
prevention of harm and those who emphasize positive
development. The youth development movement, for
example, stresses the difference between programs
designed to enhance skills and those aimed at preventing problems, arguing that prevention is not sufficient.
Should welfare policies for parents make positive con-
tributions to childrens well-being? Or is it sufficient to

do no harm?
Finally, some maintain that child well-being should be
defined solely by the childs characteristics, whereas
others believe the context in which the child lives
should also be considered. In models and research
investigating policy effects on children, the child outcomes are the childs behaviors and attributes, such as
school achievement, social skills, behavioral problems,
and health status. Developmental psychologists and
policy researchers alike think of the child as the locus of
development. But others have begun to focus more on
the contexts provided to children. For example, among
the rights listed in the United Nations Convention on
the Rights of the Child is an obligation of the state to
provide assistance to parents, such as income supports,
child care services, childrens health care services, and
education.6 A consortium of state policymakers recently brought together to select a common set of indicators to evaluate welfare reform also focused on
contexts, nominating outcomes that included many
aspects of childrens environments, such as reduced
child poverty and adequate housing.7 Such indicators
reflect the view that a policys effects on the home environment, for example, are as relevant as effects on a
childs school performance.
These contrasting perspectives illustrate the multiple
and sometimes contradictory criteria and objectives for
public policies affecting children. The most reasonable
approach to assessing the impact of welfare and income
support policies on children would keep various objectives in balance: emphasizing a childs present wellbeing as well as his or her future as an adult;
considering a childs well-being as well as his or her
productivity; expecting policies to promote positive
development as well as prevent harm; and defining
child outcomes to include healthy contexts as well as
skills and behaviors. Goals for healthy child development that encompass all these dimensions are summarized in Box 1.
61
Huston
Box 1
Goals for Healthy Child Development
Health and physical comfort, including shelter, nourishment,
freedom from pain and abuse, and medical care.
Indicators: housing stability versus homelessness; food sufficiency and nutrition; freedom from child abuse, use of foster
care; health care and immunization; absence of physical disability.
Family or adults who care, are reasonably constant and reliable, and who provide love and encouragement. Consistency
of caregivers and settings.
Indicators: child living out of home; parentchild relationship;
parenting warmth; social supports from other adults.
Development of intellectual and other capabilities to their
fullest, such as language skill; school achievement; and skill in
athletics, music, or art.
Indicators: language; cognitive ability; literacy; school achievement (short- and long-term); achievement in other domains.
Indicators: positive social behavior; low externalizing or

behavior problems; social skills with peers; social skills with
adults; social cognitive skills.
Responsibility and morality, including the ability to guide ones
own behavior and act in accord with societal standards of right
and wrong.
Indicators: absence of delinquency and antisocial behavior;
conformity to social expectations; sexual responsibility.
In adulthood, ability to support self and family, be a good parent,
contribute to society, be mentally and physically healthy, and not
commit crimes or abuse substances.
Emotional well-being and mental health, including self-worth,

sense of personal control, and freedom from depression and
anxiety.
Indicators: educational and occupational attainment; absence of

criminal activity or substance abuse; mental and physical health.
Indicators: low internalizing problems and anxiety; high perceived self-worth; low referrals for mental health problems.
Source: For a review of indicators and measures, see Hauser, R.M., Brown, B.V.,
and Prosser, W.R., eds. Indicators of childrens well-being. New York, NY: Russell
Sage Press, 1997.
How Welfare and Income Support Policies

Affect Children
Welfare reform and efforts to support working poor
families are designed to change the economic and personal behavior of parents. Any effects on children
occur indirectly, most likely through an impact on parents employment, family resources, use of early education and child care, and other family circumstances.
Figure 1 displays a schematic of some of the major
pathways by which policies could affect childrens
development.8 Welfare and income support policies are
designed to influence parents employment (especially
62
Skills in relating to others, both adults and peers, including, for

example, assertiveness without violence, sociability, cooperation, understanding others perspectives, complying with adult
expectations, and leadership.
maternal employment), income and material resources,

and family structure. Changes in employment,
resources, and family structure are likely in turn to
influence childrens physical and material environment,
family environment, and child care, school, neighborhood, and community environments.
The following sections summarize the literature about
how the factors in each of these boxes might affect
children.9 Rather than beginning with the policies and
tracking their potential influence through adult activities and behavior, the first section focuses on the two
right-hand columns of the schematic, summarizing
what is known about how environmental contexts
Volume 12, Number 1
relate to childrens healthy development. This approach

helps to locate domains of childrens development that
have been neglected in research and focuses directly on
the well-being of children, especially those in lowincome families. Working backward in the model, the
second section then summarizes what is known about
how parents employment, family material resources,
and family structure might affect childrens life experiences and development.
The Links between Environmental Contexts and

Childrens Development
Environmental contexts and child characteristics are
both potentially important, albeit indirect, outcomes
of public policy directed at families. Environmental

contexts that support a decent quality of life can be justified on that basis alone, but they are also important
because they can affect childrens physical, intellectual,
and socioemotional development. Random-assignment experiments, which randomly assign some participants to the program and others to a control group
have established the causal influences in some
instancesthat is, whether the context affects the
child, the child affects the context, or both are caused
by other factors. But it is also recognized that children
often play a role in both selecting and responding to
available contexts, and they in turn are influenced by
their experiences. Thus, most current child develop-
Figure 1
Conceptual Model for Policy Effects on Children
Welfare and Income
Support Policies
Intended Direct
Effects on Parents
Indirect Effects
on Childs Contexts
Indirect Effects on
Child Development
Work requirements
Sanctions
Time limits
Parental employment
Physical and material

environment
Adequate standard
of living
Health care services
Safety
Physical development
Healthy, normal growth
Healthy behavior
Human capital
development
Resources
Earning supplements
Disregardsa
Subsidies for health care
Subsidies for child care
Housing assistance
Increased income and

material resources
Strengthened family
structure
Reduced teen
childbearing
Marriage
Paternal responsibility
Fertility and parenting

Family caps
Adolescent residence
with parents
Child support
enforcement
Family environment
Parent well-being
Parenting practices
Father involvement
Social and community

environment
Child care and afterschool settings
Neighborhoods, schools,
and communities
Intellectual
development
School performance
Adult attainmentb
Life skills
Social and emotional
development
Mental health and
emotional well-being
Social competence
Morality
Responsibility
aDisregards refers to policies that allow recipients to earn more before losing benefits.
bAdult attainment refers to the process of maturing mentally and physically.
63
Huston
For infants and preschool children, the cognitive and social

environment provided in the home predicts childrens language
development, intelligence, school readiness, and school achievement
in reading and math.
ment theories maintain that effects on children and
context are bidirectionalthat is, they are mutual and
intertwined.
Physical and Material Environment
Three key elements contribute to a healthy physical
and material environment for children: (1) an adequate
standard of living (particularly sufficient food, clothing, and housing); (2) health care services, including
prevention and treatment (such as immunization); and
(3) safety from injury, violence, and environmental
hazards. In contrast, the physical and material environments that poor children experience are often characterized by hardship, hunger, homelessness, exposure to
toxins and other dangerous substances, exposure to
violence and other hazards to physical safety, and inadequate preventive health care. When families and communities do not haveor do not providethe
necessary preventive resources, children are at greater
risk of injuries, failure to thrive, mortality, untreated
medical and dental conditions, and health problems
such as asthma.10,11
Policies that improve family income or provide assistance with food, housing, and child care increase families material resources. Income is often used as a
proxy for material resources, but it is not a direct index
of the material goods and services available to children.12 Although intensive qualitative interviews with
mothers living in poverty show that most families
budget money in ways that will benefit children,13 families vary in how much of their income they spend on
children. By middle childhood, and certainly in adolescence, childrens perceptions of their familys economic strain or material hardship can be an important
influence on their sense of well-being.14
Family Environment
Parenting and the quality of the home environment
play an extremely important role in childrens lives. For
healthy development, children need adult family members and other adults who are mentally healthy, respon-
64
sible, constant, and reliable to provide them with love

and encouragement. The parentchild relationship
should be characterized by warmth, open communication, and firm (but not harsh) discipline. And children
benefit when both parents assume responsibility for
their emotional and financial support.
Welfare and income support policies that change parents employment status, family resources, and family
structure could affect childrens home environments by
changing parents time at home, parents personal sense
of well-being, and the relationships among family members. Experimental evidence has shown that the quality
of the home environment can be improved by interventions that teach mothers about child rearing, but no
evidence documents the effects of interventions less
directly focused on child rearing.15
In the very large amount of research relating parenting
and the home environment to childrens cognitive and
intellectual development, some consistent patterns have
emerged. First, although children who develop strong
intellectual skills also tend to demonstrate positive social
and emotional behavior, the home and family characteristics associated with intellectual versus social development are somewhat different. For infants and
preschool children, the cognitive and social environment provided in the home predicts childrens language
development, intelligence, school readiness, and school
achievement in reading and math, even when controlling for numerous other family and demographic characteristics.1619 Maternal intellectual ability also is a
strong predictor of childrens academic performance, a
finding that suggests both genetic and environmental
contributions to a childs performance.
In comparison to cognitive and academic outcomes,
childrens social behavior and emotional well-being
are more strongly related to the quality of
parentchild interactions, and less strongly to income
and cognitive environment.20 Maternal depression is
associated with behavioral problems much more con-
Volume 12, Number 1
Stone/Lawrence Migdale
sistently than with problems in cognitive or intellectual functioning,21 probably because of its effects on parenting practices. Children and youth who
demonstrate positive social behavior and low levels of
psychological distress perceive their relations with parents positively, and they tend to have mothers who are
warm and who avoid harsh punishment.10 Young people who engage in deviant and delinquent behavior
tend to have parents who provide low levels of supervision, monitoring, and control.22
Virtually all of this literature is subject to some methodological problems that should be of concern in evaluating policy effects. First, links between parent and child
characteristics are probably due, at least in part, to
genetic similarities.23 Second, childrens behavior elicits
and influences parenting and parent well-being as well
as responds to it, so the association of parent and child
behaviors probably indicates bidirectional causation.
For example, in one recent series of studies, better outcomes were found among adolescents whose parents
monitor their whereabouts. Investigators argue persuasively that these outcomes are not due to parents vigilant supervision, but instead are due largely to the fact
that better-adjusted adolescents volunteer such information about their whereabouts.24 Finally, in many of
the studies evaluating welfare policies, childrens social
behavior is measured by maternal report only, so the
measure reflects the mothers reaction to the child as
well as the childs behavior. (For a more detailed discussion of this topic, see the article by Chase-Lansdale
and Pittman in this journal issue.)
Social and Community Environment

Child care, schools, out-of-school activities, mass
media, social supports, and neighborhoods are sources
of formative experiences for children in general. Such
resources may be especially important for children with
a single, employed parent. Research suggests that
healthy development can be promoted by child care
that is reliable, safe, and of high quality; by education
and out-of-school activities appropriate to a childs
ability and culture; and by neighborhoods and communities that are safe and that have resources for serving children.
Child Care and After-School Programs
From the time children are infants through their
school-age and teen years, the environments they experience while their parents work can play an important
role in their development. Federal, state, and local programs provide child care assistance to low-income families, especially those moving from welfare to work, in
a variety of ways, including subsidies and resource and
referral services. Research suggests that the impact of
these policies depends on the age of the child, and the
quality, type, and stability of care.25 (See the article by
Fuller and colleagues in this journal issue.)
Even very young children are now spending a large
amount of time in child care. Current welfare policy
permits states to require parents participation in workrelated activities when children are infants, leading to
concerns that a mothers time away from her infant will
interfere with early development of a positive
motherchild relationship and a secure attachment.26
Some evidence suggests that mothers who return to
work very early in their infants lives (before about three
months of age) or whose infants have extensive early
child care are less sensitive to their children than are their
counterparts.27 Maternal sensitivity is associated with the
development of secure attachments for infants, but there
is little evidence that time with a child per se affects the
development of secure attachment.28,29
The quality of care is important, however. An abundance of evidence shows that high-quality child care
care that provides intellectual stimulation, social
involvement of adults with children, and language
interactionscan have a positive influence on childrens intellectual development, particularly among
65
Stone/Charles Thatcher
Huston
children in low-income families.26 Longitudinal, correlational studies consistently show a relationship

between high quality of care and small but significant
gains in childrens cognitive and language development, even with controls for demographic characteristics and parenting behavior.17,29,30 Some have argued
that these results could be due to unobserved characteristics that differentiate families who choose highand low-quality care,31 and that some of the results do
not endure into later childhood and adolescence.32
Nevertheless, experimental studies show that highquality programs for low-income infants and
preschoolers lead to short-term and, in some cases,
long-term improvements in school performance.33
The type of care is also important. Studies suggest that
preschool children who receive care in formal centerbased settings show better cognitive and language
66
development on average than do those cared for by relatives or nonrelatives in home-based settings.17,34,35 Similarly, somewhat older children, especially those living in
low-income areas, who participate in formal afterschool programs have been shown to perform better in
school compared with those in other types of afterschool arrangements.36 Between about third and fifth
grade, many children phase out of formal child care
programs and spend less and less time under the direct
supervision of their parents.37 During this transition,
organized youth activities offer opportunities to build
skills and interact with peers with at least some adult
supervision. Youths who participate in structured activities approved by adults have better school performance
and less deviant behavior than do those who spend
after-school time in unsupervised activities with peers,
especially in low-income families and neighborhoods.38
Volume 12, Number 1
Stability of care appears to be more consistently related

to social and emotional well-being than is quality or
type of care. Low-income children who experience
unstable child carefrequent changes and multiple
arrangementstend to have more behavioral problems
than do those with more stable child care.34 Changing
child care arrangements may be a sign of instability in
other facets of family life, including housing, parental
employment, and family compositiona pattern that
has been described as turbulence.39 Also, mounting
evidence suggests that children who enter care early or
spend many hours in care throughout the preschool
years may have more behavioral problems than do
those who spend more time in the care of parents.40
As children get older, they have increasing control over
where they spend their time, whom they associate with,
and what activities they experience. Thus, the links
between their out-of-school environments and any
positive development probably reflects bidirectional
influencesthat is, youths who are more attached to
school and nondeviant peers select more positive activities, which in turn reinforce positive behavior. An
important developmental outcome may be the
choices that young people make about how and with
whom they spend their time.41
Neighborhoods, Schools, and Communities
Although most research shows that the effects of
neighborhoods and communities are considerably
smaller than family influences, childrens development
nevertheless is influenced by where they live. Young
children and adolescents who live in middle-class
neighborhoods perform better in school than do children from comparable families living in less affluent
neighborhoods, even with controls for individual family characteristics.42
Safety and school quality are neighborhood features
with important potential consequences for childrens
development. Results from the Boston site of Move to
Opportunity, a random-assignment experiment in
which families were given vouchers to move from public housing to low-poverty neighborhoods, suggest
benefits for children. Children and youths in the experimental groups had lower levels of behavioral problems
(boys only), fewer asthma attacks and injuries, and were
less often the victims of violence than were the children
in the control group.43 Parents reported less gunfire and

less drug activity in the neighborhood. No measures of
cognitive or academic performance were reported, but
in an earlier quasi-experiment (the Gautreaux Project)
in which families moved from inner-city public housing
to suburban communities, positive effects on childrens
school performance were found.44
In sum, the evidence linking environmental contexts to
child development, especially among low-income families, is quite strong. Children who are deprived of adequate physical and material resources, who lack attention
from a warm and loving family, or who spend lengthy
hours in low-quality child care settings tend to show
negative intellectual and behavioral outcomes. Thus,
policies that work to ameliorate these conditions are likely to have a positive impact on childrens development.
The Links between Effects on Parents and Effects

on Children
The direct goals of welfare and income support policies
affecting parents employment, income, and decisions
about family structureare shown in Figure 1 as potential influences on childrens environmental contexts. This
section discusses briefly what is known about the effects
of parental employment, income, and family structure (in
conjunction with parents age) on childrens life experiences and development in low-income families.
Parental Employment
Since 1996, federal welfare policy has been designed to
move welfare recipients quickly into employment and,
in the context of a strong economy and other income
support policies, it has succeeded. Major welfare policy
strategies include work requirements, time limits on
welfare receipt, sanctions for failing to participate in
work-related activities, casework, financial incentives
such as increased income disregards, which allow recipients to earn more before losing benefits, quotas and
incentives for states, and transitional Medicaid and child
care assistance. States can and do use federal dollars and
their own funds for a wide range of other strategies.
And, of course, the EITC, the largest federal antipoverty program, provides a strong work incentive for lowincome families. Because single mothers are much more
likely to be poor than are single fathers or married parents, these policies are especially likely to increase
maternal employment in single-mother families.
67
Huston
Although there is little disagreement that families fare better

when the father is employed, researchers have long debated the
effects on children when a mother is employed.
Although there is little disagreement that families fare
better when the father is employed, researchers have
long debated the effects on children when a mother is
employed.10,45 The effects of maternal employment on
children in low-income families probably depend on
whether overall family resources increase, characteristics
of the job(s) (benefits, stimulation, schedule), characteristics of the child (age, sex, temperament), characteristics
of the mother (health, mental health, attitudes and
beliefs), and available support (family, child care, community programs).
If maternal employment increases material resources, its
effects on children are likely to be positive.46 Data indicate, however, that trading welfare assistance for employment earnings does not always change family income,47
and resources may decline when the costs of employment, such as transportation, child care, payroll taxes,
and clothing, are considered. Mandatory employment in
conjunction with time limits could result in reduced disposable income in some families. Some types of employment might increase parents skills (human capital) or
contacts with adults (social capital), which in turn could
have positive effects on the home environment.48
Proponents of welfare reform emphasize the value of
parental modeling; children who see parents going to a
job regularly and bringing home a paycheck will accept
employment as the norm of adult life. Parents work lives
can serve as a model for their children,49 but both positive and negative messages could be conveyed. Mothers
who have a positive attitude about combining work with
family are more satisfied and report less role strain than
those who believe they should be at home with their
children.50 When parents jobs are routine and repetitious or have very low wages, their children may infer
that work is dull and boring.48,51
Parents employment reduces the time they have available for child care and activities with the child. Young
children are in someone elses care when mothers are at
work, and the quality of that care is likely to mediate
effects of parental employment. But many school-age
children and youths are in self-care while their moth-
68
ers work. Although parents often monitor their children

by telephone and through neighbors,37 children without
direct supervision are vulnerable to antisocial peers or, if
they stay home, to many hours in front of the television
set.36 Self-care in late childhood and early adolescence
does not appear to pose risks for middle-class children,
but, for those who live in poverty and in dangerous
neighborhoods, the odds of behavioral problems
increase when young adolescents are unsupervised.52 In
fact, parents appear to recognize the differential risk;
middle-class and white children are more likely than
poor and minority children to be in self-care.53 (See Figure 2 in the article by Fuller and colleagues in this journal issue.)
The effects of maternal employment vary with a childs
age, but potential harmful effects do not necessarily
decline as children get older. In fact, researchers have
found that maternal employment during the adolescent
years is associated with adolescent delinquency,22 lowered educational attainment,54 and low adolescent wellbeing.55 One reason may be that, unlike younger
children, adolescents with employed mothers are generally not in supervised alternative settings.
Maternal employment can also affect childrens responsibilities and family routines.56 Policymakers often assert
that employment will provide a structure for family
schedules of getting up, having meals, and going to bed.
If a parent has a regular work schedule, this result might
occur, but many low-income jobs have irregular and
unconventional hours,57 and many parents work more
than one job. In addition, studies suggest that unconventional and extended work hours may result in children assuming more responsibility for household tasks
or caring for siblings, rather than spending more time in
child care.58 Such a shift in responsibility could lead to
either positive or negative consequences for children.
Perhaps it should not be surprising that available data
provide no simple answers about the effects of maternal
employment on childrens environments and development. Most of the literature on low-income families suggests that childrens cognitive and social development is
Volume 12, Number 1
Policies that require single mothers to seek and hold

employment are likely to vary in effect, depending on the
childs age, the childs experiences when the mother is employed, and
the mothers individual beliefs and characteristics.
In short, policies that require single mothers to seek and

hold employment are likely to vary in effect, depending
on the childs age, the childs experiences when the
mother is employed, and the mothers individual beliefs
and characteristics. Children might benefit if the job pays
decent wages and provides some opportunity for social or
cognitive stimulation; if children are in good-quality child
care, youth programs, or neighborhoods that support
positive development; if the mother does not feel excessive role strain, depression, or conflict about her dual
roles; and if family economic or social resources are
improved. Children might suffer, however, when these
conditions do not pertain. Moreover, the consequences
of a mother not being employed may be more negative
under welfare reform because the time limits on receiving
welfare threaten seriously reduced income and because
people may be discouraged by their failure to meet societys expectation that they support their families.
Income and Material Resources
Welfare policies that include wage supplements,
income disregards (which allow recipients to earn more
before losing benefits), and subsidies for necessary
expenses such as child care increase overall family
income and resources.47 If income disregards or other
subsidies are part of states new welfare programs
under TANF, however, they end when a recipient

reaches a time limit or leaves welfare for any reason.
Programs outside the welfare system, such as the
EITC, improve family resources with no time limit as
long as the parent is employed. Although there is considerable debate about the influence of income per se
on childrens cognitive and social development,61 to
the extent that increased resources improve a familys
material circumstances, the quality of the home and
child care environments, opportunities for children to
participate in beneficial activities, and the neighborhood in which a family lives, results for children are
positive. Family income during the preschool years
appears to be particularly important in predicting cognitive and educational attainment,62 probably because
income affects the quality of the home and child care
environments. The home is an important source of
intellectual stimulation before children enter formal
schooling, and analyses show that the degree of cognitive and social stimulation at home varies with family
income.63 Higher-income families also are more likely
than low-income families to use center-based care or
higher-quality home-based care for preschoolers.64
Stephen Shames/Matrix
more positive in families with employed mothers than in

those with unemployed mothers.59 But it appears that
much if not all of this difference is a function of preexisting differences in demographic attributes, skills, personality, and child-rearing practices between employed
and unemployed mothers rather than being a product of
employment.19 In a large sample of mothers and infants,
employed mothers provided a higher-quality home environment than unemployed mothers did, but these differences were entirely accounted for by maternal
characteristics.60 Parents with more skills, and better psychological and physical health, are probably more likely
to find and maintain paid employment. Parents with
those same qualities are also more likely to provide better environments for their children.
69
Huston
A series of income-maintenance experiments, conducted in the 1960s and 1970s, tested the effects of a guaranteed minimum income. Income guarantees led to
slightly better school achievement for elementaryschool-age children (but not for high school students)
and to better nutrition for a very disadvantaged rural
sample.65 Income supports also led to a higher probability of separation and divorce for couples experiencing high levels of conflict, but not for those with low
levels of conflict.66 Environmental contexts were not
assessed, but families in the experiment did buy homes
or improve the quality of their rental units more often
than controls did, probably leading to better neighborhood conditions for their families.67
Poverty and income loss affect the socioemotional climate of the home, which in turn influences childrens
psychological well-being and behavior. Studies of poor
families and families that experienced large reductions
in income indicate that the effects on children are
mediated primarily by parents psychological distress,
which may be reflected in practices such as low warmth
and frequent use of harsh punishment.10,68
Policy researchers not only debate how much good
might come from raising poor families income, some
argue that welfare income has harmful effects. Comparisons of families receiving welfare with comparable
poor families not receiving welfare show that longterm welfare recipients provide lower-quality home
environments than do other poor families.69 Although
many studies show little or no relation of welfare
receipt to childrens cognitive and social development
when differences in demographic and family characteristics are taken into account,35 some studies have
shown that welfare predicts higher rates of grade retention,70 lower academic achievement,71 and lower completed schooling among children.54 By contrast,
adolescents in welfare families placed more importance
on school than did those in poor, nonwelfare families.72
It is clearly important for policymakers to understand
why and how families receiving welfare differ from
other poor families, and why welfare income may not
confer the same benefits as other income. But it is very
difficult to find samples with truly comparable material resources and parental characteristics. People receiving welfare by definition have almost no assets, whereas
70
other people with low incomes may have more

resources of various kinds. Also, unlike other income
supports (such as the EITC), welfare carries a stigma
that may affect the way others treat parents and children, as well as the way they think about themselves.
The poor are not immune from the value placed on
work in U.S. society and from the strong implication of
failure for those who receive public assistance. Children
as young as second grade express negative stereotypes
about poor childrens behavior73 and motivations.74
In addition, entry into and out of welfare is often associated with other transitions and changes, such as job
loss or entry, parents separating or acquiring new partners, and changes in child care. Such changes affect parents and childrens well-being, parenting, and the family
environment. Mothers report high levels of behavioral
problems among children in families who have made a
transition either into or out of welfare, compared with
families whose status has not changed.69,75
To understand how policies that affect family material
resources also affect child development, studies are
needed to examine how a familys overall resources are
perceived and deployed, and how these behaviors
affect the contexts experienced by children. Experimental studies that randomly assign participants to different types of programs are extremely important;
otherwise it is difficult to isolate which differences are
a result of policies and which are due to personal characteristics. (See the article by Zaslow and colleagues in
this journal issue for a detailed review of experimental
studies.)
Family Structure
Some of the major goals of the federal welfare reform
law of 1996 were to discourage women, particularly
adolescents, from having children outside of marriage;
to encourage people with children to marry; and to
promote paternal financial responsibility for children.
The strategies intended to advance these goals include
incentives to states to reduce their out-of-wedlock
childbearing rates without increasing abortions; family
caps; a requirement that adolescent welfare recipients
attend school and live with a parent or responsible
adult; and a requirement that applicants for cash assistance identify the father of the child, and cooperate
with efforts to collect child support. (See the article by
Volume 12, Number 1
Associated Press, AP
Greenberg and colleagues in this journal issue for an

overview of these policies.)
Both adolescent and single parenting are associated
with developmental risks for children and lowered educational and occupational attainment by mothers and
children. Young women who delay childbearing receive
more education and have fewer children than those
who give birth early, but poor school achievement can
be a motivation for pregnancy as well as a consequence.
Although preexisting differences between unwed teen
mothers and other young women account for most of
the differences between them, some of the negative
effects on their children can be attributed to early childbearing and single parenthood.76
Young and single mothers provide home environments
that are, on average, less stimulating and supportive
than do older and married mothers. In an investigation
of mothers ranging in age from their late teens to their
forties, older mothers spent more time with their
infants, were more sensitive and stimulating to them,
and provided higher-quality home environments than
did younger mothers, even after controlling for demographic and marital status. Single mothers also were
less sensitive and provided lower-quality home environments than did married mothers, but these differences were largely accounted for by differences in age,
ethnicity, and maternal education.77 In a sample of children ages 4 to 16, those with single mothers spent
more time watching television on weekends, but also

spent more time in school or preschool on weekdays
than did children of married mothers.78
Despite these variations in home environment, few differences in young childrens cognitive or social behavior are associated with single parenting per se. For
example, in one study of low-income families, marital
status was not a significant predictor of school readiness or behavioral problems.18 In another study, a large
sample of children of single mothers performed slightly better than those from two-parent families, after
controlling for income and demographics.79 When
children reach adolescence, however, those with single
mothers are more likely than those living with both
biological parents to drop out of school and to have
low educational attainment.80
Because single mothers are at high risk of poverty,
many children in these families are subject to material
hardship. Numerous analyses have been designed to
separate the effects of poverty from those of family
structure on adolescent school completion, pregnancy,
and deviant behavior. One summary of the literature
concluded that about half of the difference between
children of single-mother and two-parent families was
explained by income.76 In a later analysis of several longitudinal studies, poverty was relatively more important than family structure in explaining differences in
intellectual performance, but family structure was rela-
71
Huston
To connect welfare policies, childrens environmental contexts,

and healthy child development, ... a closer look at a complex set
of links is needed.
tively more important than poverty in predicting
deviant social behavior.81
Policies designed to promote paternal responsibility
for children, through either marriage or child support, are motivated in part by the high risk of poverty for single-mother families, but many observers also
consider father involvement important for childrens
socialization and well-being. (See the article by
McLanahan and Carlson in this journal issue.) Child
support reform began in the mid-1970s, but by
1990, only 6 of every 10 eligible mothers had child
support awards, and the rate was much lower for
those with children born out of wedlock than for
those who had been divorced.82 Of mothers with
awards, about 25% did not receive any payments.83 A
recent demonstration program, Parents Fair Share,
was designed to spur low-income fathers to provide
for their children. The program increased employment and formal child support payments, but in some
cases, informal supports to children were reduced.
Moreover, paternal involvement was found to be a
double-edged sword. Slight increases in fathers
efforts to be active parents led to increased disagreements with mothers.84
Although much is known about how family structure,
maternal age, and child support relate to childrens
development, some major questions remain about how
welfare and income support policies designed to
change fertility patterns, family formation, and decisions about marriage might affect children. First, do
policy variations have any effect on decisions about fertility and marriage, particularly for adolescents? In the
United States, the availability of AFDC may have provided an incentive for adolescent childbearing. In contrast, Swedens family leave and other policies provide
incentives to begin work before having children, and
the teen birthrate is low compared with that of older
women. Sweden is one of the few countries where the
probability of having a child is higher for more-educated than for less-educated women, controlling for other
background variables.85
72
Second, if adults were induced to marry before having

children, would benefits accrue for their children? Children in two-parent stable families generally show more
positive development than do those in never-married,
divorced, and remarried families, but these differences
may be due to preexisting differences in parents individual characteristics or the degree of conflict between
adults who remain married and those who do not. Both
domestic violence and physical abuse of children occur
in a large number of low-income families. For example,
in one large sample of welfare recipients, some 28% had
been abused by an intimate partner in the previous
year.86 Most research on divorce shows that family conflict, including domestic violence, before and after a
divorce is a major predictor of childrens emotional and
behavioral problems.87 Even for couples who do not
engage in high conflict, one cannot infer that inducing
parents to marry or to stay married would provide conditions comparable to those in families where parents
choose marriage without inducement.
In addition, children do not appear to be better off
when their mother marries someone other than their
father, even though family income is, on average, substantially greater. In a longitudinal analysis of children
ages 5 to 10, those from blended families did not perform better on intellectual tests than did those in single-mother families.88 Similarly, children in stepparent
and single-mother families have similar rates of dropping out of school and adolescent pregnancy.89
In sum, the available literature offers provocative
hypotheses, but results are limited in several respects.
Virtually all of the findings are correlational, so selection effects and unobserved confounding variables
make it difficult to establish causal explanations. In
addition, much of the information predates 1996, so
parents decisions about employment and fertility were
made in a different policy context than the current
one.90 Policies that attempt to improve parents
employment, resources, and family structure probably
affect childrens development, but the effects can be
either positive or negative.
Volume 12, Number 1
Conclusions and Questions for Future

Research
To ensure positive impacts for children, welfare and
income support policies must move beyond their
exclusive focus on adults and include goals that aim to
improve childrens social and physical environments at
home and in the community. To connect welfare policies, childrens environmental contexts, and healthy
child development, however, a closer look at a complex
set of links is needed. There is currently a black box
quality about some of the data relating economic and
family structure variables to child and adult behavior;
not much is known about the processes underlying
these connections. To gain a better understanding, it is
important to consider the array of policies for working
and nonworking people with low incomes instead of
focusing solely on welfare programs involving cash
assistance. It is also important to examine whether and
for whom policies have the intended effects. Who benefits? Who is left out?
Relatively little is known about variations in policy
impacts across the many diverse ethnic groups affected.
Many legal as well as illegal immigrants are being
excluded from welfare benefits. Even for people who
are eligible for welfare or income support, important
ethnic and cultural variations may influence the relation of policies to childrens environments and, as a
result, to childrens development. For example, it is
widely believed that African-American parents are
more favorably disposed to putting their children in
child care centers, whereas Hispanic Americans prefer
to rely on family members for child care. If that is true,
then people in these ethnic groups might react differently to policies mandating employment or offering
child care assistance. (See the article by Fuller and colleagues in this journal issue for further discussion of
this topic.)
whom alternate care is not readily available. With a better understanding of individual and group differences,
policies can be tailored to meet different needs and circumstances.
A more complete understanding of the impact of certain policies targeted to parents will require more rigorous research identifying and isolating the causal links
inside the black box. Much of the current research
suffers from problems inherent in naturalistic studies
and survey methods. When comparing people with different incomes, employment levels, or welfare histories,
it is always possible that differences are due to unmeasured differences in ability, personal qualities, health, or
other characteristics. In fact, the literature suggests that
such selection variables are very important. Random-assignment experiments can ensure that differences are due to the policy treatment and not to
other factors. Similarly, surveys provide valuable information about large, representative samples, but measurement alone is necessarily a superficial way of looking
at important processes. Such data are much more useful if complemented by direct observation, in-depth
testing and interviewing, and ethnographic techniques,
all of which help in understanding process as well as
outcome.
Most important, however, is the need to keep children
in the foreground. Although two-thirds of welfare
recipients were children in 1996, the consequences of
welfare reform for children get scant attention. The
1996 welfare law eliminated the word children from
the name of the new program, Temporary Assistance
for Needy Families, signaling a shift in emphasis from
children to adults. But childrens well-being should
not be an afterthought. Healthy development of happy
children should be a front-line goal, along with establishing whether policies have the intended consequences for adult behavior. The welfare of children is
the true barometer of our success as a society.
Individual differences in parents academic and intellectual skills, psychological adjustment, and beliefs also
play a role in determining their responses to policies
and how these policies affect childrens environments.
Both low levels of literacy and high levels of depression
are implicated as barriers to achieving the goal of selfsufficiency envisioned by welfare reformers.91 Many
mothers have disabled children or family members for
73
Huston
ENDNOTES
1. Zaslow, M.J., Tout, K., Smith, S., and Moore, K.A. Implications
of the 1996 welfare legislation for children: A research perspective. Social Policy Report (1998) 12(3):134.
13. Edin, K., and Lein, L. Making ends meet: How single mothers survive welfare and low-wage work. New York: Russell Sage Foundation, 1997.
2. See Morris, P.A., Huston, A.C., Duncan, G.J., et al. How welfare
and work policies affect children: A synthesis of the research. New
York: Manpower Demonstration Research Corp., 2001.
14. For example, in studies of African-American single-mother families, adolescents feelings of psychological distress were related
even more strongly to their perceptions of family economic strain
than to the objective levels of family income. See McLoyd, V.C.,
Jayaratne, T.E., Ceballo, R., and Borquez, J. Unemployment and
work interruption among African American single mothers:
Effects on parenting and adolescent socioemotional functioning.
Child Development (1994) 65:56289. See also note 10, McLoyd.
3. Raikes, H. Investigating child care subsidy: What are we buying?

Social Policy Report: Society for Research in Child Development
(1998) 12(2):118; and Phillips, D.A., and Bridgmann, A. Child
care for low-income families: Summary of two workshops. Washington, DC: National Academy Press, 1995.
4. Devaney, B.L., Ellwood, M.R., and Love, J.M. Programs that
mitigate the effects of poverty on children. The Future of Children (Summer/Fall 1997) 7(2):88112.
5. About 70% of adults claiming the EITC are single parents, many
of whom have received welfare. See Smeeding, T.M., Ross, K.E.,
OConnor, M., and Simon, M. The economic impact of the
Earned Income Tax Credit (EITC). Unpublished manuscript,
Syracuse University, Syracuse, NY, 1999.
6. The United Nations General Assembly unanimously adopted the
Convention on the Rights of the Child on November 20, 1989.
The Convention spells out the rights to which every child under
age 18 is entitled, ranging from civil and political rights to economic, social, and cultural rights. As of 2001, some 191 United
Nations member countries had ratified the treaty; only two had
notthe United States and Somalia. The convention is controversial in the United States because some believe it undermines
the rights of parents to make decisions regarding the health, education, and religious upbringing of their children, among other
reasons. See the United Nations Web site at
http://www.unicef.org/crc/convention.htm; and Fagan, P. How
U.N. conventions on womens and childrens rights undermine
family, religion, and sovereignty. Backgrounder no. 1407. Washington, DC: The Heritage Foundation, February 5, 2001.
7. Moore, K.A. How do state policymakers think about family process
and child development in low-income families? Washington, DC:
Child Trends, 2001.
8. Figure 1 borrows heavily from earlier work, including: Zaslow,
M.J., Moore, K.A., Morrison, D.R., and Coiro, M.J. The Family
Support Act and children: Potential pathways of influence. Children and Youth Services Review (1995) 17:23149; Bos, J.M.,
Huston, A.C., Granger, R.C., et al. New hope for people with low
incomes: Two-year results of a program to reduce poverty and
reform welfare. New York: Manpower Demonstration Research
Corp., 1999.
16. The Home Observation Measure of the Environment (HOME),

a measure of the learning and social environment provided by the
parent(s), predicts both short-term and long-term school
achievement. See Bradley, R.H., Caldwell, B.M., and Rock, S.L.
Home environment and school performance: A ten-year followup and examination of three models of environmental action.
Child Development (1988) 59:85267; Duncan, G.J., BrooksGunn, J., and Klebanov, P.K. Economic deprivation and early
childhood development. Child Development (1994) 65:296318;
and Moore, K.A., and Snyder, N.O. Cognitive attainment among
firstborn children of adolescent mothers. American Sociological
Review (1991) 56:61224.
17. NICHD Early Child Care Network. The relations of early child
care to cognitive and language development. Child Development
(2000) 71:95878.
18. Zaslow, M.J., McGroder, S.M., Moore, K.A., and Le Menestrel,
S. Behavioral problems and cognitive school readiness among
children in families with a history of welfare receipt: Diverging
patterns and their predictors. Paper presented at the Biennial
Meeting of the Society for Research in Child Development.
Albuquerque, NM. April 1518, 1999.
19. Zaslow, M.J., McGroder, S.M., Cave, G., and Mariner, C.L.
Maternal employment and measures of childrens health and
development among families with some history of welfare receipt.
In Research in the sociology of work. Vol. 7, Work and family. R.
Hodson and T. L. Parcel, eds. Stamford, CT: JAI Press, 1999,
pp. 23359.

well-being of Americas children and youth: 2000. Washington,
DC: U.S. Government Printing Office, 2000.
20. For example, one study found that cognitive competencies in

children appeared to be linked to family income, but socioemotional competencies appeared to be associated with quality of
time spent with parents. See Amato, P., and Ochiltree, G. Family
resources and the development of child competence. Journal of
Marriage and the Family (1986) 48:4756. Similarly, another
study found that although socioeconomic status (SES) and
mothers mental health predicted both cognitive outcomes and
social competence, SES explained more variance in cognitive outcomes than social competence, whereas mothers mental health
explained more variance in social competence. See Sameroff, A.J.,
and Seifer, R. Familial and child competence. Child Development
(1983) 54:125468.
12. Mayer, S., and Jencks, C. Poverty and the distribution of material
hardship. Journal of Human Resources (1989) 24:88114.
21. Leadbeater, B.J., and Bishop, S.J. Predictors of behavioral problems in preschool children of inner-city Afro-American and Puer-
9. The topic is vast, and because the relations differ considerably for
children of different ages, for boys and girls, and so on, this summary is by definition incomplete.
74
15. Brooks-Gunn, J., Klebanov, P.K., and Liaw, F. The learning,

physical, and emotional environment of the home in the context
of poverty: The infant health and development program. Children and Youth Services Review (1995) 17:25176.
Volume 12, Number 1
to Rican adolescent mothers. Child Development (1994)

65:63848.
22. Sampson, R.J., and Laub, J.H. Urban poverty and the family
context of delinquency: A new look at structure and process in a
classic study. Child Development (1994) 65:52340.
23. Harris, J.R. The nurture assumption: Why children turn out the
way they do. New York: Free Press, 1998; and Rodgers, J.L., and
Rowe, D.C. Poverty and behavior: Are environmental measures
nature and nurture? Developmental Review (1997) 17:35875.
24. Kerr, M., and Stattin, H. What parents know, how they know it,
and several forms of adolescent adjustment: Further support for a
reinterpretation of monitoring. Developmental Psychology (2000)
36:36680.
25. See note 3, Raikes.
26. See earlier journal issue on caring for infants and toddlers in The
Future of Children: (Spring/Summer 2001) 11(1).
27. Clark, R., Hyde, J.S., Essex, M.J., and Klein, M.H. Length of
maternity leave and quality of motherinfant interactions. Child
Development (1997) 68:36483; and NICHD Early Child Care
Research Network. Child care and motherchild interaction in
the first 3 years of life. Developmental Psychology (1999)
35:13991413.
28. Huston, A.C. Child care and child development. In International encyclopedia of social and behavioral sciences. N. Eisenberg, ed.
Oxford: Elsevier Sciences, 2000.
29. Lamb, M.E. Nonparental child care: Context, quality, correlates,
and consequences. In Child psychology in practice. Vol. 4, Handbook of child psychology. 5th ed. W. Damon, series ed., and I. Sigel
and K.A. Renninger, eds. New York: John Wiley & Sons, 1997,
pp. 73134.
30. See also Peisner-Feinberg, E.S., and Burchinal, M.R. Concurrent
relations between preschool childrens child-care experiences and
concurrent development: The Cost, Quality, and Outcomes Study.
Merrill-Palmer Quarterly (1997) 43:45177; Peisner-Feinberg,
E.S., Burchinal, M.R., Clifford, R.M., et al. The children of the
Cost, Quality, and Outcomes Study go to school. Chapel Hill: University of North Carolina Press, 1999; Vandell, D.L., and Wolfe, B.
Child care quality: Does it matter and does it need to be improved?
Madison: Institute for Research on Poverty, University of WisconsinMadison, 2000; and Shonkoff, J.P., and Phillips, D.A., eds.
From neurons to neighborhoods: The science of early childhood development. Washington, DC: National Academy Press, 2000.
31. Duncan, G.J., and Gibson, C. Selection and attrition in the
NICHD childcare studys analyses of the impacts of childcare quality on child outcomes. Unpublished paper. Evanston, IL: Northwestern University, March 29, 2000.
32. Scarr, S. American child care today. American Psychologist (1998)
53:95108.
33. Lazar, I., and Darlington, R. Lasting effects of early education: A
report from the Consortium for Longitudinal Studies. Monographs
of the Society for Research in Child Development (1982) 47:1151;
and Ramey, C.T., Campbell, F.A., Burchinal, M.R., et al. Persistent effects of early childhood education on high-risk children and
their mothers. Applied Developmental Science (2000) 4:214. See
also The Future of Children (Winter 1995) 5(3).
34. Bos, J.M., and Granger, R.C. Estimating effects of day care use
on child outcomes. Paper presented at the Biennial Meeting of
the Society for Research in Child Development. Albuquerque,
NM. April 1518, 1999.
35. Yoshikawa, H. Welfare dynamics, support services, mothers earnings, and child cognitive development: Implications for contemporary welfare reform. Child Development (1999) 70:779801.
See also the article by Fuller and colleagues in this journal issue.
36. Posner, J.K., and Vandell, D.L. Low-income childrens afterschool care: Are there beneficial effects of after-school programs?
Child Development (1994) 65:44056; and Posner, J.K., and
Vandell, D.L. After-school activities and the development of lowincome urban children: A longitudinal study. Developmental Psychology (1999) 35:86879.
37. Belle, D. The after-school lives of children. Mahwah, NJ: Lawrence
Erlbaum Associates, 1999.
38. Carnegie Corporation of New York. Starting points: Meeting the
needs of our youngest children. New York: Carnegie Corporation,
1994; Pettit, G.S., Bates, J.E., and Dodge, K.A. Supportive parenting, ecological context, and childrens adjustment: A sevenyear longitudinal study. Child Development (1997) 68:90824;
Catalano, R.F., Berglund, M.L., Ryan, J.A.M., et al. Positive
youth development in the United States: Research findings on evaluations of positive youth development programs. Seattle: Social
Development Research Group, University of Washington School
of Social Work, 1999. See also The Future of Children (Fall 1999)
9(2).
39. Child Trends. Children and welfare reform: A guide to evaluating
the effects of state welfare policies on children. Washington, DC:
Child Trends, 1999.
40. Blair, C. The early identification of risk for grade retention
among African American children at risk for school difficulty.
Applied Developmental Science (2001) 5:3750; Hofferth, S.L.
Child care in the first three years of life and preschoolers language and behavior. Paper presented at the Biennial Meeting of
the Society for Research in Child Development. Albuquerque,
NM. April 1518, 1999; NICHD Early Child Care Research
Network. Early child care and self-control, compliance, and problem behavior at twenty-four and thirty-six months. Child Development (1998) 69:114570. See also NICHD Early Child Care
Research Network. Quantity of child care and problem behavior.
Paper presented at the Biennial Meeting of the Society for
Research in Child Development. Minneapolis, MN. April 1922,
2001.
41. In a large-scale study of adolescents in Philadelphia, for example,
two patterns of successful development were identified. One of
these was defined by active involvement in organized community
and school activities. See Furstenberg, F.F., Cook, T.D., Eccles,
J.S., et al. Managing to make it: Urban families and adolescent
success. Chicago: University of Chicago Press, 1999.
42. Brooks-Gunn, J., Duncan, G.J., Leventhal, T., and Aber, J.L.
Lessons learned and future directions for research on the neighborhoods in which children live. In Neighborhood poverty. Vol. 1,
Context and consequences for children. J. Brooks-Gunn, G.J. Duncan, and J.L. Aber, eds. New York: Russell Sage Foundation,
1997, pp. 27997.
43. Kling, J.R., Liebman, J.B., and Katz, L.F. Bullets dont got no
75
Huston
name: Consequences of fear in the ghetto. Paper presented at the

Conference on Mixed Methods, sponsored by the MacArthur
Network on Successful Pathways through Middle Childhood.
Santa Monica, CA. January 2527, 2001.
44. Rosenbaum, J.E. Changing the geography of opportunity by
expanding residential choice: Lessons from the Gautreaux program. Housing Policy Debate (1995) 6:23169.
59. Zaslow, M.J., Rabinovich, B.A., and Suwalsky, J.T.D. From

maternal employment to child outcomes: Preexisting group differences and moderating variables. In Employed mothers and their
children. J.V. Lerner and N.L. Galambos, eds. New York: Garland
Publishing, 1991, pp. 23782.
45. McLoyd, V.C. Socialization and development in a changing

economy: The effects of paternal job and income loss on children. American Psychologist (1989) 44:293302.
60. Quality based on assessments using the HOME and observed

maternal sensitivity and cognitive stimulation. See note 16,
Bradley, et al, and note 58, Aronson and Huston.
46. Harvey, E. Short-term and long-term effects of early parental

employment on children of the National Longitudinal Survey of
Youth. Developmental Psychology (1999) 35:44559.
61. Mayer, S.E. What money cant buy: Family income and childrens
life chances. Cambridge, MA: Harvard University Press, 1997.
47. Bloom, D., and Michalopoulos, C. How welfare and work policies
affect employment and income: A synthesis of research. New York:
Manpower Demonstration Research Corp., 2001.
48. Parcel, T.L., and Menaghan, E.G. Parents jobs and childrens
lives. New York: Aldine de Gruyter, 1994.
49. Ripke, M., Huston, A.C., and Mistry, R.S. Characteristics of parents jobs and childrens occupational aspirations and expectations
in low-income families. Paper presented at the Biennial Meeting
of the Society for Research on Child Development. Minneapolis,
MN. April 1922, 2001.
50. Jackson, A.P. Black, single, working mothers in poverty: Preferences for employment, well-being, and perceptions of children.
Social Work (1993) 38:2634. See also Chang, Y.E., and Huston,
A.C. The relations of maternal beliefs about employment and
infant child care to maternal well-being. Paper presented at the
Biennial Meeting of the Society for Research in Child Development. Minneapolis, MN. April 1922, 2001.
51. Parcel, T.L., and Menaghan, E.G. Effects of low-wage employment on family well-being. The Future of Children (Summer/Fall
1997) 7(2):11621; and Moore, K.A., and Driscoll, A.K. Lowwage maternal employment and outcomes for children: A study.
The Future of Children (Summer/Fall 1997) 7(2):12227.
52. See note 38, Pettit, et al.
53. Capizzano, J., Tout, K., and Adams, G. Child care patterns of
school-age children with employed mothers. Washington, DC:
Urban Institute Press, 2000.
54. Duncan, G.J., and Yeung, W-J.J. Extent and consequences of
welfare dependence among Americas children. Children and
Youth Services Review (1995) 17:15782.
55. Crouter, A.C., Bumpus, M.F., and McHale, S.M. Linking parents work pressure and adolescents well-being: Insights into
dynamics in dual-earner families. Developmental Psychology (1999)
35:145361.
56. Weisner, T.S. Human development, child well-being, and the cultural project of development. New Directions for Child Development (1998) 81:6985.
57. Hofferth, S.L. Caring for children at the poverty line. Children
and Youth Services Review (1995) 17:6190.
58. Time use analyses show that a mothers work time has less influence on the amount of time children spend in child care than on
maternal activities such as leisure and housework. See Bianchi,
S.M. Maternal employment and time with children: Dramatic
change or surprising continuity? Demography (2000) 37:40114.
See also Aronson, S.R., and Huston, A.C. Work time, infant
time, and the motherinfant relationship. Paper presented at the
76
Biennial Meeting of the Society for Research in Child Development. Minneapolis, MN. April 1922, 2001.
62. See note 16, Duncan, et al.

63. Garrett, P., Ngandu, N., and Ferron, J. Poverty experiences of
young children and the quality of their home environments.
64. Capizzano, J., Adams, G., and Sonenstein, F. Child care arrangements for children under five: Variation across states. Washington,
DC: Urban Institute Press, 2000; NICHD Early Child Care
Research Network. Familial factors associated with the characteristics of nonmaternal care for infants. Journal of Marriage and the
Family (1997) 59:389408.
65. Salkind, N.J., and Haskins, R. Negative income tax: The impact
on children from low-income families. Journal of Family Issues
(1982) 3(2):16580.
66. U.S. Department of Health, Education and Welfare. Overview of
the Seattle/Denver income maintenance experiments: Final report.
Washington, DC: DHEW, Office of Income Security, 1983.
67. Kershaw, D., and Fair, J. The New Jersey income maintenance
experiment. New York: Academic Press, 1976.
68. Conger, R.D., and Elder Jr., G.H. Families in troubled times:
Adapting to change in rural America. New York: Aldine de
Gruyter, 1994.
69. Moore, K.A., Morrison, D.R., Zaslow, M.J., and Glei, D.A.
Ebbing and flowing, learning and growing: Family economic
resources and childrens development; and Smith, J.R., and
Brooks-Gunn, J. Developmental effects of natural transitions in
welfare receipt. Both papers presented at a research briefing of
the National Academy of Sciences, Board on Children and Families. Washington, DC. December 56, 1994. See also Zill, N.,
Moore, K.A., Smith, E.W., et al. The life circumstances and
development of children in welfare families: A profile based on
national survey data. In Escape from poverty: What makes a difference for children? P.L. Chase-Lansdale and J. Brooks-Gunn, eds.
New York: Cambridge University Press, 1995, pp. 3859.
70. Guo, G., Brooks-Gunn, J., and Harris, K.M. Parents labor force
attachment and grade retention among urban black children.
Sociology of Education (1996) 69:21636.
71. See note 69, Smith and Brooks-Gunn.
72. Kalil, A., and Eccles, J.S. Does welfare affect family processes and
adolescent adjustment? Child Development (1998)
69:15971629.
73. Mistry, R.S. A multidimensional analysis of young childrens
beliefs about wealth and poverty. Ph.D. diss., University of Texas
at Austin, 1999.
74. Crosby, D.A. Childrens causal attributions for economic inequality: Relation to age and socioeconomic environment. Ph.D. diss.,
University of Texas at Austin, 2000.
Volume 12, Number 1
75. Hofferth, S.L., Smith, J., McLoyd, V.C., and Finkelstein, J.

Achievement and behavior among children of welfare recipients,
welfare leavers, and low-income single mothers. Journal of Social
Issues (2000) 56(4):74774.
Press, 1994. See also Maynard, R. Kids having kids: Economic
costs and social consequences of teen pregnancy. Washington, DC:
77. Aronson, S.R., and Huston, A.C. The motherinfant relationship
in single, cohabiting and married families: A case for marriage?
Unpublished paper. Austin: University of Texas at Austin, 2001.
78. Timmer, S.G., Eccles, J.S., and OBrien, K. How children use
time. In Time, goods, and well-being. F.T. Juster and F.P. Stafford,
eds. Ann Arbor, MI: University of Michigan, Institute for Social
Research, 1985, pp. 35382.
79. Smith, J.R., Brooks-Gunn, J., and Klebanov, P.K. Consequences
of living in poverty for young childrens cognitive and verbal ability and early school achievement. In Consequences of growing up
poor. G.J. Duncan and J. Brooks-Gunn, eds. New York: Russell
Sage Foundation, 1997, pp. 13289.
84. Knox, V., and Redcross, C. Parenting and providing: The impact
of Parents Fair Share on paternal involvement. New York: Manpower Demonstration Research Corp., 2000.
85. Gustafson, S.S., and Stafford, F.P. Links between early childhood
programs and maternal employment in three countries. The
Future of Children (Winter 1995) 5(3):16174.
86. Gennetian, L.A., and Miller, C. Reforming welfare and rewarding work: Final report on the Minnesota Family Investment Program. New York: Manpower Demonstration Research Corp.,
2000.
87. Hetherington, E.M., Bridges, M., and Insabella, G.M. What matters? What does not? Five perspectives on the association between
marital transitions and childrens adjustment. American Psychologist (1998) 53:16784.
88. Gennetian, L.A. One or two parents? Half or step siblings? The
effect of family composition on cognitive outcomes for young
children. Unpublished manuscript. Manpower Demonstration
Research Corp. New York, 1999.
80. See note 76, McLanahan and Sandefur.
89. McLanahan, S.S., Astone, N.M., and Marks, N. The role of

mother-only families in reproducing poverty. In Children in
poverty: Child development and public policy. A.C. Huston, ed.
New York: Cambridge University Press, 1991, pp. 5178.
81. McLanahan, S.S. Parent absence or poverty: Which matters

more? In Consequences of growing up poor. G.J. Duncan and J.
Brooks-Gunn, eds. New York: Russell Sage Foundation, 1997,
pp. 3548.
90. See the article by Zaslow and colleagues in this journal issue for a
summary of the results of experimental studies testing the effects
of such new policies as time limits, mandatory participation, and
financial incentives.
82. Garfinkel, I., McLanahan, S.S., and Robins, P.K. Child support
and child well-being: What have we learned? In Child support
and child well-being. I. Garfinkel, S.S. McLanahan, and P.K.
Robins, eds. Washington, DC: Urban Institute Press, 1994.
91. Zaslow, M.J., Hair, E.C., Dion, M.R., et al. Maternal depressive
symptoms and low literacy as potential barriers to employment in
a sample of families receiving welfare: Are there two-generational
implications? Women & Health (2001) 32(3):21151.

well-being of Americas children and youth: 2000. Washington,
DC: U.S. Government Printing Office, 2000.
77
Experimental Studies of
Welfare Reform and Children
Martha J. Zaslow, Kristin A. Moore, Jennifer L. Brooks,
Pamela A. Morris, Kathryn Tout, Zakia A. Redd, and Carol A. Emig
SUMMARY
Even prior to passage of federal welfare reform,

many demonstration programs anticipated key features of the 1996 law, such as work-first strategies, time limits on welfare receipt, and financial
incentives to work. Over the past decade, 10
experimental evaluations of these programs have
extended their studies to examine the impacts on
children. This article provides a synthesis of findings from the first seven of these studies to release
results concerning child impacts. Key observations
include the following:
Across the different types of welfare-to-work programs examined, researchers found neither widespread harm nor widespread benefit to young
children, but some significant impacts did occur.
Favorable impacts tended to occur in programs
that improved family economic status or maternal education, but these programs still did not
bring children to the level of national norms for
positive child development.
Unfavorable impacts tended to occur when families did not show economic progress or when
their economic situation worsened, when the
children were adolescents, andunexpectedly
when the families were believed to be at lower
risk for long-term welfare receipt.
Thus, although impacts were not widespread,

these programs did have the potential to affect
children for both better and worse across a range
of developmental outcomes. The authors conclude that these findings underscore the importance of strengthening program approaches to
enhance developmental outcomes for children in
families being served by the welfare system.
Martha J. Zaslow, Ph.D., is vice president for research

and director of the content area on welfare and
poverty at Child Trends, a nonpartisan research
organization in Washington, D.C., that studies children and families.
Kristin A. Moore, Ph.D., is president and senior
scholar at Child Trends.
Jennifer L. Brooks, Ph.D., is a senior research associate at Child Trends.
Pamela A. Morris, Ph.D., is a senior research associate at Manpower Demonstration Research Corporation, a nonprofit social policy research organization
dedicated to learning how to improve the well-being
of low-income people.
Kathryn Tout, Ph.D., is a senior research associate at
Child Trends.
Zakia A. Redd, M.P.P., is a research analyst at Child
Trends.
Carol A. Emig, M.P.P., is vice president and director
of public information and policy at Child Trends.
79
Zaslow, Moore, Brooks, et al.
n enduring irony of welfare policy in the

United States is that welfare programs
launched out of concern for children are
evaluated primarily on adult outcomes,
such as welfare receipt, employment, earnings, income,
and marriage.1 Fortunately for those concerned with
child well-being, within the plethora of welfare studies
is a small group of rigorous experimental studies that
focus on how welfare programs and policies affect children as well as adults.2 Experimental studies randomly
assign families to either an experimental group that can
participate in the program, or a control group that is
not eligible to participate. Random assignment assures
that any differences that emerge between the experimental and control groups are due to the program and
not to other differing characteristics of families and
children. Over the past decade, 10 experimental evaluations of welfare programs have extended their studies
to examine the impacts on children.3 Box 1 describes
the programs studied in these evaluations, grouped
into four categories based on the strength of their work
incentives, as well as their barriers or penalties for not
working or not leaving welfare.4
All the programs were implemented before the 1996
federal welfare reform law was passed, but had provisions that anticipated key features of the new law, such
as work-first strategies, time limits on welfare
receipt, and financial incentives to work. Similar to the
welfare-to-work programs being implemented today,
these programs generally focused on improving family
economic self-sufficiency, and did not include components aimed directly at improving outcomes for children such as screening for child health or
developmental problems, or providing high-quality
early childhood intervention programs. Nevertheless,
many in the field believe that impacts on children
should be considered alongside impacts on adults
when examining the success of these programs, first
because an underlying goal of welfare policies is to protect the well-being of children, and second because
changes in adult economic outcomes (or other aspects
of family life, such as parenting behavior or maternal
psychological well-being) could be important to childrens development. (See the article by Huston in this
journal issue.)
80
To this end, the experimental studies of earlier welfareto-work programs can provide many useful insights
about how various types of welfare-to-work programs
affect children. Above all, the experimental design identifies impacts that are attributable to a welfare-to-work
program rather than other factors. At the same time,
because the evaluations tend to study packages of
program components, it is difficult to isolate exactly
which aspect of the program may be causing an impact.
Also, program impacts are affected by general economic conditions and the other benefits available to lowincome families (such as food stamps and child care
subsidies) at the time and place of the study. The evaluations reveal little about how policies might affect families under different economic conditions or families not
directly examined, such as low-income families not
receiving welfare. Although these studies examine the
major welfare-to-work strategies, the full range of program approaches that have been implemented more
recently are not included. Even so, these studies can be
very useful in determining how impacts on children differ given different program goals and components and
whether programs changed adult factors, such as family
economic status, parenting behavior, and maternal psychological well-being, that can in turn affect children.
To date, results are available from 7 of the 10 evaluations. This article provides a synthesis of findings from
these seven studies, distilling and presenting the evidence concerning welfare reforms effects on children.5,6 The first section summarizes the impacts on
adults and children by type of program. The second
section explores patterns related to the favorable and
unfavorable impacts for children, and the key factors
that appear to be associated with these patterns. Based
on these findings, the final section offers recommendations to guide federal, state, and local policymakers as
they consider ways to help children in families moving
from welfare to work.
Program Impacts on Adults and Children

Children are not the primary focus of welfare-to-work
programs, so understanding how these programs affect
adults is important to understanding the impacts on
children. Although programs varied in their expectations regarding such adult outcomes as program par-
Volume 12, Number 1
Experimental Studies of Welfare Reform
Broadly Targeted Programs for Teenage Recipients
Courtney Bennett
Both programs targeted to teenage welfare recipients

New Chance8 and the Teenage Parent Demonstration
(TPD) program9provided a broad array of services,
addressing personal development as well as economic
issues, in keeping with concerns about the range of difficulties faced by young parents.10 For example, both
programs included components focusing on life skills
and personal development (such as parenting behavior
and psychological well-being), in addition to education
and work preparation.
ticipation, maternal education, and welfare receipt, all

the programs measured a common core of indicators
of economic progress: employment, earnings, overall
family income, and poverty. Findings related to these
measures reveal a broad range of impacts on adults:
some mothers achieved sustained economic improvement, others experienced no economic progress or a
setback, and many fell somewhere in between.
With respect to impacts on children, findings are summarized across three areas: academic progress and cognitive development (academic/cognitive); behavioral
and emotional adjustment (behavioral/emotional);
and health and safety. These categories help to identify
whether patterns are consistent across different aspects
of development. The following synthesis also explores
whether the programs affected noneconomic aspects of
family life that were not explicitly targeted by the programs, but that may shape results for children, such as
parenting, fathers involvment, and mothers mental
health. As was the case with adult impacts, a broad
range of child impacts also emergedsome favorable,
some unfavorable, and often only weak or no impacts
were detected in these evaluations.
A detailed discussion of these findings follows, organized
by the four categories of programs outlined in Box 1.7
(The pattern of adult and child impacts from each study
is displayed in the appendix at the end of this article.)
At the outset, teen mothers in these programs participated more in education and work preparation activities and received less welfare than control group
mothers did, but these initial gains often faded over
time. Program enrollment did not generally result in
sustained improvements in family economic status. As
noted by the researchers, these findings underscore the
special challenges of bringing about lasting change for
the particularly disadvantaged subgroup of welfare
recipients who are teen parents.
Meanwhile, many impacts on children in these programs were neutralthat is, no impacts or weak
impactsand some were unfavorable. In the Newark
site of the TPD program, unfavorable impacts were
found on two direct assessments of childrens cognitive
development11 as well as a measure of the childrens
perception of their school.12 In addition, mothers in
this program rated their children less positively on
expressiveness, although their childrens overall social
behavior scores did not differ from those of control
group children.13
It is particularly surprising that the New Chance program also had unfavorable impacts for children in
both academic/cognitive and behavioral development. Of all the programs included in this synthesis,
New Chance had the most explicitly two-generational
focus, seeking to improve outcomes for children as
well as mothers by providing parenting education,
developmentally appropriate child care, and access to
health care. It was expected that the program would
lead to increases in maternal education and improvements in family economic status, which would contribute to improved outcomes for children. However,
81
mothers in New Chance rated their childrens academic performance as significantly lower than did
mothers in the control group.14 In addition, mothers
reports of childrens behavioral problems were higher
in the program group than in the control group,
whereas reports of their childrens positive behaviors
were significantly lower.
These findings caution that it may be particularly difficult

to achieve not only positive economic impacts, but also
positive impacts on childrens development, in families
headed by young mothers on welfare. Teen parents now
are subject to more stringent requirements than they
were under the previous welfare program, Aid to Families with Dependent Children. With limited exceptions,
Box 1
Overview of Welfare Programs Studied
Broadly Targeted Programs for Teenage Welfare Recipients
New Chance Demonstration: A mostly voluntary program for
teenage mothers on welfare who had dropped out of school.
The program provided comprehensive education, training, and
other services intended to increase the long-term self-sufficiency and well-being of teenage mothers and their children.
This is the only program examined in this article that was
explicitly two-generational, articulating goals for improving
outcomes for children as well as their mothers. The program
operated in 16 sites across 10 states between 1989 and 1992.
Impacts for children ages 3.5 to 10 at the 3.5-year follow-up
are aggregated and reported across the 16 sites.
Teenage Parent Demonstration (TPD): A mandatory program
for first-time teenage parents applying for welfare, with sanctions for those who did not comply with requirements to participate in education, job training, or employment-related
activities. The program provided case management, life skills
and personal development workshops, child care assistance,
and transportation assistance. The program operated in Chicago and in two sites in New Jersey, Newark and Camden,
between 1987 and 1991. Impacts for first-born children ages 5
to 10 (with the majority of children between ages 6 and 8) at
the final follow-up (conducted, on average, 6.5 years after
intake) are reported separately for the three study sites.
Mandatory Education-First and Work-First Programs
Job Opportunities and Basic Skills Training (JOBS): A welfare-to-work program established under the Family Support
Act of 1988. This program aimed to reduce welfare dependen-
82
cy through a combination of intensified case management and

participation in work or education activities. Failure to comply
could result in a reduction in welfare benefits. The program
was implemented nationwide between 1988 and 1996.
Impacts for children ages 5 to 7 at the 2-year follow-up are
reported from JOBS programs in three sites: Atlanta, Georgia;
Grand Rapids, Michigan; and Riverside, California. In each site,
two different program approaches were studied: a human capital development or education-first approach, and a labor
force attachment or work-first approach.
Strong Financial Work Incentive Programs
Minnesota Family Investment Program (MFIP): This program
included generous financial incentives for work, streamlined
receipt of cash assistance and food stamps, and direct payments to child care providers. The program was implemented
under waivers to federal rules as a pilot in three urban counties and four rural counties (the study sites) between 1994 and
1998. A modified version of the program was implemented
statewide following passage of federal reform. Two MFIP variations were studied: a program with financial incentives only,
and a full program with both financial incentives and a work
mandate. Impacts for school-age children (ages 5 to 12) and
adolescents (ages 13 to 18) at the 3-year follow-up are reported separately for the children of long-term welfare recipients
and recent welfare applicants in each variant of the program.
New Hope Project: This voluntary demonstration program
offered an earnings supplement to raise the income of full-time
workers to the poverty level; health insurance; child care subsidies; and community service job opportunities for partici-
Volume 12, Number 1
the current welfare reform law prohibits states from using

federal welfare funds for unmarried teen parents who do
not go to school and do not live in an adult-supervised
setting. (See the article by Greenberg and colleagues in
this journal issue.) It will be critical to determine whether
these participation and residence requirements result in
more favorable impacts for both mothers and children.
Mandatory Education-First and

Work-First Programs
Various versions of the mandatory education-first
and work-first program, Job Opportunities and
Basic Skills Training (JOBS), were implemented
nationwide during the late 1980s and early 1990s.
Results reported here are from the Child Outcomes
Box 1
continued
pants who were not able to find work. The program operated in
two low-income areas in Milwaukee, Wisconsin, between
August 1994 and December 1998. Impacts for children ages 3
to 5 and ages 6 to 12 at the 2-year follow-up are reported separately by age and gender.
Canadas Self-Sufficiency Project (SSP): This demonstration
program offered substantial financial incentives to participants
who worked at least 30 hours per week and left welfare. Individuals could receive the financial supplement within a year of
random assignment to the program and could continue to
receive the supplement for up to 3 years if they worked full
time and did not receive welfare. The program operated in New
Brunswick and British Columbia between November 1992 and
March 1995. Impacts are reported for children ages 3 to 5, 6 to
11, and 12 to 18 at the 3-year follow-up.
Programs with a Time Limit or Benefit Termination Component
Florida Family Transition Program (FTP): This program included a time limit of 24 months of welfare receipt in any 60-month
period for most applicants (36 months in any 72-month period
for the least job-ready), along with a small financial work incentive and parental responsibility mandates that included ensuring child immunizations and school attendance. The program
operated under waivers to federal rules from 1994 to 1999, initially in Escambia County. Other counties were phased in gradually, and the program then served as a model for the statewide
welfare program implemented following passage of federal
reform. Impacts are reported for school-age children (ages 5 to
12) and adolescents (ages 13 to 17) at the 4-year follow-up.
Connecticuts Jobs First Program: This program included a

21-month time limit on benefits; mandatory employment services; earned income disregards; and a partial family cap (limiting benefits for children born to families while on welfare),
among other provisions. The program was implemented
statewide in 1996 under waivers to federal rules, and with
modifications, remained the states welfare program following
passage of federal reform. The evaluation focuses on Manchester and New Haven. Child impact study results for this program are still pending.
Indiana Welfare Reform Evaluation: Provisions of this program
include a focus on work first and labor force attachment,
time limits on cash assistance, sanctions for not meeting
parental responsibilities (including child immunization and
school attendance), and a family cap. The program began to be
implemented statewide in 1995 under waivers to federal rules,
with additional modifications in 1997 in response to passage of
federal reform. Child impact study results for this program are
still pending.
Iowas Family Investment Program (FIP): This program
includes financial incentives for employment, as well as components facilitating asset accumulation and family stability by
broadening eligibility rules for two-parent working families.
Parents must create and adhere to an agreement specifying
their steps to self-sufficiency, or their benefits can be reduced
or terminated. The program was implemented in October 1993
under waivers to federal rules, and many original FIP provisions were retained following passage of federal reform. The
program is being evaluated in nine counties. Child impact
study results are still pending.
83
Study of the national evaluation of this program

(National Evaluation of Welfare-to-Work Strategies, or
NEWWS).15 Data gathering focused on mothers who
were age 19 or older who had a 3- to 5-year-old at the
time of enrollment, and who were participating in
either the education-first or work-first components of
the JOBS programs in three sites: Atlanta, Georgia;
Grand Rapids, Michigan; and Riverside, California.16
Like the programs targeted to teens, this set of programs increased adult participation in work preparation
activities, and also in a number of programs (primarily
the education-first programs), educational attainment.
Overall, these programs more consistently increased
employment than did the teen-targeted programs, at
least through the follow-up two years after mothers
were enrolled in the evaluation.17 The programs did
not affect income levels, however, and results were
mixed with regard to the proportion of families living
in poverty. In one of the six programs studied, the percentage of families living at or above the poverty line
increased, whereas it decreased in two other programs.
84
compared with the control group. Moreover, unfavorable impacts were found for childrens health outcomes
in two programs. In Riverside, mothers gave lower
overall health ratings to their children in both the education-first and work-first programs, and were less likely to rate their childrens health as very good or
excellent.
In the present policy context, welfare-to-work programs have more stringent employment requirements
and consequences for noncompliance than was the case
in most JOBS programs. Yet impacts on children in
JOBS programs, especially programs requiring quick
entry into the labor force, may be particularly informative with respect to current programs that seek to
increase employment without attempting explicitly to
increase income. While some unfavorable impacts
occurred, at least some of the impacts among JOBS
programs were favorable, in contrast to the broadly targeted programs for teenage recipients, which showed
no favorable patterns of impacts on children.
Moreover, evaluations of these programs detected mostly weak or no impacts on children ages five to seven.
When impacts did occur, they sometimes resulted in
improved outcomes, and sometimes in worsened outcomes. Positive impacts were found in academic/cognitive development, based on improved scores in direct
assessments of childrens cognitive skills. Children in
Atlanta and Grand Rapids scored higher on these assessments.18 Interestingly, the favorable impacts on cognitive
development all occurred in programs in which mothers
showed increases in educational attainment, though one
program (the education-first program in Riverside, California) increased maternal education, and this did not
result in favorable cognitive impacts on children.
The three strong financial work incentive programs

Minnesota Family Investment Program (MFIP),19 New
Hope Project,20 and Canadas Self-Sufficiency Project
(SSP)21all provided financial incentives and other
supports to make work pay. They all had an explicit
goal of reducing poverty by providing a generous cash
supplement tied to work, though the programs differed in the form and source of the supplement.22
These programs stand apart because they not only
increased employment, but they also improved income
and reduced poverty (albeit not for all subgroups).
Furthermore, with only a few exceptions, these positive
impacts had not faded when follow-up assessments
were conducted two to three years later.23
While impacts on cognitive development, when they

occurred, were all favorable, impacts on behavioral
outcomes were mixed. For example, mothers assigned
to Atlantas work-first program reported that their
young children had more positive behavioral/emotional outcomes in terms of less frequent externalizing
behavioral problems, on average, compared with the
control group. However, mothers in Grand Rapids
work-first program reported that their children had
more behavioral problems of this kind, on average,
The patterns of impacts on children within these programs varied considerably by the age of the child.
Among children under age five at the time of the follow-up, only weak or no impacts were found. However, among children ages 5 to 12 at follow-up, clear
patterns of favorable impacts emerged in the areas of
academic/cognitive development and, to a lesser
extent, behavioral/emotional adjustment. For example, in SSP, school-age children in the program group
scored higher on a test of math achievement, and their
Volume 12, Number 1
Janet Brown McCracken
mothers rated their academic achievement higher,

compared with school-age children in the control
group. In New Hope, teachers gave school-age boys in
the program group higher ratings for academics,24 as
well as for classroom behavior and independence. In
both versions of MFIP, long-term recipient mothers in
the program group rated their school-age children
higher for school performance and school engagement, and reported fewer behavioral problems.25 Confidence in the findings of favorable cognitive and
behavioral impacts for school-age children is bolstered
by the varied sources showing these impactsfrom
achievement test scores and teacher ratings to child
self-report measures and maternal ratings.
However, impacts were not uniformly favorable across all
the strong financial work incentive programs studied. In
particular, unfavorable impacts were found for adolescent
children ages 13 to 18 in MFIP and ages 12 to 18 in SSP,
and in some instances, for school-age children of recent
MFIP applicants. For example, in MFIP, the proportion
of adolescents whose mothers rated their school performance as above average was significantly lower among
both recent applicant families and long-term recipients
compared with the control groups. Also, a higher percentage of recent applicant mothers reported having
been contacted by their teens school about a behavioral
problem. In SSP, mothers in the program group rated
their adolescent childrens academic achievement slightly
less favorably, and reported more school behavioral problems for their adolescents, compared with mothers in the
control group. Adolescents themselves reported more
frequent smoking, drinking once a week or more, and
drug use, compared with adolescents in the control
group, and older adolescents reported more delinquent
activity. SSP researchers caution that there was attrition in
the adolescent sample, and that overall sample size for the
adolescent-reported impacts was small and perhaps
biased. Still, these findings raise the possibility that unfavorable impacts on adolescents may occur even when the
same programs have favorable behavioral impacts on
younger children.
With respect to health, very few significant impacts
were found in this set of programs, and the few that did
occur were in opposite directions. Among long-term
recipients in MFIP, an increased proportion of families
reported that a child had suffered an accident or injury
requiring a visit to an emergency room or clinic. This
might indicate either an increase in accidents or injuries
among these children or an increase in the use of emergency rooms by program group families. By contrast,
in SSP, program group mothers gave their school-age
children higher overall health ratings and were less likely to indicate that their children had long-term health
problems, compared with control group mothers.
In general, these studies help to identify the effects on
children when both employment and income increase
in the context of strong financial work incentive programs, and when such changes are fairly strong and
sustained. In the new policy context, numerous states
are using financial incentives to reward work.26 But
because current state welfare programs generally do
not provide incentives as generous as those offered in
the demonstration programs, the impacts on both
adults and children may be different. Nevertheless, the
findings emerging from these evaluations highlight the
potential of financial incentive programs to bring about
positive impacts for school-age children though perhaps leading to negative impacts for adolescents.
85
Across the different welfare-to-work programs studied so far,

there was neither widespread harm nor benefit to young
children. Yet some significant impacts on children did occur.
Programs with a Time Limit
Floridas Family Transition Program (FTP)27 was the single program among those discussed here to include a time
limit on welfare receipt. Within this program, mothers
increased their employment, earnings, and overall income,
but the impact on income was modest, and by the time of
the final follow-up four years later, the impacts on employment, earnings, and income had all disappeared. There
was even some indication that the proportion of families
with extremely low income may have increased.
Findings regarding FTP impacts on children were few
and mixed. For children ages 5 to 12 at the follow-up,
there were no impacts on academic/cognitive development, a slight indication of unfavorable impacts on
behavioral/emotional adjustment, and favorable
impacts on health and safety, according to maternal
reports. Researchers hypothesized that the favorable
health impacts may be related to the program groups
fewer reported housing problems, such as exposure to
dust or pollution that could trigger asthma.
As was the case for the programs with financial work
incentives, however, some unfavorable impacts
emerged for adolescents at the follow-up. Adolescent
children of mothers assigned to FTP were more likely
to have been suspended from school; however, they
were no more likely to have had contact with the
police, to have been arrested or convicted, or to have
had a baby. In addition, their mothers rated their
achievement slightly lower, although the adolescents
were not more likely to be below average in achievement or to be in special education.
Although no consistent or strong pattern of impacts
emerged among young children in families in FTP,
only about one-fifth of the families in the FTP study
had reached the time limit when follow-up interviews
were conducted. Moreover, results come only from a
single study, so that the consistency of findings across
programs could not be assessed. In the future, results
from two additional studies of programs with time limits will be available to help clarify the effects of this type
of program on children.
86
Nevertheless, this first look at child impacts in a program

with a time limit is particularly important when considering the effects of programs that combine enforcement
strategies with incentives and services. Such combinations are increasingly prevalent under the current welfare
law, which requires that states mandate employment participation and place time limits on welfare receipt. In
addition, this study sheds light on outcomes for children
when families experience a period of modest improvement in economic status that is not sustained, a common
experience among families transitioning off welfare in
the current policy and economic environment. (See the
article by Zedlewski in this journal issue.)
In sum, many of the results in the evaluations of these
different programs point only to weak impacts or an
absence of impacts for young children. Given the limited emphasis placed directly on services or program
components for children, perhaps this is not surprising.
Nevertheless, some impacts did occur that were both
favorable and unfavorable in each of the three aspects of
childrens development examined: academic/cognitive,
behavioral/emotional, and health and safety. Thus, the
programs did have the potential to affect childrens outcomes for both better and worse. Moreover, the magnitude of these impacts, when they did occur, fell within
the range of impacts for programs that target children
directly, but were not as strong as the largest impacts
found in some of the most successful early intervention
programs.28 As explored further below, various patterns
emerge from this synthesis of findings that help explain
the programs impacts on children.
Patterns Related to Favorable and

Unfavorable Impacts
Overall, this synthesis of findings suggests that across
the different welfare-to-work programs studied so far,
there was neither widespread harm nor benefit to
young children. Yet some significant impacts on children did occur. The question of whether and how different types of programs and different patterns of
economic impacts correspond to different patterns of
Volume 12, Number 1
impacts on children is of critical importance to policymakers. Of central concern is whether impacts on children correspond more closely to the type of program
(the programs blueprint), or to the actual economic results in terms of mothers employment and
income, regardless of a programs orientation or intent.
Box 2.) In addition, preliminary work suggests that

nontargeted family factors, such as parenting and
maternal mental health, also may play a role. These
influences differed across developmental areas, with
patterns generally similar for academic/cognitive and
behavioral outcomes, but not for health and safety.
In the studies reviewed here, four sets of factors appear

to tip program impacts on children toward favorable
or unfavorable: (1) program goals and components,
(2) the patterns of economic impacts, (3) family characteristics, and (4) child age. Favorable impacts on children tended to occur in programs that improved family
economic status or maternal education. Unfavorable
impacts on children tended to occur when families in
the program did not show economic progress or experienced a setback, when the children were adolescents,
andunexpectedlywhen the families were believed
to be at lower risk of long-term welfare receipt. (See
Program Goals and Components

The findings show relatively few impacts for children in
three of the four program types. Programs targeted to
teenage mothers, mandatory work-first or educationfirst programs, and the program with a time limit all had
limited impacts on child outcomes. Where patterns of
impacts did emerge, they were unfavorable in the teen
programs, but mixed in work-first and education-first
programs as well as in the time-limited program.
Favorable impacts were found much more consistently
for young children in the programs involving strong
financial work incentives, but even in these programs,
Box 2
Favorable and Unfavorable Impacts on Children
Favorable impacts tended to occur:
Unfavorable impacts tended to occur:
For school-age children in programs that resulted in improvements in family economic status, not only in terms of employment and earnings, but also in terms of overall family income
and proportion of families in poverty. This pattern of economic
impacts occurred most consistently, though not only, in programs that had strong financial incentives and supports for
working.
When families in the program did not show progress, on average, or experienced setbacks on any of the core economic outcomes (employment and earnings, overall income, or
proportion of families in poverty), despite program supports
and requirements.
On cognitive and academic outcomes for school-age children

in programs that resulted in increased maternal educational
attainment. This pattern of impacts for adults usually occurred
in programs that included education and training components
as precursors to job search and employment.
When children of the adults targeted by the programs were

adolescents. Unfavorable impacts for adolescent children of
welfare recipients occurred in programs taking very different
approaches (for example, programs emphasizing financial
incentives and supports for working, and programs focusing
more heavily on enforcement strategies such as sanctions and
time limits).
For children in families that were at lower rather than higher
initial levels of disadvantage. This unexpected pattern cut
across different ways of defining the initial level of disadvantage, and also across different program approaches.
87
The findings on economic patterns underscore the importance

of efforts to support sustained economic progress for families
in order to increase favorable impacts for children.
a consistent pattern did not emerge for all families. As
discussed further below, it is important to take into
account economic impacts and family and child characteristics to understand the variation in findings within and across program types.
Patterns of Economic Impacts

The correspondence between program type and pattern of economic impacts was not perfect for the programs examined here. Even within a single program
type, economic impacts for specific sites or program
variants sometimes differed. Similarly, the correspondence between patterns of economic impacts and child
outcomes also varied. The findings suggest that favorable impacts on income facilitate, but do not assure,
favorable child outcomes.
In general, young children in families making stronger
economic progress showed more favorable impacts on
their academic/cognitive and behavioral/emotional
outcomes, whereas children in families making no economic progress or experiencing a setback showed less
favorable impacts. However, for children in families
whose economic progress was somewhere in between
that is, with improvements that faded over time, or
increases in employment without increased income
the impacts on children tended to be weak. At the same
time, an unfavorable pattern of economic impacts did
not always correspond to unfavorable child impacts,
and similarly, a favorable pattern of economic impacts
did not always correspond to favorable child impacts.
For example, unfavorable impacts for adolescents cut
across different patterns of economic outcomes.
In addition, when mothers made progress in educational attainment, favorable child academic/cognitive
impacts were sometimes found even in the absence of
positive economic outcomes. For example, favorable
academic/cognitive impacts occurred in three of the
four JOBS programs in which mothers achieved higher
levels of education. One of the TPD programs in which
there were impacts on mothers education, TPD Chicago, also showed a weak pattern of favorable impacts for
child academic/cognitive outcomes. Not all programs
88
in which mothers made educational progress resulted in

favorable impacts on child academic/cognitive outcomes; for example, no impacts were found in the
JOBS education-first program in Riverside and the
TPD program in Camden. Nevertheless, these findings
suggest that improving the educational levels of mothers receiving welfare may have positive implications for
their childrens academic/cognitive development.29
For the most part, however, the findings on economic
patterns underscore the importance of efforts to support
sustained economic progress for families in order to
increase favorable impacts for children. The results also
show that in some instances, welfare-to-work programs
have had no or unfavorable economic impacts for families (especially families headed by teenage recipients), and
that such programs sometimes have had unfavorable
impacts for children. It is equally important to understand when and why these less-favorable economic
impact patterns occur and to seek alternative approaches.
Family Characteristics
Results taking into account family characteristics revealed
a pattern contrary to expectations. Findings suggest that
unfavorable impacts on children are more likely in families traditionally considered at lower risk for long-term
welfare receipt, such as families new to welfare, or those
who had received welfare for less than two years. In contrast, findings for children in families traditionally viewed
as higher risk, such as those who had received welfare
longer, show more favorable impacts for academic/cognitive and behavioral/emotional outcomes.30
Various explanations have been suggested as to why
lower family risk might be associated with unfavorable
child impacts. One hypothesis is that families traditionally viewed as lower risk are, in fact, experiencing acute
as opposed to chronic stress. For example, families new
to welfare may be applying because of a fairly recent
crisis, such as loss of a job, separation or divorce, or
domestic violence, or because of a major life change,
such as a babys birth. If such families are then encouraged or pushed toward work, this may add further
transitions to the lives of children already adapting to
Volume 12, Number 1
major changes, leading to unfavorable impacts. Another hypothesis is that applying for welfare may be associated with greater stigma for lower-risk families.31 As a
result, these families may have a greater sense of obligation and anxiety about needing to fulfill program
requirements, or a greater sense of shame about any
difficulty in meeting requirements. Such responses
could affect children, resulting in unfavorable impacts.
Unexpected findings were revealed at both ends of the
age range, for the youngest and the oldest children in
these studies. Other research suggests that changes in
family economic well-being, especially movement into
and out of poverty, as well as length of time in deep
poverty, appear to have the strongest effects on children
in the first years of life, particularly in the area of cognitive development.32 Thus, favorable impacts for children,
especially in the studies involving favorable economic
impacts, might be expected to be particularly evident for
very young children. Similarly, some evidence suggests
that maternal employment during infancy may sometimes have negative effects on children.33 However,
results pertaining to the youngest children in these studies do not support this hypothesis. In general, whereas
school-age children tended to show either favorable or
unfavorable impacts, the younger children in these studies showed little indication of significant impacts.
2001 Stephen Shames/Matrix
Child Age
age range, and the greater difficulty of obtaining reliable measures of their development. Similarly, the
impacts on adolescents were examined in only a minority of the studies. Still, these findings suggest that the
childs age is an important consideration across a range
of program approaches. Further investigation is needed of the impacts on infants and toddlers, as well as
adolescents, of differing welfare-to-work approaches.36
Nontargeted Family Factors
At the other end of the age continuum, some

researchers expected that adolescents would be the
least affected by their mothers involvement in a welfare-to-work program. Yet this set of evaluations points
fairly consistently to unfavorable patterns of impacts for
adolescents across program types and economic impact
patterns.34 Possible explanations for these impacts
range from an erosion in parenting quality and monitoring to a substantial increase in adolescent responsibilities within the household.35 Each of these
possibilities has different implications for programs
aimed at helping adolescents; therefore, careful consideration of the processes at play is warranted to mount
the most effective programs to address these issues.
Although these programs directly attempted to

improve adult economic outcomes such as employment, earnings, and income, an important question is
whether nontargeted aspects of family life, such as parenting behavior or maternal psychological well-being,
also were affected by the programs. Overall, nontargeted aspects of family life appear to have been affected less frequently than targeted economic outcomes,
yet such impacts did occur in these studies. Further,
some of the impacts occurred on outcomes that could
be very important for the quality of childrens experiences in the home, such as harsh parenting, maternal
depressive symptoms, residence of mother apart from
the child, and domestic violence.
At the same time, data are limited with respect to both

age groups. The very limited evidence of impacts on
the youngest children in these studies reflects a lack of
detailed and direct assessments of the children in this
Researchers are exploring whether and how impacts on

these nontargeted aspects of family life help to shape
program impacts on children. Preliminary work suggests that such impacts may play a role in determining
89
outcomes for children.37 Different programs appear to

activate different pathways of influence, however, and
results do not point to a single pathway across all studies. Indeed, it appears that child impacts for any one
program may reflect the simultaneous influences of
impacts on multiple targeted and nontargeted family
outcomes, sometimes functioning in opposing directions. (See the article by Huston in this journal issue.)
Differences across Developmental Areas

The patterns of influence relating to program type, economic impacts, family characteristics, and child age differed across developmental areas. In general, patterns
were similar for academic/cognitive and behavioral/
emotional outcomes, but not for health and safety.
Impacts on child academic/cognitive outcomes were the
most pervasive. Previous research suggests that childrens
academic/cognitive development may be particularly
affected by the cumulative influence over time of family
economic status. Thus, the greater impacts for academic/cognitive outcomes may reflect the fact that childrens
academic and achievement outcomes were influenced by
economic change over the follow-up period.
Impacts on child behavioral/emotional outcomes were
somewhat less consistent. Some evidence indicates that
behavioral outcomes in children are sensitive to transitions
and can change more rapidly.38 This sensitivity to change
may help explain the generally less pervasive patterns of
behavioral/emotional outcomes in this set of studies.
Impacts on child health and safety outcomes, such as
overall health ratings and incidence of accidents and
injuries, were rare for this set of evaluation studies.
When impacts did occur, they showed a very different
pattern from the other developmental areas, both in
terms of their frequency and their relationship to adult
economic impacts, suggesting different factors may be
at play. For instance, a significant factor that may influence health and safety measures may be whether parents have access to health care either through
employment or public assistance. Thus, the economic
patterns that help to explain the pattern of impacts for
academic/cognitive and behavioral/emotional functioning may be less important in explaining health and
safety outcomes; instead, health coverage, which was
not examined in detail here, may be key.
90
Different methods of measurement may also contribute

to the different patterns across areas of development.
Measures of academic/cognitive development, for
example, included standardized assessments and reports
from teachers and children, as well as from mothers. In
contrast, behavioral/emotional development was most
often measured only through maternal report. More or
different impacts might have been detected in the
behavioral/emotional area if information had been
obtained through other means, such as teachers
reports, or direct observation of childrens behavior.39
Similarly, it would also be helpful to go beyond maternal report measures of health and safety to include, for
example, a physicians assessment of the childs overall
health and of any limiting conditions.
Policy Implications
As reauthorization of welfare reform approaches, several implications from this synthesis of findings may be
of interest to policymakers at the federal, state, and
local levels who look to welfare policy as a means to
benefit children. In particular, the research suggests
that it may be useful to:
Focus on welfare programs that lead to sustained
financial gains for families. This analysis, along
with others, suggests that programs leading to sustained financial gains have the potential to improve
young childrens well-being, whereas programs that
fail to increase employment and income over the
long term are more likely to have neutral or negative
implications for children.
Take steps to identify and implement programs to
help families that have only recently begun receiving welfare or are considered to be at lower risk
of long-term welfare dependency. Somewhat surprisingly, children in these families seem to benefit
least, and indeed sometimes show unfavorable
impacts, when their mothers participate in welfareto-work programs. It is therefore important to consider ways to minimize these unintended negative
consequences for the families traditionally considered
at lower risk, and to reexamine the concept of risk
among families applying for welfare.
Explore the underlying causes of the initial indications of negative impacts of welfare-to-work
Volume 12, Number 1
programs on adolescent children and use this

information to implement better programs or
policies to help this age group. Few data are available to explain why adolescents score lower on various academic/cognitive and behavioral/emotional
measures, including their own self-reports of delinquent activities, when their mothers participate in a
welfare-to-work program. Before steps can be taken
to help, it is important to understand what is leading
to these negative effects.
Focus on how the new residence and education
requirements for teen parent recipients are affecting these families. Very little is known about how
these provisions are affecting teen parent families.
This is an important area for future experimental and
nonexperimental research.
Obtain more data on the effects on infants and
toddlers. Very little is known about how the
youngest children are affected when their mothers
are required to fulfill work requirements as a condition of welfare receipt. It would be extremely informative to systematically examine any variation in
impacts on infants and toddlers based on varying
policies concerning the age of the child when parents work requirements commence (for example,
varying the timing when recipients are required to
work to earlier versus later in an infants first year)
and the number of hours of work required for recipients with infants and toddlers.
Explore further the possibility that programs that
increase mothers education may benefit children.
Some programs studied in this analysis did not
increase family income during the time frame exam-
ined, but did increase mothers education and childrens academic/cognitive skills. Hence, these findings suggest another way that children might benefit
from their mothers participation in a welfare-towork program. The implications for children of programs combining education- and work-first strategies
may also be worth exploring.
Even when welfare-to-work programs have favorable
impacts on children, a high level of risk for poor developmental outcomes remains. The favorable impacts
occurring in these programs did not bring children to
the level of national norms on measures of
academic/cognitive development or behavioral/emotional adjustment. Thus, these programs by no means
offer a panacea for the developmental problems faced
by many low-income children or children whose parents receive welfare. Further, unfavorable impacts worsened already elevated levels of risk in some children.
These findings underscore the importance of strengthening program approaches that enhance developmental
outcomes for children in families being served by the
welfare system. Although improving child outcomes is
not typically the direct goal of welfare-to-work programs, much can be done to improve the way these
programs are structured to help ensure more positive
impacts for children.
The authors gratefully acknowledge the Edna

McConnell Clark Foundation and The David and
Lucile Packard Foundation for their support for
preparation of this article and the longer paper on
which it is based.
ENDNOTES
1. Zaslow, M., Tout, K., Smith, S., and Moore, K. Implications of
the 1996 welfare legislation for children: A research perspective.
Social Policy Report (1998) 12(3):134.
2. See reviews of some of the experimental studies in Duncan, G.J.,
and Chase-Lansdale, P.L. Welfare reform and childrens well-being.
In The new world of welfare. R.M. Blank and R. Haskins, eds.
Washington, DC: Brookings Institution Press, 2001, pp. 391417;
and Morris, P.A., Huston, A.C., Duncan, G.J., et al. How welfare
and work policies affect children: A synthesis of research. New York:

Manpower Demonstration Research Corp., 2001.
3. One of the programs studied, New Hope, was not restricted to
families receiving welfare, but many of the families had applied for
or were receiving welfare at the start of the evaluation. Thus, for
the sake of brevity, all programs considered here are referred to as
welfare programs.
4. Weil, A. Program redesign by states in the wake of welfare reform:
91
Making sense of the effects of devolution. In For better or for

G.J. Duncan and P.L. Chase-Lansdale, eds. New York: Russell
Sage Foundation, January 2002, pp. 6380.
5. For a more detailed review of the child impact findings in the
evaluations that formed the groundwork for this paper, see
Zaslow, M.J., Brooks, J.L., Moore, K.A., et al. Experimental
studies of welfare reform and children. Washington, DC: Child
Trends, 2001. Available online at http://www.childtrends.org.
6. A common set of child outcome and family process measures
across 5 of the 10 evaluations was created by Child Trends based
on the conceptual framework developed by the Project on StateLevel Child Outcomes team, building on work in the Child Outcomes Study of the National Evaluation of Welfare-to-Work
Strategies (NEWWS). The Project on State-Level Child Outcomes includes evaluations of waiver policies in Minnesota and
Florida (results included here), and Connecticut, Indiana, and
Iowa (results still pending). The evaluation teams for these waiver studies include researchers from the Manpower Demonstration
Research Corporation (for programs in Minnesota, Florida, and
Connecticut), Abt Associates (for the program in Indiana), and
Mathematica Policy Research (for the program in Iowa). Together with staff from the U.S. Department of Health and Human
Services and the NICHD Family and Child Well-Being Research
Network, researchers from Child Trends coordinate the Project
on State-Level Child Outcomes and have provided starting
points for discussions on conceptualization, measurement, and
analysis.
7. For a detailed summary of impacts, see note 5, Zaslow, et al.,
tables and appendices.
8. See Quint, J., Bos, J., and Polit, D. New Chance: Final report on
a comprehensive program for young mothers in poverty and their
1997; and Zaslow, M.J., and Eldred, C.A., eds. Parenting behavior in a sample of young mothers in poverty: Results of the New
Chance Observational Study. New York: Manpower Demonstration Research Corp., 1998.
9. Kisker, E.E., Rangarajan, A., and Boller, K. Moving into adulthood: Were the impacts of mandatory programs for welfare-dependent teenage parents sustained after the programs ended? Princeton,
NJ: Mathematica Policy Research, 1998.
10. Quint, J., and Musick, J. Lives of promise, lives of pain: Young
mothers after New Chance. New York: Manpower Demonstration
11. The tools used in these assessments were the Woodcock-JohnsonRevised Letter-Word Identification test and Calculation test.
12. Based on childrens evaluations of their teachers.
13. Based on mothers ratings using the Adaptive Social Behavior
Index.
14. In New Chance, teachers reported on a similar range of child
outcomes for a select subset of the sample. Interestingly, teachers
reported no impacts on the childrens behavioral or academic
functioning. The discrepancy in reports by teachers and mothers
might indicate that the program-related differences in childrens
behavior and indications of academic functioning occurred only
in the home, but not at school. Alternatively, the differences may
reveal changes in mothers perceptions of their children as a
result of the program, rather than differences in the childrens
actual functioning. Although the latter explanation suggests that
the impacts are not real, in the sense that they do not capture
actual changes in child functioning, parents who perceive their
92
children as behaving problematically or performing poorly in

school might treat their children less positively, which may have
important implications for their development.
15. In the NEWWS evaluation, a limited set of questions was asked
of mothers regarding all of their children. The impact results for
these questions are summarized in Freedman, S., Friedlander, D.,
Hamilton, G., et al. National Evaluation of Welfare-to-Work
Strategies: Evaluating alternative welfare-to-work approaches: Twoyear impacts for eleven programs. Washington, DC: U.S. Department of Health and Human Services and U.S. Department of
Education, 2000; and in Hamilton, G., with Freedman, S., and
McGroder, S.M. Do mandatory welfare-to-work programs affect
the well-being of children? A synthesis of child research conducted as
part of the National Evaluation of Welfare-to-Work Strategies.
Washington, DC: U.S. Department of Health and Human Services and U.S. Department of Education, 2000. This article
focuses on the results within each study that concern a particular
focal child, rather than results of items that pertain to multiple
children across a very wide age range. Accordingly, the summary
of NEWWS findings presented here comes from the child outcome studys results reporting on the focal child.
16. McGroder, S.M., Zaslow, M.J., Moore, K.A., and Le Menestrel,
S. National Evaluation of Welfare-to-Work Strategies: Impacts on
young children and their families two years after enrollment: Findings from the Child Outcomes Study. Washington, DC: U.S.
Department of Health and Human Services and U.S. Department of Education, 2000. See also Zaslow, M.J., McGroder,
S.M., and Moore, K.A. Summary report on the two-year follow-up
of the Child Outcomes Study: National Evaluation of Welfare-toWork Strategies. Washington, DC: U.S. Department of Health
and Human Services and U.S. Department of Education, 2000.
See also note 15, Freedman, et al.; and Hamilton, et al.
17. A second follow-up, five years after enrollment, has just been
released.
18. These impacts reflected higher mean scores on the Bracken Basic
Concept Scale School Readiness Composite for Atlantas work-first
component, and better distributional scores on this measure (that
is, fewer children scoring at the low end of the distribution or
more at the high end) for Atlantas education-first and work-first
components, and for Grand Rapids education-first component.
19. Gennetian, L.A., and Miller, C. Reforming welfare and rewarding work: Final report on the Minnesota Family Investment Program. Vol. 2, Effects on children. New York: Manpower
Demonstration Research Corp., 2000. See also Knox, V., Miller,
C., and Gennetian, L.A. Reforming welfare and rewarding work:
A summary of the final report on the Minnesota Family Investment
Program. New York: Manpower Demonstration Research Corp.,
2000.
20. Bos, H., Huston, A.C., Granger, R., et al. New Hope for people
with low incomes: Two-year results of a program to reduce poverty
and reform welfare. New York: Manpower Demonstration
Research Corp., 1999. See also Huston, A.C., Duncan, G.J.,
Granger, R., et al. Work-based antipoverty programs for parents
can enhance the school performance and social behavior of children. Child Development (2001) 72(1):31836.
21. Michalopoulos, C., Card, D., Gennetian, L.A., et al. The SelfSufficiency Project at 36 months: Effects of a financial work incentive on employment and income. Ottawa, Canada: Social Research
and Demonstration Corp., 2000. See also Morris, P., and
Michalopoulos, C. The Self-Sufficiency Project at 36 months:
Effects on children of a program that increased parental employment and income. Ottawa, Ontario: Social Research and Demonstration Corp., 2000.
Volume 12, Number 1
22. For example, MFIP generally provided earnings supplements

within the welfare system so that both income and welfare receipt
increased. In contrast, New Hope and SSP provided earnings
supplements outside the welfare system so that income increased
while welfare receipt decreased.
23. Only the recent applicant group within MFIP (who did not experience income gains or poverty reduction) and the participants in
the New Hope Project (who showed fade-out of the income
impact over time) depart from this pattern.
24. These ratings were from the Academic Subscale of the Social
Skills Rating System. Boys in the New Hope Project group were
also more likely to indicate that they expected to attend college,
and were more likely to have higher occupational expectations.
For girls in the program group, however, little impact on measures of cognitive development was found.
25. However, researchers found no favorable behavioral impacts for
school-age children of recent welfare recipients in the MFIP variant involving incentives only. See note 5, Zaslow, et al.
26. The most common financial incentive to reward work is the
earned income disregard, whereby a portion of earned income is
not counted when determining eligibility for benefits or calculating a familys welfare benefit. (See the article by Zedlewski in this
journal issue.)
27. Bloom, D., Kemple, J.J., Morris, P., et al. The Family Transition
Program: Final report on Floridas initial time-limited welfare program. New York: Manpower Demonstration Research Corp.,
2000.
28. See Cohen, J. Statistical power analysis for the behavioral sciences.
2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates, 1988.
Cohen offers statistical guidelines for estimating the practical significance of a relationship based on its effect size (here, the difference between the program and control group divided by the
standard deviation of the control group). He recommended that,
when comparing two groups, an effect size of .20 might be considered small, an effect size of .50 might be considered medium, and an effect size of .80 might be considered large.
Under these guidelines, when welfare programs do have impacts
on children, the impacts generally are in the small to medium
range, the same range as found for impacts in two-generational
programs and some child-focused programs.
29. This hypothesis is being explored further in work taking into
account factors that influence which mothers obtain further education in the context of the JOBS welfare-to-work programs.
Findings to date support the hypothesis that increasing mothers
educational attainment may be another means of improving outcomes for children in welfare-receiving families, specifically in the
area of academic/cognitive development. See Magnuson, K., and
McGroder, S.M. From ABEs to 1, 2, 3s: The effects of maternal
education on young childrens school readiness. Poster presented at
the annual meeting of the Population Association of America.
Washington, DC. March 2931, 2001.
30. Further findings from the Child Outcomes Study (within
NEWWS) and from FTP, not included in the synthesis, point to
similar conclusions.
31. This hypothesis was suggested by Kate Marsland, Laura Sosinsky,
and Sean Moundas of the Department of Psychology, Yale University, in response to a presentation by Martha Zaslow at the
Bush Center in Child Development and Social Policy, Yale University, March 2, 2001.
1997. However, few studies have investigated this point, and

those that exist are not all in agreement. For instance, another
study suggests that, whereas cognitive skills are most affected by
poverty in early childhood, academic achievement is more sensitive to poverty in adolescence. See Guo, G. The timing of the
influences of cumulative poverty on childrens cognitive ability
and achievement. Social Forces (1998) 77:25788. Yet another
study suggested that the relationship between early poverty and
childrens outcomes is an artifact of the relationship between the
timing and duration of poverty, with those experiencing poverty
at younger ages more likely to experience it for longer periods of
time. See Brooks, J.L., and Shanahan, M.J. The impact of family
economic history on pre-adolescent achievement. Unpublished
manuscript, 2000.
33. Baydar, N., and Brooks-Gunn, J. Effects of maternal employment
and child-care arrangements on preschoolers cognitive and
behavioral outcomes: Evidence from the children of the National
Longitudinal Survey of Youth. Developmental Psychology (1991)
27:93245. See also Belsky, J., and Eggebeen, D. Early and
extensive maternal employment and young childrens socioemotional development: Children of the National Longitudinal Survey of Youth. Journal of Marriage and the Family (1991)
53:10831110.
34. In two of the three programs with negative impacts on adolescents, MFIP and FTP, data were not available to indicate whether
the pattern of economic impacts differed for families with adolescents than for families with younger children. However, the SSP
findings suggest that economic and employment impacts are
identical in direction (though perhaps somewhat smaller) in families with older children and in families with younger children.
Hence, in the absence of data specific to families with adolescents, the pattern of economic impacts for families with younger
children is used here.
35. Brooks, J., Hair, E., and Zaslow, M. Welfare reforms impact on
adolescents: Early warning signs. Research brief. Washington, DC:
Child Trends, July 2001.
36. See, for example, Shonkoff, J.P., and Phillips, D.A, eds. From
neurons to neighborhoods: The science of early childhood development. Washington, DC: National Academy Press, 2000; and
Larner, M., Behrman, R., Young, M., and Reich, K. Caring for
infants and toddlers: Analysis and recommendations. The Future
of Children (Spring/Summer 2001) 11(1):719.
37. McGroder, S.M., Zaslow, M.J., Moore, K.A., et al. The role of
parenting in shaping the impacts of welfare-to-work programs on
children. In Parenting and the childs world: Influences on academic, intellectual, and social-emotional development. J.G.
Borkowski, S.L. Ramey, and M. Bristol-Powers, eds. Mahwah,
NJ: Lawrence Erlbaum Associates. In press.
38. Hofferth, S.L., Smith, J., McLoyd, V.C., and Finkelstein, J.
Achievement and behavior among children of welfare recipients,
welfare leavers, and low-income single mothers. Journal of Social
Issues (2000) 56(4):74774. See also Moore, K., Glei, D.,
Driscoll, A., et al. Welfare and poverty patterns: Implications for
children. Washington, DC: Child Trends, 2000.
39. Teachers reports for behavioral and emotional outcomes were
collected in two of the studies included in this synthesis of findings (New Hope and New Chance). However, the majority of
studies included in this synthesis relied solely on mothers reports
of childrens behavioral and emotional outcomes
32. See, for example, Duncan, G.J., and Brooks-Gunn, J., eds. Consequences of growing up poor. New York: Russell Sage Foundation,
93
APPENDIX
Profile of Welfare Studies Favorable and Unfavorable Impacts on Children

The following table displays the net proportion of statistically significant favorable and unfavorable impacts on childrens academic/cognitive,
behavioral/emotional, and health and safety outcomes, by patterns of impacts on family economic resources and parental education. Proportions are based on the number of favorable impacts, minus the number of unfavorable impacts, divided by the total number of child outcome
measures examined within the domain of development. An impact refers to the difference in the average level of functioning of children in the
program and control groups. (Ages of children at the time of the follow-up study in parentheses.)
Adult Outcomes
Child Outcomes
Economic
resourcesa
Educational
attainmentb
New Chance (ages 3.510)c
No improvement/setback
No differenced
TPD (Camden) (ages 510)

TPD (Chicago) (ages 510)
TPD (Newark) (ages 510)
Sustained improvement
Employment only
Academic/
cognitive
Behavioral/
emotional
Health and
safety
0.2
0.39
0.1
More
More
Less
0
0.06
0.19
0
0
0.11
0
0
0
JOBS/education-first (Atlanta) (ages 57)

Employment only
JOBS/education-first (Grand Rapids) (ages 57) No improvement/setback
JOBS/education-first (Riverside) (ages 57)
Employment only
More
More
More
0.25
0.25
0
0
0
0
0
0
0.67
JOBS/work-first (Atlanta) (ages 57)

JOBS/work-first (Grand Rapids) (ages 57)
JOBS/work-first (Riverside) (ages 57)
Employment only
Employment only
Employment only
More
Less
No difference
0.75
0
0
0.09
0.18
0
0
0
0.67
MFIP/LT/F (ages 512)e

MFIP/LT/I (ages 512)
MFIP/LT/F (ages 1318)
No difference
More
No difference
0.6
0.4
0.25
0.3
0.5
0
0.5
0.5
MFIP/R/F (ages 512)

MFIP/R/I (ages 512)
MFIP/R/F (ages 1318)
Employment only
Employment only
No difference
No difference
No difference
0
0.4
0.75
0.25
0
-1
0
0
New Hope/boys (ages 35)

New Hope/boys (ages 612)
New Hope/girls (ages 35)
New Hope/girls (ages 612)
No difference
No difference
No difference
No difference
0
0.27
0
0.03
Programs for Teenage Welfare Recipients
Mandatory Education-First and

Work-First Programs
94
0.53
f
0.06
f
f
f
Volume 12, Number 1
Adult Outcomes
Child Outcomes
Economic
resourcesa
Educational
attainmentb
Fade-out
Fade-out
Health and
safety
Academic/
cognitive
Behavioral/
emotional
0
0.29
0.33
0
0
0.63
0
0.5
0
0
0.33
0.13
0.2
0.67
Strong Financial Work Incentive Programs (continued)

SSP (ages 35)
SSP (ages 611)
SSP (ages 1218)
f
f
Programs with a Time Limit Component

FTP (ages 512)
FTP (ages 1317)
a Economic resources refers to the pattern of impacts on average parental employ-
ment and average family income.

No improvement/setback: The program group did not have a higher average level
of employment or income, or had a lower level of employment or income, than
the control group.
Employment only: The program group had a higher average level of employment,
but not income, than the control group.
Fade-out: The program group had a higher average level of both employment and
income than the control group, but at least one of these effects faded out by
the end of the follow-up.
Sustained improvement: The program group had a higher average level of both
employment and income through the end of the follow-up period.
b Educational attainment refers to the pattern of impacts on average parental edu-
cational advancement.
More: The program group had a higher average level of educational attainment
than the control group.
Less: The program group had a lower average level of educational attainment
than the control group.
No difference: No statistically significant difference was found between the average level of educational attainment in the program and control groups.
completion, but a negative impact on high school completion. These mixed

results are summarized here as no impact.
e MFIP data are reported separately for mothers who had been receiving welfare
for 24 of the 36 months immediately prior to random assignment (long-term

recipients/LT) versus those who did not meet this criterion (recent applicants/R);
also, data are reported separately for mothers who received financial incentives
only (incentives/I) versus those who were enrolled in the full program with both
financial incentives and a work mandate (full/F). For long-term recipients, the
work mandates in the full program went into effect immediately upon assignment to the program. For recent applicants, however, the work requirement did
not go into effect until the recipient had been receiving welfare for 24 months.
The MFIP results are presented for single-parent recipients in urban counties
only.
f No assessment of this indicator in this study
Acronyms: FTP = Floridas Family Transition Program

JOBS = Job Opportunities and Basic Skills Training
MFIP = Minnesota Family Investment Program
SSP = Canadas Self Sufficiency Project
TPD = Teenage Parent Demonstration
c Includes results from all program sites: 16 locations across 10 states.

d New Chance had a positive impact on General Educational Development (GED)
For a more detailed discussion of these findings, see the Child Trends Web site at
http://www.childtrends.org.
95
Welfare Reform and Child Care

Options for Low-Income Families
Bruce Fuller, Sharon L. Kagan, Gretchen L. Caspary, and Christiane A. Gauthier
SUMMARY
For the changes under welfare reform to positively affect children, the gains that mothers
make from employment must lead to improvements in childrens daily settings at home, in
child care, at school, or in the community. This
article focuses on the role child care can play in
promoting the development of, and life opportunities for, low-income children. Key observations include:
Total federal and state funding for child care
for welfare and working poor families has increased dramatically since welfare reform,
from $2.8 billion in 1995 to $8.0 billion in
2000.
The majority of welfare mothers tend to rely
on informal child care arrangements when
first participating in welfare-to-work programs, but as they move off welfare and into
more stable jobs, they are more likely to
choose a center or a family child care home.
Although children from poor households
stand to benefit the most from high-quality
care, they are less likely to be enrolled in highquality programs than are children from affluent families, partly due to uneven access to
high-quality options in their neighborhoods.
Less than one-quarter of all eligible families

use child care subsidies, and usage varies widely across states and local areas reflecting various barriers to access and scarcity of quality
center-based care.
The authors conclude that to achieve welfare
reforms ultimate goal of breaking the cycle
of intergenerational poverty and dependence
on government benefits, welfare-to-work
programs should promote learning and
development among children in welfare and
working poor families by increasing access to
high-quality child care in low-income neighborhoods.
Bruce Fuller, Ph.D., is professor of education and public policy, and director of Policy Analysis for California Education (PACE) at the University of
California, Berkeley.
Sharon L. Kagan, Ed.D., is professor of early childhood and family policy, and codirector of The Center
for Children and Families at Teachers College,
Columbia University.
Gretchen L. Caspary is assistant director of the Growing Up in Poverty Project at PACE.
Christiane A. Gauthier, M.S.Ed., is field operations
coordinator at PACE.
97
Fuller, Kagan, Caspary, and Gauthier
central goal of the welfare reforms undertaken in the 1990s was to increase parents
self-sufficiency and end dependence on
government benefits.1 For this goal to be
realized, not just for the current generation but also for
the next, attention must be paid to the early development and long-term advancement of children in welfare and working poor families. Mothers employment
gains are of little consequence to childrens development unless such gains lead to improvements in childrens daily environments at home, in child care, at
school, or in the community.
This article focuses on the effects of welfare reform on
how and where low-income children spend their days,
and on the role child care can play in improving their
lives. The first section reviews the history of public
interest and support for child care. The second section
examines patterns of child care use among low-income
families, changes in family life spurred by welfare
reform, and factors affecting parents choice of care.
The third section summarizes what is known about the
quality of care in various settings and how the quality
of care affects childrens development. The fourth section discusses strategies for crafting more effective policies to advance child care options for low-income
families. Finally, the article concludes with some
thoughts about steps needed to help achieve the policy aim of ending the inheritance of family poverty.
The Public Interest in Child Care

Society has a stake in families child care choices, both
because child care enables parents to work and because
it can influence childrens development. Separate
strategies and funding streams have evolved over the
past century in response to each of these concerns.
The settlement house movement, which began in the
late 1800s, included a push to expand child care centers for single mothers who had to work. Congress
redoubled this effort during World War II, rapidly
expanding center-based programs for female factory
workers when the labor power of young mothers was
sorely needed.2 A parallel effort focused on providing a
wholesome environment for children in poverty. This
movement first emerged in the 1930s, when federally
funded nursery schools were established to create jobs
98
for unemployed teachers, nurses, and others.3 Statefunded preschools emphasizing early education and
school readiness evolved out of this tradition, most
notably Head Start, a child development program created in 1965 to serve low-income children and their
families. Then in 1988, Congress enacted three welfare-related child care programs to subsidize care as a
support for parents who were engaged in work preparation activities or work itself, and who were on welfare, leaving welfare, or at risk of becoming dependent
on welfare. In 1990, Congress also created the Child
Care and Development Block Grant to subsidize child
care for a wider range of low-income working parents.
The welfare reform law of 1996 enacted further
changes to federal child care programs. Growing out of
an interest in enabling work, but touching on concerns
Box 1
New Goals in the 1996 Law for
Federal Support of Child Care
To allow maximum flexibility for states to develop child
care programs and policies that best suit the needs of children and parents in each state.
To promote parental choice and empower working parents
to make their own decisions about the child care that best
suits their familys needs.
To encourage states to provide consumer education information to help parents make informed choices about child
care.
To assist states to provide child care to parents trying to
achieve independence from public assistance.
To assist states in implementing the health, safety, licensing, and registration standards established in the states
child care regulations.
Source: Personal Responsibility and Work Opportunity Reconciliation Act of
1996. Public Law 104-193, section 602, 110 Stat. 2279, August 22, 1996.
Volume 12, Number 1
Child Care Options for Low-Income Families
for childrens development, the Child Care and Development Block Grant was expanded and consolidated
with the other welfare-related funding streams
described above. (See the article by Greenberg and colleagues in this journal issue.) The new goals established
for the expanded block grant, referred to in federal regulations as the Child Care and Development Fund
(CCDF), are summarized in Box 1. In addition to
increasing funds for child care, the law also allows
states to spend funds allocated to the new welfare program, Temporary Assistance for Needy Families
(TANF), directly for child care, and to transfer up to
30% of their TANF funds into the CCDF.
Meanwhile, spending on preschools and early education programs also increased. Federal spending on
Head Start preschools, for example, grew from $1.2
billion in 1990 to $5.3 billion in 2000 ($3.8 billion in

constant 1990 dollars).4 The Early Head Start program
was established in 1994, and preschool support from
Title I of the Elementary and Secondary Education Act
began to grow rapidly in the mid-1990s as well.
Figure 1 summarizes the growth in federal appropriations for major child care and early childhood programs over the past decade. Only the federal
Dependent Care Tax Credit, a nonrefundable tax credit for taxpayers who pay out-of-pocket for child care,
declined during this period.5 The use and significance
of this tax credit are likely to increase, however, as the
Bush administration has agreed to make the credit
refundable beginning in 2002.
In addition, with their added flexibility under TANF,
some states have aggressively reallocated welfare dollars
Figure 1
Head Start
Rising Federal
Commitment to
Child Care and
Early Education
Early Head Start and Even Start

Child Care and Development
Block Granta
ESEAb-Title I early education
Child care food program
TANF transfers to child carec
KEY:
FY 1991
FY 2000
Dependent Care Tax Credit

|
Billions of 1990 dollars, adjusted for inflation

a
Fiscal year 1991 funding for the Child Care and Development Fund reflects the sum of the amounts provided for various programs that were consolidated into the fund in 1996, including AFDC-related child care and the Child Care and Development Block Grant.
b
ESEA = Elementary and Secondary Education Act
TANF transfers to child care includes direct spending on child care as well as transfers to state Child Care and Development Funds.
Sources: U.S. General Accounting Office. Education and care: Early childhood programs and services for low-income families. HEHS-00-11. Washington, DC: GAO, 1999; Office
of Management and Budget. The U.S. budget, fiscal year 1992. Washington, DC: Government Printing Office, 1991; Hayes, C., Palmer, J., and Zaslow, M. Who cares for Americas
children? Washington, DC: National Academy Press, 1991; U.S. General Accounting Office. Preschool education: Federal investment for low-income children significant but
effectiveness unclear. GAO/T-HEHS-00-88. Washington, DC: GAO, 2000; and personal communication with Sandy Brown, U.S. Department of Education, and Nazanin Samari,
American Enterprise Institute (who shared historical data assembled by Douglas J. Besharov).
99
to child care and after-school programs. Total federal

and state expenditures for child care under the CCDF
and welfare-related programs grew from $2.8 billion in
1995 to $8.0 billion in 2000, including $2 billion in
funds transferred from TANF.6
programs for children ages 3 to 6 grew from just $180

million in 1987 to over $2 billion in 1999. Georgia is
the only state to provide universal access for all fouryear-olds whose parents seek preschool programs, but
state-funded programs serve sizeable shares of lowincome children in California, Maryland, Massachusetts, New York, and North Carolina.7 According to a
recent report from the National Center for Children in
Poverty, total state funding for early childhood initiatives, including infant and toddler programs and an
States have also stepped up their funding for early education. At least 43 states now support preschool programs for low-income families, enrolling more than
750,000 children. State funding for early education
Figure 2
Types of Child Care Use by Poverty Status, Fall 1995
Data include children from all families, regardless of parents employment status.
Preschool
10
KEY:
10
0
Tota
l rel
ative
care a
Tota
l no
nrel
ative
care b
Orga
nize
d fa
cility
Othe
r no
nrel
ative
s
Tota
l rel
ative
care a
Desi
gnat
ed p
aren
t
Othe
r pa
rent
Gran
dpar
ent
20
In poverty
Self
-car
e
20
30
Tota
l No
nrel
ative
care b
Enri
chm
ent
acti
vitie c
s
30
40
Gran
dpar
ent
40
50
Sibli
ng
Percentage of Preschoolers
50
60
Othe
r pa
rent
Percentage of School-Age Children
Grade School
60
Not in poverty
Total includes care by designated parents, other parents, grandparents, siblings, and other relatives.
Total includes care in an organized facility or by other nonrelatives.
Enrichment activities consist of organized sports, lessons (such as music, art, dance, language, and computer), clubs, and before- or afterschool programs.
Source: Based on Census Bureau data from the Survey of Income and Program Participation, 1993 panel wave 9, as summarized in Smith, K. Whos minding the kids? Child
care arrangements: Fall 1995. Current Population Reports, Household Economic Studies, P70-70. Washington, DC: U.S. Census Bureau, October 2000, tables 4 and 11. Because
of multiple arrangements, total percentages may exceed the total number of children.
100
Volume 12, Number 1
Figure 3
Child Care Settings by Poverty Level in Four States (1997)
Percentages are based on
the number of children
under age five with
employed mothers.
Families above twice

the poverty line
Alabama
California
13%
50%
20%
Massachusetts
16%
24%
10%
Texas
41%
12%
KEY:
% in child care
centers
% in family
child care homes
Families below twice

the poverty line
Alabama
California
15%
17%
16%
Massachusetts
Texas
Source: Compiled from data from

the National Survey of Americas
Families, Urban Institute, 1997.
array of child development and family support efforts,

exceeded $3.7 billion in 2000.8
Child care funding at both the state and federal level
has risen significantly, and children are spending
increasing amounts of time in care, but the role child
care plays in the lives of children and parents is not well
understood. The remaining sections of this article
examine the implications of welfare reforms changes
to child care as an increasing number of low-income
mothers move into jobs.
Patterns of Child Care Use among

Low-Income Families
At the dawn of welfare reform in the mid-1990s, a fair
amount of research had examined the type of child care
relied upon by low-income families. As illustrated in
Figure 2, national data indicate that families living
below the poverty level relied heavily on relatives to
care for both their preschoolers and school-age children in 1995, as did families living above poverty.
However, families who were better off used nonrelative
24%
38%
28%
9%
10%
Percentage of Preschool Children
care almost as frequently as relative care for their

preschoolers, interspersing different types of care.9
Research shows, for example, that more than 70% of
four-year-olds from affluent families were enrolled in a
center or preschool in 1995, compared with 45% of
those from low-income households.10 Affluent families
were also much more likely to provide multiple types
of enrichment activities for their children in grade
school. School-age children in households earning
more than $55,000 a year were almost three times as
likely to participate in sports, and more than twice as
likely to take lessons after school, than were children
from families earning under $18,000 annually.11
Studies show that the types of care families select also
vary widely across states. To illustrate these betweenstate differences, Figure 3 shows the share of preschool
children who attended a center-based program or family child care home in 1997 among families with a working mother in four different states, by poverty level. The
share of children from households earning less than
twice the poverty line who attended centers ranged
from 17% in California to 38% in Massachusetts.12
101
Changes in Child Care Spurred by Welfare Reform

Before welfare reform, a significant number of mothers
living in poverty cared for their children themselves, as
they could usually rely on welfare without working.
Caseload data from 1995 indicate that only about 20%
of families receiving welfare were either engaged in
work activities or employed, and 4.9 million families
(monthly average) were on the rolls.13 But the 1996
federal welfare law required for the first time almost all
parents, including those with preschool-age children
and younger, to participate in work activities.14 By
1998, the percentage of families engaged in work activities or employed had grown to 35%, and only 3.2 million families were on the rolls.13 As welfare mothers
moved into jobs, their childrenespecially preschoolage childrenmoved into nonmaternal child care
arrangements. Many welfare mothers did not have a
child care provider prior to the welfare-to-work
requirements, and even for mothers who did, the number of hours their children spent in nonmaternal care
likely increased.15
A BerkeleyYale research team estimated that at least

one million preschool-age children moved into new
child care settings between 1996 and 1998, following
changes under welfare reform.16 This estimate may be
conservative, as it includes only families who were
enrolled in work activities and those employed about a
year after leaving welfare. Also, the estimate did not
include the increasing number of older children who
began spending time at home alone after school while
their mothers were still at work.
Data on the types of child care selected by welfare parents are emerging from a number of studies in several
states and cities across the country. The majority of
welfare families rely on informal arrangements when
they begin to participate in work activities. For example, in Vermont, most of the growth in use of nonmaternal child care following welfare reform involved
relatives and informal providers, up 26% in the early
years of the states welfare-to-work demonstration program, whereas use of licensed centers and family child
care homes increased by only 5%.17 But in some cities
Box 2
The Growing Up in Poverty Project
The Growing Up in Poverty (GUP) Project is a five-year longitudinal
study, launched in 1997, to learn how childrens upbringing and
development may be affected by the push to move mothers from
welfare into the workforce in the wake of welfare reform.
Researchers are tracking 948 single women who have preschoolage children and who participate in welfare-to-work programs.
The projects major goals are to measure the effects of welfare
reform on children and their mothers; to assess the type and quality of child care used by families receiving TANF; to determine how
differences in neighborhoods affect young children; and to make
recommendations for the next generation of welfare reforms.
Participants were recruited from welfare-to-work programs in
five cities across three states: San Francisco and San Jose, California; Tampa, Florida; and New Haven and Manchester, Connecticut. Each mother provided responses in a detailed interview
covering issues such as parenting, home environment, sources of
102
income, living costs, and stress. After mothers became employed

or began job training, information was gathered on the child care
settings chosen for their children, and the child care providers
were visited while the child was present to observe the setting
and interview the primary provider. The second year of data collection, completed in 2000, focused in greater depth on the childrens home environment, mothers experiences in the job
market, and the effect of those experiences on motherchild
relationships.
The GUP Project is run jointly by the University of California at
Berkeley, Yale University, and Teachers College, Columbia University, in collaboration with Mathematica Policy Research, Inc., and
the Manpower Demonstration Research Corporation. The project
receives support from nine foundations and government research agencies. For more details, see the GUP Web site at
http://pace.berkeley.edu.
Volume 12, Number 1
Data from early welfare-to-work programs show that when

participation in such programs is complemented by strong child
care and after-school support for families, use of center-based care rises.
and counties across the country, more than 40% of welfare families with preschool-age children select centerbased care as they move into work activities.18
The Growing Up in Poverty (GUP) Project is detailing the
wide variability in child care histories of young children in
California, Connecticut, and Florida. (See Box 2.) The first
wave of maternal interviews and child care observations,
collected in 1998, revealed that 70% of participating Florida mothers selected center-based care after entering welfareto-work programs, compared with the 29% and 13% of
mothers who selected centers in California and Connecticut, respectively.19 This suggests that local implementation
and center supply conditions may be important factors
affecting mothers child care selections.
Moreover, data from early welfare-to-work programs show that when participation in such programs is complemented by strong child care and
after-school support for families, use of center-based
care rises. For instance, in Minnesota, the selection
of centers or family child care homes was significantly higher for families participating in the welfare-to-work program (the experimental group)
than for nonparticipating families (the control
group): 53% versus 42%, respectively.20 The majority of welfare parents not in the experimental group
continued to rely on parental care, relatives, or
other informal child care arrangements.
Data from the National Evaluation of Welfare-to-Work
Strategies (NEWWS) identified several maternal and
family attributes that helped to predict which welfare
mothers would select centers rather than home-based
arrangements for their three- to five-year-olds. Most
important were higher levels of maternal education and
employment, a higher level of cognitive stimulation
provided in the home, and residence not in public
housing.18 These findings are similar to results from the
GUP Project that showed that welfare mothers with
stronger labor force experience were more likely to
select center-based care, after taking into account child
age and a variety of other factors.
Furthermore, working poor families have been found to

choose center-based care more often than welfare families do. As parents move off welfare and into more stable, full-time employment, they are more likely to
choose centers over less formal types of care. For example, among welfare recipients in Los Angeles with children ages 2 to 4, 27% selected center-based care in
1999; 51% relied on informal care, and the remaining
22% used a family child care home.21 In contrast, working poor parents in Los Angeles with similar access to
subsidies tended to select centers more often. About
59% selected center-based care, just 15% used informal
care, and the remaining 26% selected a family child care
home. For school-age children, 42% of parents in the
county welfare system relied on informal care; among
working poor parents, just 16% used informal providers.
Factors Affecting Choice of Care

Even as Head Start, state-funded centers, and preschool initiatives are expanding in many neighborhoods, many low-income parents continue to rely on
informal arrangements with relatives, neighbors, or
babysitters (often referred to as kith and kin) for
child care. Some analysts argue that low-income parents hold an a priori preference for informal child care
arrangements. Evidence suggests, however, that other
factors also play a role.
Over the past decade, much has been learned about the
factors that influence parents propensity to use child
care and the type of care they select. Mostly this work
has drawn from national samples of parents; only a portion has centered on low-income families. Findings
from these studies suggest that the age of the child,
trust and flexibility, cost, and accessibility figure prominently in parents decision making about their childrens care.
Age of the Child
Studies show that parents likelihood of selecting formal
care varies dramatically by the childs age. According to
1995 nationwide data, only 19% of families used formal
care for their children less than age 1 during the hours
a parent was engaged in school or work, but 50% use
103
The Image Bank/Ross Whitaker
child care. Interviews with these mothers suggest that

they often trust kin members or friends more than center-based caregivers because kith and kin offer familiar
child-rearing practices and speak their language, both
figuratively and literally. In addition, kith and kin often
have more flexibility than other providers to care for
children early in the morning or later in the evening,
which is important for many low-income mothers who
work odd-hour shifts.
such care for their children ages 3 to 4. For school-age

children, while only 3% of families used a formal child
care facility, 35% used some type of enrichment activityincluding sports, lessons, clubs, or before- and
after-school programsto help cover these hours, and
over 20% left their children in self-care at times.22
Similar patterns are reflected in studies of parents participating in welfare-to-work programs. For example,
among the subgroup of welfare families who participated in Californias GAIN program (a precursor to
contemporary welfare-to-work programs), most relied
on kith and kin to care for their infants and toddlers.23
Only 23% of those with children ages 0 to 2 selected
center-based care, compared with 47% of those with
children ages 3 to 5. Those with school-age children
also relied more heavily on centers or after-school programs than on kith and kin.
Trust and Flexibility
Many welfare-to-work programs require quick entry
into orientation sessions, job clubs, or job search activities, so mothers entering these programs must rapidly
find a trusted organization or individual to provide
104
In the GUP study, for instance, mothers entering new

welfare programs in 1998 were asked to rank the flexibility, trustworthiness, and interpersonal openness of
their child care provider, as well as the extent to which
their child received individual attention. Mothers
scored kith and kin higher than centers on all four
dimensions.24 Other studies including interviews with
women on welfare confirm this trust in and flexibility
of kith and kin, especially when it comes to care for
infants and young toddlers.25
Language concerns, in particular, may affect mothers
trust in informal arrangements. The GUP study found
that members of language minority groups (Latinas and
Vietnamese Americans) are less likely to select centerbased care.26 Also, it appears that welfare mothers are
less apt to use centers when they can rely upon more
supportive kin or coresident adults for their childrens
care. At the same time, while Latina mothers in the Los
Angeles welfare system were less likely to select centers
(33%) than were Anglo clients (45%), such ethnic differences were not found among the working poor.21
Cost
The cost of child care is a significant consideration for
all families, but especially for low-income families. A
survey of welfare parents in Illinois, conducted in
1990, revealed that 81% worried about the cost of
child care and just over half said they had serious difficulty finding a caregiver.27 Formal care is generally
more costly than informal care, which is often unpaid.
Thus, many low-income families require a subsidy to
gain access to center-based care or a family child care
home. Data from the GUP Project corroborate this
point. Researchers examined the flow of subsidies to
single mothers who selected a child care provider after
entering welfare-to-work programs in California and
Florida, and found that subsidy use was heaviest
Volume 12, Number 1
Unless the full range of child care options is truly available and
affordable, low-income parents continued reliance on kith and
kin cannot necessarily be interpreted as their true preference.
among mothers who selected center-based care. Nearly 90% of mothers using centers received a full subsidy,
whereas only 39% of mothers selecting a home-based
setting received a subsidy.28
When welfare or working poor parents do not receive
a subsidy or cannot find publicly supported child care,
they must pay out-of-pocket for whatever type of care
they use, and these costs can be substantial. Results
from the Urban Institutes National Survey of Americas Families found that a third of all working parents
who had children under age 13 paid for child care.
Among those who paid, parents spent, on average,
$190 per month (or 23% of earnings annually) on care
in 1997.29 For low-income families with a working
mother, the percentage of family income spent on child
care is even higher. According to 1995 survey data
from the U.S. Census Bureau, among families who had
an employed mother and paid for child care, those
earning less than $18,000 a year spent an average of
30% of their annual income on care. In contrast, affluent families (those with annual incomes of approximately $54,000 or more) spent only 5% of their
income, on average, for child care.30
Given that the costs of child care can quickly become
substantial, it is not surprising that many low-income
parents put their names on waiting lists for subsidies
and vouchers. For example, in a random survey of parents on such lists in Santa Clara County, California,
researchers found hundreds of working poor mothers
waiting for a subsidized child care slot or voucher.31
Two in five expressed concerns about the quality of
their current child care provider and were eager to
obtain support in order to afford another caregiver.
Similar findings emerged from a study of families on
waiting lists in Minnesota.32
Accessibility
Neighborhood conditions and basic access to particular child care options also shape parents choice of child
care. When centers are available in poor neighborhoods, parents choose this form of care more frequently, especially as their children reach age three or
four. When the supply of centers or family child care

homes is scarce, working mothers must rely on kith or
kin for child care or forgo their jobs.
The stock of child care organizations that has sprouted
within states and neighborhoods varies remarkably. For
example, the GUP Project studied provider markets
across five counties in California, Connecticut, and
Florida. The data suggest that the differences in child
care selection patterns by welfare mothers could be
explained, in part, by the differences in per capita supply
of slots in centers and family child care homes in neighborhoods where the mothers resided. For example, the
low use of center-based care in Connecticut is due, in
part, to the low supply of centers in the research sites of
New Haven and Manchester.28
An analysis of California zip code data also found a
close association between the share of welfare parents
who selected a center or family child care home and the
per capita supply of these organizations in the surrounding communities.21 And both the use and supply
of formal care settings were found to be closely associated with average levels of maternal education, as illustrated in Figure 4. Other data corroborate that
mothers with higher education levels are more likely to
choose centers and preschools over kith or kin.33 At the
same time, maternal education is highly correlated with
maternal employment rates and income levels. These
may be the underlying factors driving both the
increased demand for center-based care and the greater
supply of centers in neighborhoods with higher maternal education levels.
In sum, many factors help to explain welfare parents
selection of care for their children, including the age
of the child and the mothers level of trust and education. But as welfare parents enter the workforce and
their incomes rise, so does the likelihood that they
will choose a more formal child care arrangement
either a center or a licensed family child care home.
Still, parents choices concerning child care are influenced by the cost and accessibility of various options
within their neighborhoods. Unless the full range of
105
child care options is truly available and affordable,

low-income parents frequent reliance on kith and kin
cannot necessarily be interpreted as their true preference.
tics of the child care setting, such as the child-to-adult

ratio, the size of each group of children, and the formal
education and training of caregivers. These two aspects
of quality are often interrelatedthat is, high process
quality tends to be associated with high-quality structural characteristics.34
The Effects of Child Care Quality on

Development
Of all the quality indicators, the nature of the daily

interaction between the child and the caregiver has
been found to be very important.35 High-quality interactions are characterized by sensitivity and responsiveness, generous amounts of attention and support, and
high levels of verbal and cognitive stimulation. Compared to children in settings with less engaged caregivers, children with high-quality daily interactions
tend to display stronger cognitive and language development, school readiness, and early school achievement. High-quality care has been found to be
especially effective in improving academic outcomes
for children growing up in poverty or facing other risks
at home. Effects on childrens social development have
proven more elusive to discern, however.
As children, especially younger children, spend increasing amounts of time in child care, concerns have been
raised about the effects of child care on childrens wellbeing. Studies show that total hours of care, stability of
care, and the type of care all can have effects on childrens development, but the quality of care has by far
the greatest influence.
Elements of Quality
Over the past two decades, researchers have explored
the quality of various child care settings and its effects
on children. Key indicators of quality include the relationship between the child and the caregiver (referred
to as process quality) and the structural characteris-
Figure 4
Neighborhood Supply of Center-Based Care in California by Maternal Education Levels
KEY:
Low maternal schooling level
High maternal schooling levelc
a
Enrollment capacity is represented as the
percentage of children ages 2 to 5 for centers,
and children under age 13 for family child care
homes. Neighborhoods (zip code areas) with
fewer than 10 preschool-age children or with
no child care capacity in 2000 were excluded.
b
Neighborhoods with the lowest quartile of

average maternal education levels.
35
Neighborhoods with the highest quartile of

average maternal education levels.
Enrollment capacitya
40
b
35.0
33.7
30
25
20
15
15.0
15.0
12.3
11.2
10
7.7
5.8
5
0
Center capacity
1996
Center capacity
2000
Family child care

home capacity
1996
Family child care

home capacity
2000
Source: Compiled from data from the California Child Care Resource and Referral Network.
106
Volume 12, Number 1
The Quality of Different Types of Care

All types of child care span the range from high to
poor quality. Both center-based and home-based care
settings can at times be unstimulating, disorganized,
and even unclean or unsafe. Few studies have examined the quality and character of kith and kin
arrangements, but a consistent finding across the
small number of multicity studies that have been conducted indicates that home-based settings typically
lack the breadth of learning and play materials offered
in centers, and caregivers typically are less well educated. Researchers in one study rated about half of
the home-based settings they observed as displaying
fair quality or worse.36
Other studies have found home-based care to be better than center-based care in some situations. For
example, a study by the National Institute of Child
Health and Human Development (NICHD) has
tracked more than 1,000 children, mostly from middle
class and affluent families, to analyze their developmental progress and their care arrangements from
birth to age three. The study team reported that inhome caregivers for these infants and toddlers provided the most positive caregiving, whereas center-based
care with higher ratios of children to adults provided
the least positive care.37
At the same time, various studiesincluding the
NICHD studyhave demonstrated that if the centerbased care is of high quality, it can benefit low-income
preschoolers, especially in terms of cognitive development and engagement in learning-related tasks.38 The
high-quality settings that the study referred to included well-known intervention programs such as Head
Start, the High/Scope Perry Preschool Project, and
other early childhood education centersnot the
types of child care settings generally supported by subsidy programs for welfare families moving into work,
or for working poor families more broadly. And the
effects were modest when compared to the stronger
influence of the home environment.39 Nevertheless,
findings from the NEWWS evaluation reveal that
among children ages three to five growing up in welfare families, those who participated in center-based
care scored better on school readiness assessments
than those in home-based settings.40 In addition, a
new research paper from the GUP Project shows that

among families receiving welfare, children who spend
more time in centers display higher rates of cognitive
and language growth than children in home-based
settings, after taking into account a variety of maternal
and home attributes.41
Access to High-Quality Programs

Poor children stand to benefit most from high-quality
child care. But according to studies from before the
1996 expansion of child care funding, poor children
are less likely to be enrolled in high-quality centers
than are children from wealthier households. Such
were the findings from the Cost, Quality, and Outcomes Study, which observed centers in four states in
1995, and two earlier studies that examined centerbased care in particular cities.42 However, one of these
studies also discovered that, based on structural characteristics, the quality of centers attended by middleclass children was worse, on average, than the quality
of centers attended by poor children.43 The presence of
many moderate or high quality centers in low-income
neighborhoods no doubt reflects the 35 years of targeted federal and state child care spending on centers
in poor communities.
Other studies have confirmed that the quality of centerbased care is not uniformly low across poor communities, but that quality levels are associated with the
richness of state financing and the intensity of quality
regulation. Drawing on a 1990 national survey of child
care organizations, a Harvard research team examined
quality levels for centers in 36 states and found that
some quality indicators were relatively high among subsidized centers in low-income communities compared
to centers in middle-income communities supported
through parental fees.44 More heavily subsidized centers
and those subject to more intense regulation tended to
pay higher staff salaries and more frequently offered a
structured set of learning activities, two factors associated with positive child development.
In contrast, home-based arrangements in low-income
neighborhoods were found to be less well equipped
and less stimulating for children relative to middle-class
settings, a finding confirmed by the NICHD study
group.45 The GUP study, which focused on mothers
child care selections after entering new welfare pro-
107
EyeWire Collection
grams, also found that the quality of typical homebased care was generally low. Whether the home-based
setting was a licensed family child care home or with
kith and kin, often it did not have materials such as art
supplies and play items. Home-based providers did talk
more with individual children and displayed similar levels of warmth and positive social interaction compared
with teachers in center classrooms. Although such positive social interactions, when they occurred, helped to
explain gains in development among children in centers, they did not boost childrens development in
home-based settings.
Efforts to Improve Quality

Head Start plays an important role as a national laboratory for improving the quality of early childhood
education programs.46 The Head Start program is dedicated to addressing weaknesses in quality, meeting
new quality standards, and ensuring adequate numbers, training, and compensation of staff. By contrast,
the primary focus of child care supported by the Child
Care and Development Fund is to enable parents to
108
work.47 Even so, the 1996 law did require states to

spend at least 4% of their total CCDF expenditures
each year on activities to improve the quality and availability of child care. In addition, Congress specifically
earmarked more than $240 million in discretionary
funds in 1998 and 1999 for quality-building efforts.
To date, many states have focused CCDF quality
funds on efforts to support child care resource and
referral agencies that help parents locate care; to provide technical assistance and training to caregivers; and
to help providers meet child care standards set by state
licensing agencies.48 States report, however, that more
funding is needed to provide higher wages for caregivers to reduce turnover and promote the stability of
careimportant to parents long-term employability
and to childrens developmentand to provide adequate capacity for infant care and care during nonstandard work hours. In a recent study of child care
for low-income families, the amount of CCDF dollars
spent on quality averaged just $11.42 per child of
employed parents across the 16 states reporting.49 (See
Volume 12, Number 1
Appendix 1 at the end of this article for three examples of state initiatives to improve access to high-quality child care.)
In sum, the most important element of quality, regardless of setting, is the relationship between the child and
the caregiver. Moreover, children growing up in poverty or facing other family-based risks appear to benefit
most from high-quality child care. Low-income children ages three to five who are placed in high-quality
centers show the strongest gains in cognitive development and early learning, but access to quality programs
in poor communities is uneven. Increased investment
for quality initiatives under CCDF and state funding
streams could lead to significant improvements in childrens learning and development.
Crafting More Effective Policies to

Advance Child Care Options
Welfare reform has sparked stronger political support for
child care and early education, primarily to enable mothers to work.50 For the more than two million parents
currently on the welfare rolls to find and hold down
jobs, new child care providers must be found or current
caregivers must become available for more hours each
week. Over half the children in welfare families were
under age six, and another third were in elementary
school in 1999. To meet the rising demand for care,
federal and state governments have attempted to
expand access to various child care options in lowincome communities by increasing the availability of
vouchers or by making direct institutional efforts to
strengthen center supply. Although significant progress
has been made in expanding subsidies, take-up rates
remain low and the supply of quality options uneven.
Since enactment of the child care block grant in 1990,
federal policymakers have banked heavily on a demandside strategy, based on the idea that use of child care
vouchers will effectively raise low-income families purchasing power and spur the market to strengthen the
child care infrastructure in poor neighborhoods. As a
result of increased funding and expanded eligibility
rules, the number of children receiving child care subsidies under the CCDF and predecessor programs has
grown by about 28%, from approximately 1.4 million
in 1995 to nearly 1.8 million by 1999.51 Still, subsidy
utilization rates remain under one-quarter of all eligible parents, and are highly variable across states and
local areas.52 For example, although subsidy utilization
rates are under one-quarter in many large urban counties such as Los Angeles, other counties have moved
aggressively to raise rates: in San Francisco County,
subsidy use now exceeds two-thirds of all eligible families.53 (See Appendix 2 at the end of this article for a
listing of CCDF utilization rates, by state, in 1999.)
In devising strategies to bolster subsidy take-up rates, it
is important to remember that the child care market is
affected by both demand-side and supply-side factors.
Parents respond to policy rules and incentives in
expressing their demand for particular child care
providers. At the same time, parents live or work in
neighborhoods with variable populations of organizations and individuals who provide care. In contrast to
the demand-side strategy of bolstering parents purchasing power, there is also the older, alternative strategy involving direct public financing of new or
expanded child care centers and preschoolsa supplyside approach. Supply-side financing was how federal
policymakers originally supported child care programs
during World War II. Other examples of this institution-building approach include Head Start and statefunded preschools.
Both demand- and supply-side strategies, if effectively
implemented, can help to expand the range of child
care options in low-income communities and improve
the quality of care. Key factors to consider in improving these strategies include states eligibility criteria,
copayment policies, reimbursement rates, links to center-based care, and local neighborhood contexts, as
discussed below.
Income Eligibility and Copayments

The 1996 law increased states authority to establish eligibility criteria for child care subsidies, and raised the
allowable family income limit to qualify for a subsidy
from 75% to 85% of the state median income. As a
result, states income limits vary widely. By 1999, nine
states had raised eligibility to the new federal maximum
of 85% of the state median income. On the other hand,
Missouri and Wyoming decided that families with
incomes up to only 42% of the state median should be
eligible. Urban Institute researchers found that states
109
raised their monthly income eligibility standard by

$130, on average, after passage of the 1996 reforms.54
At the same time, eligibility does not guarantee utilization. States require most subsidized families to pay a
portion of the monthly costs of child care, or copayment, ranging from $10 to $100 or more. From a
states perspective, copayments can stretch available
child care funding so that more families can receive
subsidies. But from a familys perspective, the copayment may discourage both subsidy use and employment.55 In addition, the administrative process for
getting and retaining subsidies, involving in-person visits and extensive paperwork, can be discouraging for
working poor families who may risk losing their jobs if
they take time off to meet these bureaucratic
demands.56
Reimbursement Rates
Under the federal welfare reform law, states may now
reimburse child care providers (organizations and individuals) above the 75th percentile of local market rates,
previously the cap for welfare-related child care subsidies. About 30 states are continuing to use the 75th
percentile to set their rates, whereas others are using
their discretion to either raise or lower rates. For example, most California counties reimburse providers at
about the 90th percentile of local market rates. In contrast, Massachusetts now sets its rate at the 55th percentile. Reimbursements are constrained both by
setting rates at lower percentiles and by basing payments on old market rate surveys.55
Lower reimbursement rates allow states to provide subsidies to more families, but can make it difficult for families to find care, as fewer providers can afford to accept
the lower rates. Moreover, if a provider accepts the
lower rate, the quality of care offered may be undercut,
as providers rely on lower paid, less well-educated staff,
or skimp on learning-related supplies. Lower reimbursement rates also discourage both individuals and
organizations from entering the provider market. As an
incentive to improve both quality and access to care, an
increasing number of states are experimenting with
tiered reimbursements rates, paying higher rates to centers that are accredited, or to providers who address
special needs (such as infant or odd-hour care) or attend
training or seek certification.57
110
Links between Subsidy Use and Center-Based Care

A high correlation between use of subsidies and enrollment in center-based care has persisted since long
before the 1996 reforms. The inverse also is true: Families who rely on informal arrangements have been far
less likely to draw financial aid for this care. Because
center care often is more costly, it is understandable
that families wishing to use centers would be most likely to seek out a subsidy, but institutional factors may be
contributing to this pattern. The high use of subsidies
for center care in some states and local areas is rooted
in longstanding contracting policies that secure a set
number of center-based slots for children. The subsidycenter linkage also may be partly due to the way information about subsidies is communicated and how
center slots continue to be allocated.
Following the 1996 reforms, federal regulations
required that parents eligible for assistance under the
CCDF be given a choice of receiving a voucher or
enrolling their child in a state-funded facility. Nationwide, use of vouchers is certainly the most widely used
option. In 1999, 83% of children subsidized by the
CCDF were provided vouchers. Only 11% were using
a state-funded center or family child care home, and
the remaining 6% received a direct cash subsidy.51
However, use of state-funded facilities is much higher
in some states. For example, among the 23 states using
CCDF grants and contracts to fund facilities, the percentage of children using these facilities ranged from
only 1% in states such as Colorado, Indiana, New Mexico, and Vermont to a high of 73% in Florida.
By using subsidy dollars for grants and contracts with
selected centers and securing a stable number of slots,
welfare agencies can support the basic infrastructure at
these sites, exert greater influence to promote quality
caregiving, and encourage these centers to expand. But
tying substantial portions of subsidy funds to centers
may deter the use of subsidies by families who prefer
different types of care, and may deprive other providers
of monetary incentives necessary to remain in the field.
Even with respect to vouchers, in many welfare offices
throughout the country, it has been a tacit belief
among clients and caseworkers alike that child care aid
goes only for center-based programs. Researchers have
found that when a caseworker asks, Do you need day
Volume 12, Number 1
care?, many welfare mothers take this to mean, Do

you want to place your child in a center or preschool?
The conversation sometimes ends there, without the
caseworker explaining that a voucher could reimburse
kith or kin members for child care services.58 Since
reform, however, use of vouchers for kith and kin
providers has been growing.
Local Neighborhood Conditions

Before the mid-1990s, child care researchers rarely
focused on neighborhood contexts, particularly the
many small-scale child care organizations created over
the past 40 years. Now, as government agencies escalate efforts to help parents move from welfare to work,
state and federal officials are discovering a territory
densely populated by privately funded centers and nonprofit programs run by community-based organizations and local schools.
Recent data from the Census Bureau reveals the steady
growth in the number of formal centers and family child
care homes since 1982, as depicted in Figure 5. However, nationwide data from a recent study by the Childrens Foundation suggests that licensed centers
capacity grew only between 2% and 3% per year during

the late 1990s, not enough to keep pace with child population growth in major urban areas.59 Moreover, neither study focused on organizational growth in
low-income neighborhoods.
To begin taking stock of child care provider markets at
the local levelthe contexts in which welfare and
working poor parents must make decisionslocal child
care agencies in many states now conduct a census of
centers and family child care homes, tracking how
many are in operation and how many children they are
licensed to serve. A few states collect data on actual
child enrollments. These organization-level data allow
researchers to identify different levels of access to child
care options across diverse zip codes or census tracts. A
recent analysis of such data in California, for example,
revealed some progress between 1996 and 2000:
Capacity growth was higher overall in zip codes that
had a relatively low supply at the beginning of the fouryear period, indicating that these communities are
slowly catching up with high-supply communities.
Although center capacity was relatively high in the
poorest zip codes, it declined in working poor and
Figure 5
Growth in Formal Care Providers 19821997
KEY:
For profit
Nonprofit
aFormal
providers include both centers and

family child care homes.
Source: Based on data from the Census of
Services, as cited in Blau, D. The child care
problem: An economic analysis. New York:
Russell Sage Foundation, 2001.
Number of Formal Providersa Nationwide
70,000
60,000
50,000
40,000
30000
20,000
10,000
0
1982
1987
1992
1997
Year
111
Stronger efforts to expand high-quality child care options will

be needed to achieve the tandem goals of child care: enabling
mothers to work and enhancing the development of children.
lower middle-class communities and areas with higher
concentrations of Latino families, before rising sharply
again in affluent zip codes.60
In California, licensing data from the state welfare
agency show that centers enrollment capacity grew by
2.5% annually in the three years following passage of
the states welfare reform bill in 1997, nearly one full
percentage point behind the 3.4% annual growth rate
in the states child population.21 Similar slow rates of
center growth have been found in Illinois and Maryland.61 In addition, the number of new centers granted licenses grew at about two-thirds of the
center-growth rate, indicating that much of the expansion was through new slots within existing centers
rather than entry of new organizations. Adding slots
to existing centers does little to expand capacity for
many working-class communities and new immigrant
communities because the number of existing centers is
acutely low in those communities.
Moreover, the quality of center-based programs also
depends on state and local conditions. In some lowincome communities, public investment targeted on
carefully regulated centers and preschools has effectively
sustained programs of reasonably strong quality, at least
with respect to structural factors. In other neighborhoods
where infrastructure is weak, financial incentives have
been insufficient to sustain higher quality centers. The
mix of centers and preschools run by school districts,
nongovernment organizations, and for-profit firms also
affects average quality levels, especially when centers
operate under weak state regulation. The GUP Project,
for instance, has detailed the ample supply of low-quality
centers in Florida. Beyond efforts to increase subsidy
take-up rates, states may need to address the political and
economic forces that surround neighborhood populations of child care organizations to simultaneously bolster
supply and improve the quality of care.
In sum, federal and state governments have significantly increased spending on child care and preschools for
low-income families since 1996. Use of vouchers for
kith and kin providers has grown rapidly, and the sup-
112
ply of centers and family child care homes in major

urban states is struggling to keep pace with child population growth. The constraints on center supply may
be limiting parental choice and pushing families toward
kith and kin caregivers. The supply of quality care
options is uneven, especially in poor and middleincome communities, and the number of providers
entering the child care market may be tapering off due
to flagging subsidy take-up rates and the low reimbursement rates set by some states. Stronger efforts to
expand high-quality child care options for low-income
families will be needed to achieve the tandem goals of
child care: enabling mothers to work and enhancing
the development of children.
Conclusions
Although policymakers and private benefactors have
long argued that public agencies can effectively
strengthen the child care infrastructure and regulate
quality, progress has been slow. Meanwhile, affluent
families have built and enriched their own child care
infrastructure, privately financing expansion and quality, often through hefty fees. Ensuring that children in
welfare and working poor families have equal access to
high-quality care is a crucial challenge facing society
and all levels of government.
Important empirical lessons are emerging about the
extent to which welfare reforms have or have not
widened child care options for low-income parents
making new decisions about who cares for their children. But much remains to be learned in two crucial
areas. First, little is known about the relative benefits of
maternal versus the different types of nonmaternal care
for low-income children of different ages. It is unclear
whether the increasing use of nonmaternal care by welfare families helps or hinders early development
because information is just now beginning to emerge
about the quality of childrens home settings versus the
quality of care in settings outside the home. More
focused analyses should explore the comparative quality of different types of care and the underlying reasons
Volume 12, Number 1
parents select particular arrangements for their children. Second, too little is known about the effects on
both families and providers of alternative policies
regarding eligibility for child care aid, out-of-pocket
costs, reimbursement rates, the links between subsidies
and centers, and the effects of neighborhood supply on
subsidy take-up rates.
Successful policies need to be identified, both to support
stronger gains in mothers employability and to promote
childrens development. To begin, the following policy
adjustments could help ensure that welfare and working
poor parents are truly able to choose from a range of
quality child care options in their neighborhoods:
Welfare and working poor parents need clear, comprehensive information about their child care options
to gain purchasing power through the use of child
care vouchers and bolster growth of quality choices.
CCDF funding should be increased and states should
expand the capacity of center-based programs and
licensed family child care homes so that welfare and
working poor families have a full array of stable,
affordable options.
In support of federal, state, and local efforts to bolster
subsidy use and ensure that parents and childrens
needs are being met, better information should be
gathered on the types of child care low-income parents
prefer, the stability and quality of the care they select,
and the ways parents are paying for the arrangements.
A huge amount of political capital has been invested in
the proposition that single mothers should work to
build a better future for themselves and their children.

But maternal employment alone cannot benefit children unless it leads to improvements in childrens daily
environments. It is not enough for welfare reform simply to cause no harm. Welfare-to-work programs must
focus on policies that help promote childrens development by widening access to high-quality child care and
after-school options. Children need opportunities for
brighter futures if welfare programs are to achieve the
ultimate goal of breaking the intergenerational cycle of
poverty. Delivering on the promise of affordable, highquality child care would be an important step toward
realizing this goal.
The authors wish to warmly thank Doug Besharov,

Hilva Chan, Mark Greenberg, Lee Kreader, Joan
Lombardi, Ivelisse Martinez-Beck, Kristin Moore,
Rachel Schumacher, Margie Shields, and Karen Tvedt
for their thoughtful comments on earlier drafts; and
Naomi Karp for commissioning three independent
reviewers as well. Much of the research reviewed in this
article stems from the Child Care Research Partnership, supported by the U.S. Department of Health and
Human Services. Work on this article was supported
by the Growing Up in Poverty Project, with funding
for dissemination and policy engagement activities
provided by the MacArthur, Mailman, and Graustein
foundations, and the Walter and Elise Haas Fund, in
cooperation with the National Conference of State
Legislatures.
ENDNOTES
1. Personal Responsibility and Work Opportunity Reconciliation Act
of 1996. Public Law 104-193, section 103, 110 Stat. 2113,
August 22, 1996.
2. Michel, S. Childrens interests, mothers rights: The shaping of
Americas child care policy. New Haven, CT: Yale University Press,
1999.
2000 green book: Background material and data on programs within the jurisdiction of the Committee on Ways and Means. Washing-
ton, DC: U.S. Government Printing Office, 2001, p. 597.

4. See the fact sheets available on the U.S. Department of Health
and Human Services Head Start Bureau Web site at
http://www2.acf.dhhs.gov/programs/hsb/.
5. Reductions in the Dependent Care Tax Credit stemmed largely
from changes enacted in the Family Support Act of 1988. See
note 3, U.S. House of Representatives, Committee on Ways and
Means, pp. 81315.
113
and Means, pp. 62527, and data charts available on the U.S.
Department of Health and Human Services Child Care Bureau
Web site at http://www.acf.dhhs.gov/programs/ccb/research.
7. Blank, H., and Poersch, N. State developments in child care and
early education, 1999. Washington, DC: Childrens Defense
Fund, 2000.
8. Cauthen, N.K, Knitzer, J., and Ripple, C.H. Map and track: State
initiatives for young children and families, 2000 edition. Highlights
from findings. New York: Columbia University, National Center
for Children in Poverty, 2001, p. 5. Available online at
http://cpmcnet.columbia.edu/dept/nccp.
9. Smith, K. Whos minding the kids? Child care arrangements: Fall
1995. Current Population Reports, P70-70. Washington, DC:
U.S. Census Bureau, 2000.
10. Hofferth, S., Shauman, K., Henke, R., and West, J. Characteristics of childrens early care and education programs: Data from the
1995 National Household Education Survey. National Center for
Education Statistics, no. 98-128. Washington, DC: NCES, 1998.
11. Brimhall, D., Reaney, L., and West, J. Participation of kindergarteners through third-graders in before- and after-school care.
National Center for Education Statistics, no. 1999013. Washington, DC: NCES, 1999. See also note 9, Smith.
12. Capizzano, J., Adams, G., and Sonenstein, F. Child care arrangements for children under five: Variation across states. Washington,
DC: Urban Institute, 2000.
and Means, p. 376, table 7-4, and p. 431, table 7-25.
14. Under the Family Support Act of 1988, the Jobs Opportunities
and Basic Skills Training (JOBS) program exempted families from
work requirements if they had children under age three, or under
age one by state option. But only a small share of welfare clients
with young children actually were required to work until state
waiver programs began in 1993. See Friedlander, D., and Burtless, G. Five years after: The long-term effects of welfare-to-work
programs. New York: Russell Sage Foundation, 1995; and
ONeill, D., and ONeill, J. Lessons from welfare reform. Kalamazoo, MI: Upjohn Institute for Employment Research, 1997.
15. For example, in Connecticut, some 14% more families participating in the states welfare-to-work program were using a child care
provider for at least 10 hours per week in 1998, about 18 months
after entering the new program, compared with those not participating (that is, families in the control group). This higher use of
care was linked to the higher rate of employment among participants with at least one preschool-age child. See Growing Up in
Poverty Project. Remember the children: Mothers balance work
and child care under welfare reform. Berkeley: University of California, Policy Analysis for California Education, 2000.
16. For details of this estimate, see note 15, Growing Up in Poverty
Project, technical supplement.
17. See Raikes, H. Investigating child care subsidy: What are we buying? Social Policy Report (1998) 7(2).
18. Zaslow, M.J., Oldham, E., Moore, K.A., and Magenheim, E.
Welfare families use of early childhood care and education programs, and the implications for their childrens development.
Early Childhood Research Quarterly (1998) 13(4):53563.
19. See note 15, Growing Up in Poverty Project. Caution is warranted in making comparisons to Connecticut, because wave 1 data
for this state were collected 18 months after participating families
were randomly assigned to an experimental or control group,
114
whereas in California and Florida, the child care data were collected 6 months after entry.
20. Knox, V., Miller, C., and Gennetian, L. Reforming welfare and
rewarding work: Summary of the final report on the Minnesota
Family Investment Program. New York: Manpower Demonstration Research Corp., 2000. See also Berlin, G. Encouraging work,
reducing poverty: The impact of work incentive programs. New
York: Manpower Demonstration Research Corp., 2000, p. 17.
21. Hirshberg, D., and Fuller, B. Child care for low-income California families. Berkeley: University of California, Policy Analysis for
California Education. In press.
22. See note 9, Smith, pp. 10 and 17 (table 9).
23. Meyers, M., and Heintze, T. The performance of the child-care
subsidy system. Social Service Review (March 1999) 73(1):3764.
24. Fuller, B., Kagan, S.L., Suzuki, S., and Chang, Y. Child care
quality and affordability: Uneven opportunity for poor families in
California. Working paper 01-2. Berkeley: University of California
and New Haven, CT: Yale University: Growing Up in Poverty
Project, 2001.
25. Holloway, S., Fuller, B., Rambaud, M., and Eggers-Pirola, C.
Through my own eyes: Single mothers and the cultures of poverty.
Cambridge, MA: Harvard University Press, 1998. See also
Brown-Lyons, M., Robertson, A., and Layzer, J. Kith and kin
informal care: Highlights from recent research. New York: National
Center for Children in Poverty, 2001.
26. Liang, X., Fuller, B., and Singer, J. Ethnic differences in child
care selection: The influence of family structure, parental practices, and home language. Early Childhood Research Quarterly
(2000) 15:35784.
27. Siegel, G., and Loman, L. Child care and AFDC recipients in Illinois: Patterns, problems, and needs. St. Louis: Institute of Applied
Research, 1991.
28. Fuller, B., Kagan, S.L., Gascue, L., et al. Explaining family
demand for early education: Household factors and neighborhood organizations. Working paper. Berkeley, CA: University of
California and New Haven, CT: Yale University, 2001.
29. Giannarelli, L., and Barsimantov, J. Child care expenses of Americas families. Washington, DC: Urban Institute, 2000.
30. See note 9, Smith, pp. 2627.
31. Coonerty, C., and Levy, T. Waiting for child care: How do parents
adjust to scarce options in Santa Clara County? Berkeley: University of California, Policy Analysis for California Education, 1998.
32. Shlick, D., Daly, M., and Bradford, L. Faces on the waiting list:
Waiting for child care assistance in Ramsey County. Minneapolis:
University of Minnesota and Ramsey County Human Services,
1999.
33. Fuller, B., Holloway, S., and Liang, X. Family selection of child
care centers: The influence of household support, ethnicity, and
parental practices. Child Development (1996) 67:332037. See
also note 18, Zaslow, et al.
34. Vandell, D.L., and Wolfe, B. Child care quality: Does it matter
and does it need to be improved? Washington, DC: U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, May 2000, p. 1.
National Academy Press, 2000, pp. 30720.
Volume 12, Number 1
36. Galinsky, E., Howes, C., Kontos, S., and Shinn, M. The study of
children in family child care and relative care. New York: Families
and Work Institute, 1994.
48. U.S. General Accounting Office. Child care: States increased

spending on low-income families. GAO-01-293. Washington, DC:
GAO, February 2001, pp. 8, 2223.
37. National Institute of Child Health and Human Development

Early Child Care Research Network. Characteristics and quality
of child care for toddlers and preschoolers. Applied Developmental Science (2000) 4(3):11635. For further details on the
NICHD Study of Early Child Care and Youth Development, see
the Web site at http://public.rti.org/secc/summary.cfm.
care for low-income families: State and community substudy interim report. Washington, DC: U.S. Department of Health and
Human Services, Administration for Children and Families,
November 2000, p. 32.
38. Recent reviews include Barnett, W.S. Long-term effects of early

childhood programs on cognitive and school outcomes. The
Future of Children (Winter 1995) 5(3):2550. Burchinal, M.
Child care experiences and developmental outcomes. Annals of
the American Academy of Political and Social Science (1999)
563:7397. National Institute of Child Health and Human
Development Early Child Care Research Network. Relations
between family predictors and child outcomes: Are they weaker
for children in child care? Developmental Psychology (1998)
34:111928. Campbell, F., and Ramey, C. Effects of early intervention on intellectual and academic achievement: A follow-up
study of children from low-income families. Child Development
(1994) 65:68498.
39. National Institute of Child Health and Human Development
Early Child Care Research Network. The relation of child care to
cognitive and language development. Child Development (2000)
71(4):95878.
40. Zaslow, M., McGroder, S., Moore, K., and LeMenestrel, S.
Behavior problems and cognitive school readiness among children in families with a history of welfare receipt. Paper presented
at the Biennial Meeting of the Society for Research in Child
Development, Albuquerque, NM. April 1518, 1999. See also
note 18, Zaslow, et al.
41. Loeb, S., Fuller, B., Strath, A., and Carrol, B. Active states construct the early education sector: Economic and institutional
determinants of child care distribution. Working paper. Stanford,
CA: Stanford University, 2001.
42. Helburn, S., ed. Cost, quality, and child outcomes in child care
centers: Technical report. Denver: University of Colorado,
Department of Economics and Social Policy, 1995.
43. Phillips, D., Voran, M., Kisker, E., et al. Child care for children
in poverty: Opportunity or inequity? Child Development (1994)
65(2):47292.
44. Fuller, B., Raudenbush, S., Wei, L., and Holloway, S. Can government raise child care quality? The influence of family demand,
poverty, and policy. Educational Evaluation and Policy Analysis
(1993) 15:25578.
45. The NICHD study, however, included only a small subsample of
children in lower-income families. See note 36, Galinsky, et al.
See also NICHD Early Child Care Research Network. Poverty
and patterns of child care. In Consequences of growing up poor. G.
Duncan and J. Brooks-Gunn, eds. New York: Russell Sage Foundation, 1997.
46. U.S. Department of Health and Human Services. Administration
for Children and Families: Head Start. Fact sheet. Washington,
DC: DHHS, January 2001 update. Available online at
http://www.acf.dhhs.gov/programs/opa/facts/headst.htm.
47. See the U.S. Department of Health and Human Services Administration for Children and Families Child Care Bureau Web site at
http://www.acf.dhhs.gov/programs/ccb/geninfo/ccdfdesc.htm.
50. Numerous studies indicate that child care support holds implications for parents employability. See Berger, M., and Black, D.
Child care subsidies, quality of care and the labor supply of lowincome single mothers. Review of Economics and Statistics (1992)
74:63542. See also Gordon, R., and Chase-Lansdale, P.L.
Womens participation in market work and the availability of
child care in the United States. Working paper 9905. Chicago:
University of Chicago, Sloan Center, 1999; and Lemke, R.,
Witte, A., Queralt, M., and Witt, R. Child care and the welfare
to work transition. Working paper 7583. Cambridge, MA:
National Bureau of Economic Research, 2000.
51. See the data charts available on the U.S. Department of Health
and Human Services Child Care Bureau Web site at
http://www.acf.dhhs.gov/programs/ccb.
53. Carroll, J. How to pay for child care? Local innovations help working families. Policy brief 01-1. Berkeley: University of California,
Policy Analysis for California Education, Spring 2001.
54. Loprest, P., Schmidt, S., and Witte, A. Welfare reform under
PRWORA: Aid to children with working families? Tax Policy and
the Economy (2001) 14:157203.
55. U.S. General Accounting Office. Welfare reform: States efforts to
expand child care programs. GAO/HEHS-98-27. Washington,
DC: GAO, January 1998, pp. 1213. See also Besharov, D., and
Samari, N. Child care after welfare reform. In The New World of
Welfare. R. Blank and R. Haskins, eds. Washington, DC: Brookings Institution Press, 2001, pp. 46176.
56. Adams, G. Preliminary findings on the adequacy of child care
funding from the Assessing the New Federalism (ANF) child care
case studies. Paper presented at a Brookings Forum, Child care
funding: How much is needed and is there enough? Washington, DC. June 13, 2001.
57. See note 49, Collins, et al., pp. 8186, 94.
58. Mensing, J.F., French, D., Fuller, B., and Kagan, S.L. Child care
selection under welfare reform: How mothers balance work
requirements and parenting. Early Education and Development
(2000) 11(5):57395.
59. See the Childrens Foundation annual reports, available online at
http://www.childrensfoundation.org.
60. Such zip code-level data also allow researchers to statistically
model neighborhood-level factors that help to explain differing
growth rates in centers and family child care homes. The California data are collected by and detailed in the California Child
Care Resource and Referral Network. The California Child Care
Portfolio, 1999. San Francisco: CCCRRN, 1999.
61. Kreader, J.L., Piecyk, J.B., and Collins, A. Scant increases after
welfare reform: Regulated child care supply in Illinois and Maryland, 19961998. New York: Columbia University, National Center for Children in Poverty, 2000.
115
APPENDICES
Appendix 1
Innovative Uses of TANF Funds to Improve
Access to High-Quality Child Care
These program profiles were prepared by Kate Boyer, Ph.D., senior
researcher, and Catherine Lawrence, C.S.W., research associate, of
the Rockefeller Institute of Government as part of the Institutes
project, Beyond Symbolic Politics.
Program name:
Kansas Early Head Start
State:
Kansas
Coverage:
13 local programs throughout the state
Program goals:
Promote healthy prenatal outcomes for pregnant women, enhance the development of
very young children, and promote healthy
family functioning
Service population: Families with children ages 0 to 4 whose

household incomes are at or below 100% of
the federal poverty guidelines at enrollment;
one-tenth of enrollment slots are reserved for
children with disabilities
Funding sources:
TANF funds transferred into the states Child

Care and Development Fund, and funds from
the federal Head Start program
Description: Launched in 1998, this Kansas program constitutes the

nations first effort to provide enriched child care environments and
other services to families by merging TANF-funded child care with
the federal Head Start program. By partnering with existing child
care providers in the community, the program provides full-day, fullyear child care while parents are at school, job training, or work. The
program also provides comprehensive services including nutrition,
health and social services, parent and community involvement, and
self-sufficiency training for parents. In 2001, Kansas Early Head
Start served 825 children in 32 counties statewide.
Results: Each of the 13 local programs must adhere to performance
standards as laid out by the National Head Start program, monitored
every three years through on-site visits. Since 1998, 11 sites have
received site visits; all have met the federal performance standards.
Selected Early Head Start programs in Kansas were also included in
a national evaluation conducted by Mathematica Policy Research.
The study found that children enrolled in Early Head Start enrichment programs have significantly better cognitive, linguistic, and
social skills than children who do not participate in the program.
For further information: See the Kansas Department of Social and
Rehabilitation Services Web site at http://www.srskansas.org/
kidsnet/kehskhs.htm; and the Mathematica Web site at
http://www.mathematica-mpr.com/3rdlevel/ehstoc.htm.
116
Volume 12, Number 1
Program name:
Los Angeles County After-School Enrichment

Program
State:
California
Coverage:
153 public schools in Los Angeles County
Program goals:
Provide after-school care that enhances

academic achievement
Program name:
Building Capacity Project
State:
Washington
Coverage:
Statewide
Program goals:
Increase the supply of child care that meets

special needs
Service population: Children receiving TANF
Service population: Families earning below 225% of the federal

poverty guidelines
Funding sources:
Funding sources:
TANF
Description: For the past three years, Los Angeles County has
offered TANF funds to expand the capacity of local schools to meet
the needs of low-income families and children by operating an
after-school program in those schools with the highest percent of
TANF-eligible children. The program provides care for school-aged
children at times when parents often have difficulty finding care,
such as weekends, vacations, and holidays. In addition, the program works to improve students academic achievement by mandating a high level of teacher involvement. Some schools use the
funds to improve teacherstudent ratios, a key indicator of quality
programming. Others use the funds to resume lost activities such
as drama, art, and music. In each case, academic and enrichment
activities are required. As of April 2001, nearly 6,000 TANF children
had been enrolled in the program. The county is also working with
other cities that want to replicate the program throughout the
country.
Results: An evaluation of 30 sites is under way, and initial responses to the program are positive. Parents say the individualized attention has improved their childrens reading and writing, teachers say
they have seen some children blossom into real leaders, and principals feel the program has built more solid bridges between the
school and parents.
TANF
Description: The Building Capacity Project seeks to expand access

to certain hard-to-find forms of child care, including infant care,
middle-school child care, before- and after-school care, evening and
weekend care, and care for children with disabilities. To accomplish
this goal, the program provides training so that first-time care
providers may gain licensing, and existing child care centers may
expand their capacity. Program administration is subcontracted to
community groups. These groups must justify their choice of neighborhoods for increasing the supply of care and then must submit
monthly progress reports toward achieving their goals. In addition
to expanding child care options for low-income families, the program also seeks to provide an economic boost to small business
owners in economically-depressed areas by aiding the expansion of
child care centers.
Results: This program began in October 2000, thus implementation
is still in its early stages. Site-monitoring was planned for 2001, and
an on-site assessment is scheduled for 2002.
For further information: See the Washington State Department of
Social & Health Services Web site at http://www.wa.gov/dshs/
occp/ccdfinal.doc.
For further information: See the Los Angeles County child care
directory Web site at http://childcare.co.la.ca.us/afterschool
_enrichment_prog.htm.
117
Appendix 2
Number of children
CCDF Utilization Rates, by State, 1999
State
Serveda
Eligibleb
These rates include only those children funded under the

Child Care and Development Fund. They do not include children
participating in Head Start or other state-funded preschools or
Alabama
24,500
233,300
11%
Alaska
6,260
46,700
13%
Arizona
36,590
283,800
13%
Arkansas
11,250
180,600
6%
California
226,750
1,732,500
13%
Colorado
23,790
226,300
11%
Connecticut
9,790
187,700
5%
Delaware
5,920
50,700
12%
District of Columbia
1,040
31,500
3%
Florida
58,630
705,300
8%
Georgia
38,170
485,200
8%
Hawaii
7,110
81,200
9%
Idaho
7,560
68,200
11%
Illinois
92,030
676,000
14%
Indiana
20,230
299,800
7%
Iowa
15,720
199,200
8%
Kansas
11,570
172,800
7%
Kentucky
26,220
170,200
15%
Louisiana
38,980
219,700
18%
8,890
60,900
15%
Maryland
22,070
259,900
8%
Massachusetts
40,200
301,700
13%
Michigan
101,890
545,100
19%
Minnesota
17,200
297,400
6%
Mississippi
17,870
185,500
10%
Missouri
58,390
305,600
19%
child care programs.
Maine
118
Percent
Volume 12, Number 1
Number of children
Serveda
Eligibleb
Percent
Montana
6,430
60,800
11%
Nebraska
12,140
115,000
11%
Nevada
5,900
97,000
6%
New Hampshire
6,790
71,600
9%
New Jersey
34,000
350,500
10%
New Mexico
16,610
126,900
13%
164,200
880,900
19%
North Carolina
67,100
411,400
16%
North Dakota
4,450
37,700
12%
Ohio
58,440
577,300
10%
Oklahoma
30,820
191,100
16%
Oregon
20,490
188,500
11%
Pennsylvania
82,750
533,900
15%
Rhode Island
6,390
42,500
15%
South Carolina
17,840
231,000
8%
South Dakota
3,680
46,200
8%
Tennessee
63,090
346,000
18%
Texas
96,640
1,161,700
8%
Utah
13,260
130,400
10%
Vermont
4,980
33,400
15%
Virginia
27,120
348,100
8%
Washington
46,130
310,500
15%
West Virginia
13,310
52,700
25%
Wisconsin
24,940
365,800
7%
Wyoming
3,330
31,600
11%
U.S. Total
1,760,260
14,749,500
12%
State
New York
a Average
b
monthly number of children served in fiscal year 1999.
Number of children eligible under the maximum limit allowed

under federal law, set at 85% of the state median income.
Source: U.S. Department of Health and Human Services, Administration for Children and Families. Based on numbers from the 1999 Access to Child Care for LowIncome Working Families Report. New statistics show only small percentage of
eligible families receive child care help. Press release. Washington, DC: USDHHS,
December 6, 2000.
119
Family Economic Resources

in the Post-Reform Era
Sheila Rafferty Zedlewski
SUMMARY
Aided by the longest economic expansion in

U.S. history and other policy changes
designed to make work pay, federal welfare
reform legislation has spurred mothers to
leave welfare at an unprecedented rate. The
majority of mothers who left welfare are
working, but most have jobs with low pay
and limited benefits. This article discusses
the relationship between economic
resources and child well-being, and how
family economic resources have changed
under welfare reform. A survey of the
research conducted since reform indicates
the following:
Families economic resources clearly matter to child well-being, but the connections are complex and vary by the age of
the child.
Without the benefit of supports designed
to make work pay, many families working full time at the minimum wage have
resources beneath the poverty line, and the
poverty line itself falls substantially short of

the needs of most working families.
Although poverty overall has declined
under welfare reform, a significant segment of families are worse offin part
because after leaving welfare, many families
do not receive other government supports
designed to help them.
Most states are still struggling to design
more effective systems for delivering supports to help low-income working families
move out of poverty. The author cautions
that the evolving story of welfare reform will
need to be monitored carefully to achieve
long-term positive impacts on family economic resources and child well-being.
Sheila Rafferty Zedlewski, M.P.A., is director of the

Income and Benefits Policy Center at The Urban
Institute, a private nonprofit research institution in
Washington, D.C.
121
Zedlewski
enough to support their families. Most families will need

help paying for child care, transportation, housing, and
health expenses.
ederal welfare reform legislation, passed in

August 1996, ushered in a new era in U.S. history of providing support for vulnerable families with children. The law replaced
entitlements to cash assistance with fixed block grants to
states under a new program titled Temporary Assistance
for Needy Families (TANF). (See the article by Greenberg and colleagues in this journal issue.) Aided by the
longest economic expansion in U.S. history and other
policy changes designed to make work pay, the new legislation hastened an already rapid decline in welfare caseloads and left states with considerable federal funds to
use in new ways to help low-income working families.
Welfare reform is still evolving as families adjust to new

expectations and new support packages, and as states
adjust to their new responsibilities in setting the basic
safety net for low-income families with children. This
article examines the relationship between economic
resources and child well-being, and how family economic resources are changing as a result of reform. A
description of federal and state policies that help to make
work pay is provided, followed by a summary of the
research describing former welfare recipients earnings,
incomes, and government supports. A final section
reviews state initiatives that provide new types of noncash assistance for working poor families to help close
the gap between the apparent benefits offered to these
families and what they actually receive.
It is not yet clear how federal welfare reform ultimately

will affect family and child well-being. The majority of
mothers who left welfare are working, but most have
jobs with low pay and limited benefits. Proponents of
reform argue that mothers who rely on work rather than
welfare provide better role models for their children,
achieve greater self-esteem, and, in time, are likely to
move up the economic ladder. Opponents caution that
children will fare worse unless working mothers can earn
Economic Resources and Child WellBeing: What Is the Connection?

The preamble to the federal welfare reform bill focuses
on the goals of marriage and responsible fatherhood
and motherhood as integral to successful child rearing
and the well-being of children.1 Noticeably absent is
a goal of reducing poverty or increasing the economic
resources available to children. Part of the political
debate surrounding how much and how long government assistance should be available to low-income families with children still focuses on whether the amount
of economic resources available to families really makes
a difference to child well-being. The research on this
topic suggests that families economic resources clearly
matter to child well-being, but the connections are
complex and vary by the age of the child.
Greater Income Related to Better Outcomes
122
It is well documented that children in families with

greater incomes do better across a wide range of indicators. Economically secure children tend to be healthier and do better in school; they are less likely to be
involved in criminal behavior and are more likely to
graduate from high school and to earn higher incomes
as adults. In contrast, poorer children tend to have
fewer opportunities for success.2
Volume 12, Number 1
Economic Resources Post-Reform
Economically secure children tend to be healthier and do better

in school; they are less likely to be involved in criminal behavior
and are more likely to graduate from high school and to earn higher
incomes as adults.
Various studies have attempted to document the
importance of income relative to other factors believed
to affect child well-being. For example, one study comparing outcomes of children in single-parent versus
two-parent families found that differences in income
account for as much as half of the difference between
these two groups of children in school achievement
and early childbearing, and for a substantial portion of
the difference in risk of dropping out of high school.3
Another study followed a cohort of children born in
the late 1960s and found that, among several factors
analyzed, income had the most significant effect on
child well-being and achievement.4 Children in families
with low income were less likely to graduate from high
school and more likely to have children before age 25
when compared with children of the same demographic backgrounds with higher incomes.
Although greater income is clearly associated with better outcomes for children, the extent to which money
matters and the pathways through which a familys
economic resources affect child well-being remain
unclear. Some researchers argue that income matters
because it leads to richer learning environments
through better neighborhoods, better schools, and
extracurricular activitiesand because children who
are exposed to a richer educational environment are
less likely to repeat the barriers that lead to dependence
on federal aid, such as low educational attainment,
drug use, or teen pregnancy. Also, children may be
more highly motivated to succeed when they know
their parents can help pay for college.
Other researchers have questioned whether it is the
lack of resources that poverty entails or something
more integral to the structure or culture of poor families that affects outcomes for poor children. Critics of
the conclusion that income matters argue that it is not
money per se, but the values intrinsic in employment
that are important to combating poverty. Thus,
income derived from work is believed to have a greater
beneficial effect on children than income derived from
welfare.5 By watching their parents struggle to provide

for the family rather than simply taking money from
the government, children learn important lessons
about self-sufficiency and hard work.6 One critic has
asserted that attempting to raise the family income
artificially through welfare is very unlikely to do much
to benefit the child, but it is likely to destroy the very
values that are key to the childs success.7
The evidence suggests that programs designed to
increase family employment and economic resources
can benefit children, but the effects vary by the age of
the child. For example, a summary of findings from
five large-scale demonstration projects concluded that
programs that provided earnings supplements along
with increased parental employment led to higher
school achievement, fewer behavioral problems,
increased positive social behavior, and sometimes
improved health among preschool and school-age children.8 The higher incomes apparently helped purchase
enriching experiences for these children at a time when
it mattered most to their development.9 In contrast,
the summary revealed some negative effects on adolescents behavior10 and identified no effects on the very
young.11 (See the article by Zaslow and colleagues in
this journal issue.) Because policies that affect families
economic resources do not influence childrens behavior and development directly, any effectseither positive or negativemost likely occur indirectly through
changes in parents employment and availability, the
use and quality of child care and early education, and
other family circumstances.
In sum, greater income clearly is linked to better outcomes for children, but the pathways connecting
income to child well-being are unclear, and demonstration projects designed to increase families economic resources have had varying effects on children.
As a first step to promoting more effective programs, it
would be helpful to know the minimally adequate level
of resources sufficient to support positive outcomes.
123
Zedlewski
The data suggest that the traditional poverty line measure falls
substantially short of meeting the needs of a working mother
or couple with children.
Measuring Poverty and Families Needs
Several benchmarks, including the poverty line, alternative measures of expenditure needs, and relative economic well-being measures, have been developed to
assess income adequacy. Although none of these measures directly relate to child well-being, the data suggest
that the traditional poverty line measure falls substantially short of meeting the needs of a working mother
or couple with children.
Since the 1960s, the United States has used a measure
of poverty to indicate the minimum level of cash
income families need to obtain food and other goods
and services. The poverty line was set at three times
the cost of a healthy but minimal budget for food,
based on the observation that the average family spent
one-third of its budget on food in the mid-1950s.12
This poverty measure, updated each year to reflect the
change in the consumer price index, is often used as a
benchmark of income adequacy, child well-being, and
the need for assistance from various government programs. However, a major study by the National
Research Council of the National Academy of Sciences
in 1995 concluded that this measure no longer provides an accurate picture of the differences in the extent
of economic poverty in the United States nor of trends
over time. The measure does not include in-kind economic resources available to families through the
earned income tax credit (EITC) and food stamps, nor
does it exclude taxes and out-of-pocket child care and
health expenses. The panel proposed a new measure
that would more accurately identify family resources
and budget needs, and some research organizations,
including the U.S. Census Bureau, have begun using
this measure in surveys and other studies.
An alternative method proposed for assessing family
needs is based on how much welfare recipients would
need to earn to leave welfare for work.13 To develop
this self-reliant measure, researchers examined lowincome families expenditure patterns and the cost of
their basic needs. The data showed that, in 1991,
households consisting of mothers with two children in
124
which the mother was paying for housing needed

$1,000 per month, on average, to cover basic expenditures while on welfare, and $1,330 per month to maintain the same standard of living while working. The
researchers estimated that a working mother needed to
earn at least $8 per hour at a full-time job to reach this
level of income. Mothers earning less per hour or
working less than full time would need other income
sources (such as a partial welfare check, housing or
child care subsidies, and the EITC) to make up the difference. In 1999, these estimates translated into a wage
of $10 per hour for a full-time worker (or lower wages
or hours supplemented by other in-kind or cash
resources) to reach a monthly income of about $1,630.
Because TANF enforces strict work requirements for
most single mothers, alternative measures of income
adequacy have been proposed for working families. For
example, one proposal recommends a basic needs or
self-sufficiency budget based on estimates of the minimum income level required for a family to afford the
necessities (food, housing, transportation, health care,
child care, and taxes) without government aid.14 Others
have suggested using one-half the median income as a
benchmark to measure the adequacy of a working familys income.15 This approach measures adequacy relative
to the general populations income. But by this measure,
regardless of the countrys economic situation, onefourth of the population will always be considered poor.
These measures of income adequacy roughly bracket
an adequate level of resources for a family. (See Figure
1.) At the high end, the basic needs budget estimates
that the average cost for the basic necessities for a
working family of three is about two times the low-end
federal poverty standard. The one-half median income
approach also places a familys needs significantly above
the poverty line, but below the basic needs budget, as
does the self-reliant measure.
Although the research does not point to a specific level
of economic resources sufficient to support positive
outcomes for children, data from experimental studies
suggest that even fairly modest increases in income can
Volume 12, Number 1
Figure 1
Alternative Measures of Income
Adequacy for a Family of Three
have positive effects. For example, some programs that

increased families economic resources by $1,200 to
$4,000 per year over the level of resources previously
received on welfare had positive effects on children in
the preschool and elementary school years. If families
resources were raised to the levels suggested by the
poverty line alternatives depicted in Figure 1, the additional resources would be much greateran additional $6,200 to $15,000 for a family of threeand could
have dramatic effects on the well-being of children.
New Policies Designed to

Make Work Pay
A major goal of welfare reform is to move families from
welfare to work. Even when working, however, many
low-income families have difficulty earning enough
money to provide for their families needs. As a result,
federal and state governments have developed several
incentives and supports to increase the payoff from
work, including minimum wage standards, tax credits,
and policies that allow welfare recipients to earn more
without losing benefits. Figure 2 summarizes the
changes in policies that complement welfare reforms
goal of increasing work among TANF recipients.
Minimum Wage
Federal and state governments set minimum wage levels that employers must pay workers in jobs covered by
the Fair Labor Standards Act (FLSA), which included
$25,595
$25,000
1999 Dollars
Sources: Poverty threshold from U.S. Census Bureau. Poverty in

the United States, 1999. Series P60210. Washington, DC: U.S.
Government Printing Office, September 2000. Self-reliant measure from Edin, K., and Lein, L. Making ends meet. New York:
Russell Sage Foundation, 1997. One-half median income measure from U.S. Census Bureau. Money income in the United States:
99. Washington, DC: U.S. Government Printing Office, September
1999, p. 1, table 1. Basic needs budget based on the average
cost of the basic necessities for a family living in Baltimore,
Maryland, from Bernstein, J., Broeht, S., and Spade-Aquilar, M.
How much is enough? Basic family budgets for working families.
Washington, DC: Economic Policy Institute, 2000, p. 61.
$28,031
$30,000
$19,600
$20,000
$13,423
$15,000
$10,000
$5,000
$0
Poverty
Line
Self-Reliant
One-Half
Median
Income
Basic
Needs
Budget
72% of wage and salary workers in 1999.16 The federal

minimum wage was last increased from $4.25 per hour
to $4.75 in 1996 and to $5.15 per hour in 1997, the
first real increases in five years. The current level provides a full-time, full-year worker an income equal to
$10,712 a year, about 80% of the poverty line for a family of three in 2000. Eleven states (including the District
of Columbia) mandate higher wage floors, ranging
from $5.25 per hour in Hawaii to $6.75 in Massachusetts.17 A full-time, full-year worker paid the minimum
wage in Massachusetts would earn an income just
around the federal poverty line for a family of three.
Wage rates are central to concerns about a low-skilled
welfare population that is expected to move into paid
employment. Because the minimum wage is not
indexed to inflation, its real value and protection for
low-wage workers erodes over time. During the continual debates over further increases to the minimum
wage, objections focus on concerns that a higher minimum wage would primarily benefit teenagers not living in poor families, and that it could result in job cuts
and increased unemployment. Although a disproportionate share of minimum wage workers are teenagers,
poor parents make up a sizable share as well. In 1998,
for example, over two million minimum wage workers
were parents over age 25 living with minor children.18
In addition, although many minimum wage studies
have focused on costs to employers and possible unemployment effects as employers reduce the number of
125
Zedlewski
Figure 2
Changes in Policies to Help Make Work Pay
$6.00
80
$5.15
60
$4.25
$4.00
57%
50
Percent
1999 Dollars
70%
70
$5.00
$3.00
39%
40
36%
30
$2.00
20
$1.00
10
$0.00
0
Income Disregardsa
Minimum Wage
Benefit Levelsa
$4,500
$3,820
$4,000
KEY:
$3,500
1995
1999
1999 Dollars
$3,110
$3,000
Sources: Minimum wage data based on figures from the U.S. Bureau of Labor Statistics.
U.S. Department of Labor. Value of the federal minimum wage, 19381997. Available
online at http://www.dol.gov/dol/esa/public.minwage/main.htm. EITC data from Internal
Revenue Service tax forms. Available online at http://www.irs.gov/forms_pubs/index.html.
Income disregards (amounts allowed after 13 months of earnings) and benefit levels
from Zedlewski, S., and Giannarelli, L. Diversity among state welfare programs: Implications for reform. Assessing the New Federalism, Policy Brief A-1. Washington, DC: Urban
Institute, January 1997; and Rowe, G. State TANF policies as of July 1999. Assessing the
New Federalism: Welfare Rules Databook. Washington, DC: Urban Institute, 2000. (For a
state-by-state listing of income disregards and benefit levels in 1999, see Appendix 1 at
the end of this article.)
$2,500
$2,000
$1,500
$1,000
$500
$0
EITC
jobs to hold down costs, most of the studies find only

slight, if any, negative employment effects resulting
from an increase in the minimum wage. Thus, raising
the minimum wage is an important tool available to
the government for helping make work pay.
Tax Policy
The refundable federal EITC is currently the largest program supporting low-income families. The federal EITC
increased substantially in 1993 as a major initiative by
the Clinton administration to make work pay for lower-
126
Average nationwide amount expressed as a percentage relative to the poverty line.
income families. The tax credit provided $30 billion in

support to low-income families with children in 1997.
The maximum federal EITC for a family with two children in 2000 was $3,888, available to families with earnings between $9,700 and $12,700 (92% of poverty).19
The credit is completely phased out when earned
income reaches $31,152 (225% of poverty). The EITC
helps more families escape poverty than any other federal program. Some 4.8 million people, including the families of 2.6 million children, are able to raise their income
above the poverty line as a result of the federal EITC.20
Volume 12, Number 1
Stephen Shames/Matrix
In addition to the EITC, other federal and state tax

credits also have been targeted to low-income families.
For example, in 1998 the federal government implemented both a nonrefundable child tax credit and a
refundable additional child tax credit, for a combined total of not more than $500 for each child.
These credits were available to people with incomes
higher than those qualifying for the EITC, but were
refundable only for families with three or more children. More recently, the Tax Reform Act of 2001
increased the child credit to $600 (escalating up to
$1,000 by 2010) and included a refundable component that will benefit all working families who have
children and earn more than $10,000 per year.23
TANF Policies
State EITCs, a more recent phenomenon, have been

increasing in number and generosity since 1997. Fifteen
states, including the District of Columbia, now offer
EITCs. All but one are based on the federal credit, and
most are refundable.21 TANF regulations issued in April
1999 specifically allow states to finance the refundable
portion of their EITCs from federal TANF monies and
to count credits for low-income families toward
required levels of state expenditures under TANF.22
This ruling may increase the number of states to follow
Wisconsins lead in using TANF funds to support
refundable EITCs as part of a strategy to increase the
net income of low-income families with children.
State TANF policies also have helped create incentives to

work. For example, states typically disregard some
amount of earnings before it affects a familys benefit
level. Benefits are reduced by an increasing percentage of
earnings above the disregard until the benefit goes to
zero. Under welfare reform, many states have expanded
these earned income disregards to allow TANF recipients to keep more of what they earn in the paid labor
market without losing benefits. As illustrated in Figure
2, between 1995 and 1999 the maximum income a family could earn without losing cash benefits increased
from 57% to 70% of poverty level on average nationwide. By 1999 the earned income disregards in 32 states
were more generous than those used under the Aid to
Families with Dependent Children (AFDC) program,
and 15 states allowed families to receive combined cash
welfare and employment earnings that exceeded the
poverty line.24 (For a state-by-state listing of these provisions, see Appendix 1 at the end of this article.)
The high marginal tax rates created by EITCs have

been getting increasing attention because the net payoff from increased work falls precipitously once earnings reach the level at which the EITC is phased out.
Although more generous EITCs obviously provide
greater assistance to low-income working families, they
also create an unavoidable tradeoff: With a more generous credit, more has to be taken away as income
climbs. The common piggybacking of state EITCs
onto the federal EITC means that state EITCs exacerbate the high marginal tax rates already generated by
the federal credit.
At the same time, most states have held their TANF

benefit levels constant since federal welfare reform legislation passed in 1996.25 On average, benefit levels
nationwide actually decreased slightly (see Figure 2).
When benefits do not increase to keep pace with inflation, the decline in real value makes welfare less attractive relative to work. But lower benefit levels also erode
the value of income disregards: Given the same earned
income disregards, TANF recipients can earn more in
states with high benefit levels. In addition, in most
states, families who receive reduced benefits along with
employment earnings use up months that count
127
Zedlewski
Policies designed to make work pay can play an important

role in augmenting gains in low-income working families
economic resources.
against their lifetime limit for benefit receipt. To ensure
that these months will not count against the federal
five-year time limit on assistance, four states (Delaware,
Illinois, Maryland, and Rhode Island) move at least
some families with employment earnings into statefunded TANF programs that are not subject to the
same federal requirements.26
Many states have also increased allowable assets limits
for TANF families, thereby increasing incentives for
recipients to work. For example, the federal AFDC
program placed a $1,500 limit on the value of an automobile that recipients could own, limiting job options
for recipients without access to reliable public transportation. Under TANF, most states increased the
vehicle exemption to more than $4,650, far exceeding
the vehicle exemption under the food stamp program
(see Appendix 1). Some 24 states now allow at least
one car of any value per family.27 The AFDC program
also placed significant limits on recipients financial
assets, reducing incentives to set aside money for emergencies and unexpected expenses. Under TANF, 39
states have increased allowable unrestricted assets (that
is, savings that recipients can use for any purpose) from
$1,000 to $2,000 or more; and 23 states substantially
increased allowable restricted assets (that is, assets that
recipients can set aside for particular purposes, such as
home ownership, education, and business capitalization).28 For a state-by-state listing of these provisions,
see Appendix 1 at the end of this article.
Other Support Policies

Other policies that provide significant support to lowincome working families include the federal food
stamp program, health insurance, child care subsidies,
and child support enforcement.29 For example, the
food stamp program offers assistance to all families
with incomes below 130% of poverty and can be very
valuable to working poor families. A single parent with
two children and earnings at one-half the poverty level
(about $590 per month) could receive $335 per
month in food stamps in 2000, assuming no income
beyond earnings.30
128
Government programs that provide health insurance

coverage also are important to low-income families
who often work in low-paying jobs with no employerprovided benefits. Some 31 states now cover children
in families with incomes up to 200% of the federal
poverty line (or higher) either through Medicaid or the
State Childrens Health Insurance Program
(SCHIP).31 Medicaid is generally available to families
on TANF, and to families leaving TANF for work, for
up to a year if they are earning below 185% of the federal poverty line.32 SCHIP is available to children in
families with income below 200% of the federal poverty line, at a minimum, and may be available to families
with higher incomes, depending on the states Medicaid income threshold. In New Jersey, for example,
SCHIP is available to children in families with income
up to 350% of poverty.33
Child care is an important element in the budget of
most low-income working families, but the value of
child care subsidies is less certain. Child care is not an
entitlement, and benefits vary considerably across
states, depending on whether a family is on welfare,
leaving welfare, or not in the welfare system.
Similarly, states are required to implement programs to
help all families receiving welfare and Medicaid assistance to secure child support payments from noncustodial parents and to help other families not on assistance
who pay a small fee.34 But the amount of payments
received, if any, varies widely by state and by individual
family. States have adopted various guidelines for determining award amounts, resulting in wide differences
across states in awards for families in similar circumstances, particularly at the lower income levels.35 In
addition, many families never collect on their awards, or
never have awards established in the first place.
Policies Combined Impacts

The cumulative effect of policies designed to make work
pay can play an important role in augmenting gains in
low-income working families economic resources. Figure 3 shows the gains achievable as work and wages
increase for low-income families in Texas and California,
Volume 12, Number 1
Figure 3
Texas
Gains Achievable
for Single-Parent
Working Families in
Texas and California
Net Monthly Income after Taxes and Transfers

1999 Dollars
$1,600
$1,541
$1,425
$1,265
$1,200
Poverty Line
$921
$800
$523
$200
$0
No Work
Part Time,
Minimum
Wage
Full Time,
Minimum
Wage
Full Time,
$7/hour
Full Time,
$10/hour
$1,564
$1,541
Full Time,
$7/hour
Full Time,
$10/hour
KEY:
Net Tax Benefits

Food Stamps
TANF
Earnings
Source: Based on calculations from the

Urban Institutes State Income Calculator (STIC), an update of data used in
Acs, G., Coe, N., Watson, K., and Lerman, R. Does work pay? An analysis of
the work incentives under TANF.
Assessing the New Federalism, Occasional Paper No. 9. Washington, DC:
Urban Institute, July 1998.
Net Monthly Income after Taxes and Transfers

1999 Dollars
California
$1,600
$1,480
$1,251
$1,200
Poverty Line
$884
$800
$200
$0
No Work
two states with distinctly different policies. In 1999, a

mother with two children working full time at the minimum wage could exceed the poverty line in both states
if she received the full package of benefits available to
her. These figures also demonstrate how extra supports
from food stamps and the EITC diminish as families
move into higher-paying jobs. In California, for exam-
Part Time,
Minimum
Wage
Full Time,
Minimum
Wage
ple, net family income actually decreases as families move

from earning $7 to $10 per hour. The food stamp benefit is phased out, and the EITC becomes negligible. To
soften the transition, some policymakers have suggested
increasing the federal EITC to implement a slower
phaseout,36 or using different phaseout schedules for
states EITCs as a counterbalance.37
129
Zedlewski
Thus, it appears that a family can be boosted over the

poverty line if the full range of benefits are received.
However, even with all available supports, families
working full time at a minimum wage job do not come
close to reaching the alternative measures of the
resources required to meet their needs, as discussed
earlier. And families often do not receive the full range
of benefits available.
Families Economic Resources

since Reform
Policies that increase work incentives and make work pay
clearly have been enhanced since welfare reform, but the
data suggest that not all low-income families have benefited from these changes. Two somewhat different pictures of family income in the TANF era have emerged so
far. Studies that base their assessment solely on cash
income conclude that poverty has declined as a result of
welfare reform. Other studies, which base their assessment on total family income, including noncash benefits, conclude that a significant segment of families are
worse off as a result of welfare reform. The potential of

government supports to augment earnings and pull
many working families out of poverty is not yet a reality.
An analysis of data from the Census Bureaus Current
Population Survey (CPS) found that between 1995 and
1997, disposable income increased for most single-mother families but not for those among the poorest income
groups.38 (See Figure 4). Total resources declined by
$814 annually for the lowest income group and by $319
for the second-lowest group. As Figure 4 shows, this
decline occurred primarily because of a drop in the number of families receiving TANF and food stamps.
Other analyses of CPS data from 1976 to 1998 point in
the same direction.39 They show that overall, recent
welfare reforms increased family earnings and decreased
poverty, but that the lowest-income families are worse
off since welfare reform. For example, researchers found
that, following the passage of the 1996 legislation,
poverty among female high school dropouts appears
to have declined by 2 percentage points more than it
would have in the absence of policy changes.40
Figure 4
$12,000
$11,584
$11,265
Changes in Annual Income among

Poorest Single-Mother Families:
19951997
Other Incomea
TANF and Food Stamps
EITC
1997 Dollars
KEY:
$6,000
$5,687
$4,873
Earnings
Sources: Primus, W., Rawlings, L., Larin, K., and Porter, K. The
initial impacts of welfare reform on the economic well-being of
single mother families. Washington, DC: Center on Budget and
Policy Priorities, August, 1999. Numbers are based on an analysis of data from the Current Population Survey and include
adjustments for underreporting of welfare and food stamp benefits, which have been shown to worsen during this period.
0
a
130
Other income includes Social Security, veterans benefits,

unemployment insurance, interest and dividends, and child
support.
1995
1997
Lowest Income Decile
1995
1997
Second-Lowest Income Decile
Volume 12, Number 1
However, such analyses also revealed that income gains

did not occur in the post-TANF era among less skilled
mothers in the bottom fifth of the income distribution.
The data showed a substantial decline in income for
women in this group, although the effects were not precisely estimated. One study found an increase in deep
poverty (below 50% of the federal poverty line) among
children between 1995 and 1997.41
In short, the picture of actual incomes in the TANF era
is mixed: Some families have gained as a result of welfare reform and other policies designed to make work
pay, whereas others at the bottom of the income distribution seem to be worse off. These findings highlight
the importance of increasing wages and resources for
working poor families and making it easier for them to
participate in government support programs. Strategies may also focus on increasing the number of hours
of work among single-parent families by ensuring the
availability of low-cost, high-quality child care.42
Work Increasing, but at Low Wages

Motivated by a strong economy and policies that help
make work pay, many more single mothers are working
now compared to earlier periods.43 Between 1995 and
1999, the labor force participation rate for single mothers with children under age 6 increased from 53% to
over 68%, and rates for single mothers with older children rose from 67% to 83% (see Figure 5).44 Among
welfare recipients, participation in the paid labor force
increased from about 9% to 28% over the same period.45,46 Studies show that the proportion of families
Figure 5
Labor Force Participation among Single Mothers and AFDC/TANF Recipients (19921999)
90
80
Percent
60
50
70
Federal Welfare Reform
40
30
1998
1999
20
10
1992
1993
1994
1995
1996
1997
Year
KEY:
TANF Recipients
Single Mothers with Children under Age 6
Single Mothers with Children Ages 6 to 17
Sources: U.S. Census Bureau. Statistical abstract of the United States, 2000: The national data book. Washington, DC: U.S. Census Bureau, 2000, table 653; and Administration for Children and Families, Office of Planning, Research and Evaluation. Temporary Assistance for Needy Families: Third annual report to Congress. Washington,
DC: U.S. Department of Health and Human Services, August 2000, table 10.1.a.
131
Zedlewski
who are employed after leaving welfare has also

increased. A nationally representative Urban Institute
study found that 75% of families who left welfare
between 1995 and 1997 included at least one employed
adult, as did 79% of those who left between 1997 and
1999.47 In studies from eight states, employment rates
among single-parent welfare leavers ranged from 50%
to 64% in the first three months after leaving welfare.48
Studies to date indicate that earnings for families leaving welfare are generally low, barely reaching the
poverty line. The Urban Institute study of welfare
leavers found that median monthly wages were $1,093
in 1999, just under the federal poverty level for a family of three.49 According to studies in eight states, average monthly earnings of former welfare recipients
during the first three months after leaving welfare were
even lower, generally ranging from $733 to $900.50 It
is clear that many low-income working families do not
earn enough to support a family of three.
In addition, few studies compare postwelfare income
with the income families would have received if they had
remained on aid. Based on individuals assessments of
Figure 6
Source: Zedlewski, S. Former welfare families and the Food

Stamp program: The exodus continues. Assessing the New
Federalism, Policy Brief B-33. Washington, DC: Urban Institute,
April 2001, p. 3.
a
132
Sample taken from the Urban Institutes 1999 National Survey

of Americas Families. Includes families who received some
TANF benefits since January 1997 but were not receiving
TANF benefits when interviewed in 1999.
Work Supports Often Not Received

For families earning low wages, government supports
such as the EITC, food stamps, health insurance, and
subsidized child care are essential to maintaining sufficient economic resources. Studies show that participation in the EITC is generally strong. At least 85% of
eligible families receive the EITC, according to rough
estimates.5254 But the same is not true of other programs intended to help make work pay. Studies show
that a substantial number of families leaving welfare do
not avail themselves of other, nontax supports to augment their earnings.
60
Percent of Welfare Leavers Receiving Food Stamps
Food Stamp Program Participation

Rates after Leaving Welfare,
by Income Level (1999)
their family income, some studies suggest that one-half

to two-thirds of families who leave welfare have higher
incomes after leaving the rolls. A study in Wisconsin,
however, estimated that more than half the families leaving welfare in that state have lower incomes than they
would have if they had stayed on welfare.51 No studies
have been able to provide information about changes in
income for those who have never become welfare recipients because of new policies adopted since reform.
51%
50
44%
43%
40
30%
30
20
10
0
All Eligibles
Below 50%
of Poverty
50 to 100%
of Poverty
100 to 130%
of Poverty
Estimated Monthly Income Relative to Povertya
Volume 12, Number 1
Figure 7
Health Insurance Coverage after Leaving Welfare (1997)
Percent of Welfare Leavers with Health Insurance
60
56%
49%
50
40
35%
34%
37%
29%
30
28%
22%
20
12%
10
0
Less than 6 months
6 to 11 months
One year or more
Number of Months since Leaving Welfare

KEY:
Private Insurance
Medicaid
Uninsured
Source: Garrett, B., and Holahan, J. Welfare leavers, Medicaid, and private health insurance. Assessing the New Federalism Series, No. B-13,
Washington, DC: Urban Institute, March 2000, p. 3.
Note: Totals by age group may not add to 100% due to rounding.
Food Stamps
For example, many families do not continue to
receive food stamps once they leave welfare. One
study found that 57% of families leaving welfare were
not receiving food stamps, even though they were
eligible.55 (See Figure 6.) More surprising, even half
the families with incomes below 50% of poverty were
not receiving food stamps. State studies that examine
participation across government programs also confirm this result. Only about half the families leaving
welfare used food stamps in the first three months
after exit, and receipt was significantly lower in most
states after families had been off the rolls for a year.56
Medicaid
Recent studies indicate similar trends in Medicaid
coverage. Although Medicaid offers a transitional
benefit of 12 months coverage to adults who leave

welfare for work, states report that continued coverage among this group varied from about 30% to
60% in the first three months after exit, and declined
significantly after leavers were off welfare for a year.
Data from the Urban Institutes National Survey of
Americas Families show that although more than
half of welfare leavers had Medicaid coverage in the
first six months of leaving the rolls, this rate of coverage dropped quickly.57 About half the mothers
were uninsured one year after leaving welfare. (See
Figure 7.)
Child Care
Child care is a critical issue for families economic
resources; the cost of child care can be a significant
strain on low-income families, consuming as much as
133
Zedlewski
The 1996 law gives states the flexibility to use federal and state
TANF funds to design new supports for working poor families,
including those who have leftor who have never receivedwelfare.
20% of their income.58 At the same time, information
about low-income families receipt of child care assistance is relatively scant, although it seems that help
in paying for child care is infrequent. For example, in
the Urban Institute study of families leaving welfare
between 1995 and 1997, only 19% reported receiving government help in paying for child care after
three months.59 This figure rose to 22% for families
leaving welfare between 1997 and 1999.60 The
recent expansion of child care funding through both
TANF and the Child Care and Development Fund
should help to expand assistance to low-income families, but many policy analysts argue that funding is
not adequate. Rough estimates suggest that about 2
million children are served through current funding
streams, although as many as 15 million children live
in families that meet the federal eligibility standards.61 (For further discussion of child care and welfare reform, see the article by Fuller and colleagues
Child Support Enforcement
Stronger child support enforcement was a goal of
welfare reform, as child support can be an important
supplement to single-parent families income and
help them move off welfare. One study found that
among poor, nonwelfare families who received child
support, it provided 35% of their income.62 Unfortunately, most poor children eligible for child support do not receive it.63 The Urban Institute study
of families leaving welfare between 1995 and 1997
found that of those who stayed off welfare, only
34% received child support.58 Moreover, state studies report that only 13% to 36% of families leaving
welfare receive child support.64 Although federal
and state government policies devoted to strengthening child support enforcement over the last two
decades have increased child support receipt significantly among both never-married and previously
married single mothers,65 the poorest children have
received only modest increases.66 It is difficult to
expand child support for poor children because
absent parents who do not pay support are, on aver-
134
age, more disadvantaged than absent parents who

do pay.67 (See the article by McLanahan and Carlson
In sum, low participation rates are found across many
government programs designed to support the working poor. Researchers conducting field studies attribute families lack of participation to a number of
factors. Some blame states failure to develop effective
administrative systems to deliver assistance to the
working poor.68 State and local offices seldom get the
word out about benefit eligibility to low-income families; administrative hurdles can be formidable, especially for working families, and many caseworkers do
not understand the complex eligibility rules that vary
across different benefit programs. Some critics argue
that the low participation rates, especially the drop
over time in Medicaid coverage, may be due to
administrative complexities.69 Eligibility for the full
12 months of Medicaid benefits requires submitting
information on earnings and child care costs in the
4th, 7th, and 10th months after leaving welfare. Stigma can also play a role; for most benefits, the point of
entry is still the local welfare office. To ensure that
stigma does not present a barrier to obtaining nonwelfare assistance among working poor families, some
states are developing alternative entry points, such as
local health clinics and child care referral programs.
Meanwhile, the continuing demand for emergency food
and shelter assistance, even among working families,
emphasizes the critical need for income support programs. In a survey of city officials conducted in December 2000, it was estimated that 32% of adults requesting
food, and 26% of those requesting shelter, were
employed.70 Moreover, officials estimated that about 62%
of those requesting emergency food and 36% of those
requesting emergency shelter were families with children.
Between 1995 and 2000, the proportion of families with
children seeking such emergency assistance remained fairly constant, but a growing number of cities reported
increased demand from families for these services.
Volume 12, Number 1
New Types of Support Emerging

under TANF
The 1996 law gives states the flexibility to use federal

and state TANF funds to design new supports for
working poor families, including those who have left
or who have never receivedwelfare. In addition, the
final TANF regulations, effective in October 1999,
clarified the types of assistance that were subject to the
federal time limit and work requirements. Many supports were specifically excluded from these requirements,71 including child care; transportation;
earmarked savings accounts (referred to as Individual
Development Accounts, or IDAs); refunded portions
of state-earned income tax credits; nonrecurrent,
short-term benefits designed to deal with crisis situations; and case management. These regulations confirmed and increased states spending flexibility,
providing an incentive to broaden benefits for lowincome families with children and to experiment with
innovative services. Two broad types of innovations in
the use of TANF monies for working poor families
including families both on and off welfareare highlighted below: extended case management, and direct
supports such as transportation and housing subsidies.
Case Management, Information, and

Referral Services
Families leaving welfare typically are employed in
low-wage jobs that lack benefits and are often shortlived. Extended case management can provide such
families with a network of services to help them stay
employed and move ahead in their careers by addressing personal and work-related barriers, and to ensure
they stay connected to benefits after leaving welfare.72
For example:
Vermont has established a statewide network of 16
centers that provide early childhood and family supports, including home visiting, early childhood services, parent education, peer support for parents, and
information and referral on resources for families.
Maryland has established a network of 27 centers
across the state to improve education, training, and
job retention through intensive case management
and service coordination.
The Vocational Foundation, Inc. (VFI) in New York

City provides follow-up services to help parents
retain jobs and move ahead in their careers by
addressing personal barriers; ensuring receipt of key
work supports, including food stamps, Medicaid,
child care subsidies, and the EITC; and offering
career counseling. Case managers also assist in resolving conflicts at the workplace or in providing immediate reemployment services.
The Workforce Information System of Texas
(TWIST) has consolidated job training and employment services into one-stop centers to create a single
intake for all clients, including welfare recipients.
Case managers review all services clients have
received and assist them in finding employment.
In Rhode Island, caseworkers in employment retention services units work with TANF recipients and
135
Zedlewski
A significant minority of families at the bottom of the income

distribution seem to be worse off ... [due to] ... low wages and
failure to benefit from various support programs.
employers. Case managers build relationships with
recipients, including detailed assessments of barriers
and supports needed to prevent potential problems
on the job.
Washingtons Work First program was designed to
give all low-wage workers, not just those leaving welfare, the supports they need to find a job, learn new
skills, and move up the career ladder so they will be
self-supporting for life.73 The state has mounted
media campaigns to publicize the availability of child
care and to encourage workers to take advantage of
the EITC, and has set up a network of call centers to
provide low-wage workers with information about
benefit programs.
Direct Supports
All states use TANF monies to fund direct supports for
low-income families, but more recently, states have
begun to expand the types of supports offered and the
populations served. For example, all states use at least
some TANF funds to finance child care subsidies,74 and
many use TANF funds for initiatives to enhance child
care quality. But some states now have begun using
TANF resources to increase child care capacity in a
variety of ways.75 Iowa has used TANF funds to
strengthen local child care capacity as part of a comprehensive approach to supporting child development.
Florida has also used TANF funds to increase child care
supply, and Washington has used TANF to provide
incentives to improve pay at centers that accept subsidized child care payments.
Because transportation can be a significant barrier to
employment, some states have begun to use TANF
funds to expand transportation aid for low-income
workers. Poor working families often do not live near
job centers, have no easy public transportation, and
cannot afford the cost of a car. To address these needs,
New Mexico now allows families with incomes below
poverty level to qualify for transportation aid funded
by TANF.72 In Kansas, Michigan, Pennsylvania, and
Virginia, car ownership is facilitated either by providing
a down payment on the purchase of a car or, as in
136
Kansas, by providing up to $5,000 toward the cost of

purchasing a car. Tennessee uses TANF funds to offer
low-income families no-interest, no-down-payment
loans on high-quality used vehicles, as highlighted in
Appendix 2 at the end of this article.
Some states also have begun to use TANF monies to
expand housing assistance for low-income families.
Most poor families do not receive federal housing assistance and have to pay the full cost of housing in the
private market. They typically pay as much as half their
income on housing or live in severely substandard
structures.76 To help meet this essential need, several
states have begun using TANF monies (alone or in
combination with other funds) to establish programs
providing housing assistance. New programs implemented in California, Connecticut, Kentucky, Maryland, Minnesota, New Jersey, and North Carolina vary
in nature and in the populations targeted, and are generally modest in size, but they show how some states
are reaching beyond the provision of cash assistance to
help low-income families.77 Some programs provide
rental assistance, whereas others focus on encouraging
home ownership. Most pay housing costs in excess of
some percentage of family income, ranging from 30%
in Minnesota to 45% in New Jersey. (See Appendix 2
for highlights of the New Jersey program.)
Summary and Conclusions

Increasing families economic resources and child wellbeing were not the primary goals of welfare reform. The
reform movement initially focused on the unprecedented declines in caseloads and the relative employment
success of mothers who left welfare, and paid little
attention to how families were getting by. The importance of families economic resources to child wellbeing is just beginning to be understood. Although the
connections are complex and vary by the childs age,
research provides compelling evidence that higher
incomes can help parents purchase enriching experiences for their preschool and school-age children, leading to positive impacts on their development.
Volume 12, Number 1
More recently, attention has begun to focus on the

ability of single mothers and other working poor families to support their families on wages alone. Federal
and state policies designed to increase families
resources by making work payincluding increasing
the minimum wage, expanding the EITC, changing
welfare rules to allow recipients to collect both a paycheck and some cash assistance, and maintaining food
stamps for all low-income familiescould effectively
lift full-time workers earning the minimum wage out of
poverty. However, many studies confirm that families
who leave welfare also leave behind government assistance for food, health insurance, and child care.
Many families have gained economic resources as a result
of welfare reform and other policies implemented by
federal and state governments to make work pay, but a
significant minority of families at the bottom of the
income distribution seem to be worse off. Key factors
contributing to the plight of this group are low wages
and failure to benefit from various support programs.
Many working families subsist on incomes below the
poverty line. Moreover, even an income just above
poverty is not sufficient to provide an adequate standard
of living for families with children. Child care and housing costs can absorb large shares of their incomes, leaving little to cover other necessities. Many families need
supports to help meet these needs as they transition to
self-sufficiency. Families also need the support of both
parents. Increased child support payments from noncustodial parents can play an important role in helping bolster the resources of single-parent working poor families.
But this is still an early picture of how families are faring in the post-reform period. Survey data showing
changes in family resources lag behind policy changes
by a year or two. Many states have begun to spend larger shares of their TANF block grants on new types of
supports for low-income families. Some are stressing
case management and outreach to low-income families, hoping to increase their knowledge of benefit eligibility and to maintain essential supports for them
when they leave welfare. Others are providing direct
assistance to families leaving welfare, such as housing
and transportation. Such noncash assistance can make
a big difference in families economic resources. Also,
how states are coping with federal benefit time limits is
only beginning to be revealed.
States TANF programs continue to evolve as it

becomes clearer how the federal block grant monies
can be spent, and states become more aware that many
families who leave welfare do not receive other services. Most states are still struggling to design effective
delivery systems that allow families outside the cash
assistance system to navigate eligibility for supports
more efficiently. Effective policies to inform lowincome families of their benefit eligibility, systems that
make it easier to retain eligibility, and broader support
services in all states could change the post-TANF picture that has emerged thus far.
Newer data reflecting the post-reform period will be
required to better understand how families economic
resources have changed since TANF and how these
changes are affecting children. The evolving story of
welfare reform will need to be monitored carefully to
achieve not only smaller caseloads, but also long-term
positive impacts on family economic resources and
child well-being.
The author wishes to thank Tracy Roberts, who provided expert research assistance; Greg Duncan, who
reviewed this paper; and The David and Lucile
Packard Foundation for a very insightful, anonymous
review of the draft. This article was prepared as part
of the Urban Institutes Assessing the New Federalism
project, with primary funding from the Annie E.
Casey Foundation, the Ford Foundation, The W. K.
Kellogg Foundation, The David and Lucile Packard
Foundation, and the Robert Wood Johnson Foundation. The opinions represent those of the author and
not the Urban Institute nor its sponsors.
137
Zedlewski
ENDNOTES
1. Personal Responsibility and Work Opportunity Reconciliation Act
of 1996. Public Law 104-193. Section 101, 110 Stat. 2110,
August 22, 1996.
2. Mayer, S. Trends in the economic well-being and life chances of
Americas children. In Consequences of growing up poor. G. Duncan and J. Brooks-Gunn, eds. New York: Russell Sage Foundation, 1997, pp. 4969. See also Brooks-Gunn, J., and Duncan,
G. The effects of poverty on children. The Future of Children
(Summer/Fall 1997) 7(2):5571.
Press, 1994. This study was based on data from the Panel Study
of Income Dynamics (PSID). Surveys confirm that income
accounts for about half of the difference in key educational outcomes between children in single- and two-parent families.
4. In fact, the study, also using PSID data, showed that most of the
other factors thought to affect child well-beingsuch as low
parental education, single-motherhood, and teen pregnancyare
closely related to issues of income. See Duncan, G., BrooksGunn, J., Smith, J.R., and Yeung, W.J. Does childhood poverty
affect the life chances of children? American Sociological Review
(June 1998) 63:40623.
5. Research does not consistently support or refute the contention
that income from welfare is less beneficial to children than other
forms of income because it carries a stigma. Many studies have
found no relation between welfare receipt and childrens cognitive and social development once other factors are taken into
account. See, for example, Haveman, R., and Wolfe, B. The
determinants of childrens attainments. Journal of Economic Literature (1995) 33(4):182978; and Yoshikawa, H. Welfare dynamics, support services, mothers earnings, and child cognitive
development: Implications for contemporary welfare reform.
Child Development (1999) 70:779801. Other studies show that
transitions into welfare are associated with more behavioral problems in children, but this effect also may be associated with other
accompanying transitions (such as a loss of job or parents separation). See, for example, Smith, J.R., and Brooks-Gunn, J. Development effects of natural transitions in welfare receipt. Paper
presented at a research briefing of the National Academy of Sciences, Board on Children and Family. Washington, DC. December 56, 1994.
6. Mayer, S. What money cant buy: The effect of parental income on
childrens outcomes. Cambridge, MA: Harvard University Press,
1997.
7. Rector, R., and Fagan, P. How welfare harms kids. Backgrounder
No. 1084. Washington, DC: Heritage Foundation, June 5, 1996.
policies affect children: A synthesis of research. New York: Manpower Demonstration Research Corp., March 2001. Together, these
5 studies examined the effects of 11 different employment-based
welfare and antipoverty programs aimed primarily at single-parent
families. All the evaluations used rigorous random-assignment
research designs. Program effects are essentially measured as the
difference between outcomes for families in a program group
with new benefits and rules, and families in a control group with
benefits and rules previously in place.
9. The study did caution, however, that the improvements in child
138
well-being were modest when considered in the context of these

childrens high levels of disadvantage. The study also concluded
that programs that increased parental employment without
increasing income had few effects on children. (For a detailed
analysis of these and other studies, see the article by Zaslow and
colleagues in this journal issue.)
10. Two of the experimental programs that increased parental work
and economic resources increased both adolescent problem
behaviors and decreased adolescent school achievement. One possible reason for these findings is that adolescent development may
be more strongly influenced by the affection, supervision, role
modeling, and mentoring from parents than by increased economic resources. But some researchers question whether outcomes would have been positive for adolescents if more
age-appropriate enrichment activities had been available. See
Duncan, G., and Chase-Lansdale, P.L. Welfare reform and childrens well-being. In The new world of welfare. R.M. Blank and R.
Haskins, eds. Washington, DC: Brookings Institution Press,
2001, pp. 391417.
11. Only one experimental study has assessed the impacts on very
young children of programs that increased parental employment
and economic resources. In this study, the program showed no
effects on achievement and behavior for children ages three to
five. See Morris, P., and Michalopoulos, C. The Self-Sufficiency
Project at 36 months: Effects on children of a program that
increased parental employment and income. Ottawa, Ontario:
Social Research and Demonstration Corp., 2000. Other studies
have shown that very young children may be more vulnerable
than older children to negative effects from employment-induced
separation from their mother. See note 10, Duncan and ChaseLansdale.
12. For a full description of this measure, see Citro, C.F., and
Michael, R.T., eds. Measuring poverty: A new approach. Washington, DC: National Academy Press, 1995. See also Betson, D.,
and Michael, R. Why so many children are poor. The Future of
Children (Summer/Fall 1997) 7(2):2539.
13. Edin, K., and Lein, L. Making ends meet. New York: Russell Sage
Foundation, 1997.
14. Bernstein, J., Broeht, S., and Spade-Aquilar, M. How much is
enough? Basic family budgets for working families. Washington,
DC: Economic Policy Institute, 2000.
15. This idea originated with Victor Fuchs in the 1960s and has been
used as an alternative measure of income adequacy since then. See
Fuchs, V.R. Redefining poverty and redistributing income. The
Public Interest (Summer 1967) 8:8895.
16. See the U.S. Department of Labor Web site at
http://www.dol.gov/dol/esa/public/whd_org.htm. Note that
the majority of workers not covered by the FLSA are executive,
administrative, or professional employees.
17. The 10 states with higher minimum wages are Alaska, California,
Connecticut, Delaware, Hawaii, Massachusetts, Oregon, Rhode
Island, Vermont, and Washington.
18. Turner, M. Does the minimum wage help or hurt low-wage
workers? In The low-wage labor market. Washington, DC: U.S.
Department of Health and Human Services, December, 1999,
p. 84.
Volume 12, Number 1
19. The percent relative to poverty uses a poverty threshold for a single parent with two children. The amount of credit a family is eligible for does not depend on marital status, which leads to
potential marriage penalties for people who would each qualify
for the EITC while single but qualify for only one credit as a
couple if they marry. See Wheaton, L. Low-income families and
the marriage tax. Policy brief. Washington, DC: Urban Institute,
September 1998.
20. Johnson, N. A hand up: How state earned income tax credits help
working families escape poverty in 2000. Washington, DC: Center
on Budget and Policy Priorities, 2000.
21. Ten states (Colorado, Kansas, Maine, Maryland, Massachusetts,
Minnesota, New Jersey, New York, Vermont, and Wisconsin) and
the District of Columbia offer refundable EITCs ranging from
10% to 34% of the federal credit (for a family of two), with most
falling at the low end of this range. Four states have EITCs that
are not refundable (Illinois, Iowa, Oregon, and Rhode Island).
These nonrefundable credits also tend to be less generous, falling
around 5% of the federal credit in all states except Rhode Island,
which has a 26% credit. For an in-depth discussion of state
EITCs, see Maag, E., and Rogers, D. The new federalism and
state tax policies toward the working poor. Assessing the New Federalism Occasional Paper No. 38. Washington, DC: Urban Institute, 2000.
22. States must maintain specific levels of funding for their TANF
programs under a state maintenance-of-effort (MOE) requirement. For more details, see the article by Greenberg and colleagues in this journal issue.
23. Greenstein, R. The changes the new tax law makes in refundable
tax credits for low-income working families. Washington, DC:
Center for Budget and Policy Priorities, June 18, 2001.
24. Rowe, G. The Welfare Rules Databook: State TANF policies as of
July 1999. Assessing the New Federalism. Washington, DC:
Urban Institute, 2000, table L.4.
25. Four states (Idaho, Illinois, Oklahoma, and Wyoming) and the
District of Columbia reduced maximum TANF benefit levels
between 1996 and 1999, and 12 states (Maine, Maryland, Mississippi, Montana, Nebraska, New Mexico, North Dakota, Oregon, Utah, Vermont, West Virginia, and Wisconsin) increased
their maximum benefit levels. See note 24, Rowe.
Policy Priorities. The state policy documentation project. Washington, DC: CLASP and CBPP, June 2000. Available online at
http://www.spdp.org.
27. See note 24, Rowe, table I.C.1, p. 46.
28. Restricted assets include Individual Development Accounts
(IDAs), authorized by federal welfare legislation that allows states
to create community-based IDA programs with TANF block
grant funds and to disregard all money saved in IDAs in determining eligibility for all means-tested government assistance. All
deposits into IDAs are limited to earned income. See Sherraden,
M., Johnson, L., Clancy, M., et al. Saving patterns in IDA programs: Downpayments on the American Dream Policy DemonstrationA national demonstration of Individual Development
Accounts. St. Louis: Center for Social Development, Washington
University, January 2000.
29. See the article by Greenberg and colleagues in this journal issue
for a detailed discussion of how federal welfare reform legislation
changed these programs.
30. This calculation assumes the maximum child care cost deduction
for children older than age two, and no excess shelter costs. For a
full description of the food stamp program, see Zedlewski, S.R.
Former welfare families continue to leave the food stamp program.
Assessing the New Federalism Discussion Paper 01-05. Washington, DC: Urban Institute, March 2001.
Washington, DC: U.S. Government Printing Office, October
2000, p. 932, tables 1527.
32. Technically, these two programs were delinked by federal welfare reform, but states have generally maintained eligibility for
adults on TANF.
33. The basic rule is that states can set a SCHIP eligibility threshold
that is 50 percentage points above the level used in their Medicaid program. See note 31, U.S. House of Representatives, Committee on Ways and Means, p. 931.
and Means, p. 464.
and Means, pp. 48184.
36. For example, some members of the Democratic Leadership
Council have advocated providing the EITC to all taxpayers,
regardless of their income.
37. In Minnesota, for example, the state EITC plateaus at two
income levels to soften the effect of the federal credit phaseout.
For families with two or more children, the credit rises from
$840 (22% of the federal) to $1,222 (41% of the federal) when
earnings increase from $14,600 to $17,600, and is completely
phased out once a familys earnings reach $30,600. See note 21,
Maag and Rogers, for a complete description of how state EITCs
work.
impacts of welfare reform on the economic well-being of single mother families. Washington, DC: Center on Budget and Policy Priorities, August 1999.
39. These analyses used three alternative methods (difference-in-difference, regression models, and residual analyses) to assess
changes in income. Results are not directly comparable with
those of the Center on Budget and Policy Priorities and others
because they use a cash income measure of poverty. See Schoeni,
R., and Blank, R. What has welfare reform accomplished? Impacts
on welfare participation, employment, income, poverty, and family
structure. Santa Monica, CA: RAND and the Gerald R. Ford
School of Public Policy, University of Michigan, February 2000.
40. See note 39, Schoeni and Blank, p. 26.
41. Haskins, R. Effects of welfare reform on family income and
poverty. In The new world of welfare. R. Blank and R. Haskins,
eds. Washington, DC: Brookings Institute Press, 2001,
pp. 10336.
42. Most low-income single parents who are employed already work
139
Zedlewski
a substantial amount. For example, 68% of former welfare recipients who were working in 1999 worked full time (35 hours per
week or more). See Loprest, P. How are families that left welfare
doing? A comparison of early and recent welfare leavers. Assessing
the New Federalism, Policy Brief No. B-36. Washington, DC:
52. Eligibility is estimated from surveys including all families,

whether or not they file income taxes, compared to families who
did claim this tax credit.
43. The relative contributions of a strong economy, welfare reform,

and other federal policies to the increase in single mothers work
and earnings are as yet unclear. However, the evidence to date
suggests that all three factors matter. For example, one study
attributed one-third of the increase in employment of single
mothers between 1992 and 1996 to the expansion of the federal
EITC. See Council of Economic Advisors. The economic report of
the president. Washington, DC: U.S. Government Printing Office,
February 1999. See also note 39, Schoeni and Blank. They provide the first study designed to estimate the direct impact of welfare policy changes on work behavior, earnings levels, and income.
The authors found that welfare waivers increased labor market
involvement among less-skilled women, but the 1996 federal
reforms had little additional impact on work behavior after controlling for economic forces. That is, ongoing increases in work in
the post-1995 period were driven by the economic expansion.
54. The IRS worries more about families who are not eligible receiving tax credits than about nonparticipation. A variety of studies
place the error rate in the EITC around 15% to 33%. See Liebman, J. Who are the ineligible EITC recipients? Cambridge, MA:
Harvard University Press, October, 1995.
44. U.S. Census Bureau. Statistical abstract of the United States,

2000: The national data book. Washington, DC: U.S. Census
Bureau, 2000, table 653. Unemployment rates also decreased
sharply in 1998 for single mothers with younger and older children, to 17% and 12%, respectively.
58. Giannarelli, L., and Barsimantov, J. Child care expenses of Americas families. Assessing the New Federalism Occasional Paper No.
40. Washington, DC: Urban Institute, 2000.
45. Between 1992 and 1997 the percentage of adult welfare recipients who participated in work activities tripled, and between
1997 and 1999 the state-reported work participation rate for
TANF recipients increased from 28% to 38%. See Administration
for Children and Families, Office of Planning, Research, and
Evaluation. Temporary Assistance for Needy Families: Third annual report to Congress. Washington, DC: U.S. Department of
Health and Human Services, August 2000.
53. Another study indicates that participation may be particularly low

for Hispanics, however. See Phillips, K. Who knows about the
Earned Income Tax Credit? Policy Brief B-27. Washington, DC:
Urban Institute, January 2001.
55. See note 30, Zedlewski.

56. See note 48, Acs and Loprest, table v.3.
57. Garrett, B., and Holahan, J. Welfare leavers, Medicaid coverage,
and private health insurance. Assessing the New Federalism
Series, No. B-13. Washington DC: Urban Institute, March 2000.
59. See note 47, Loprest.

60. Loprest, P. How are families who left welfare doing over time? A
comparison of two cohorts of welfare leavers. In Economic policy
review. New York: The New York Federal Reserve, September
2001, pp. 919.
61. Layzer, J. Child care funding: How much is needed and is there
enough? Statement presented at the Brookings Institute Forum.
Washington, DC. June 13, 2001.
46. According to the National Survey of Americas Families, some

33% of adult TANF recipients had earnings at some point during
1999, compared with 22% in 1997. See Zedlewski, S., and Alderson, D. Do families on welfare in the post-TANF era differ from
their pre-TANF counterparts? Assessing the New Federalism Discussion Paper 01-03. Washington, DC: Urban Institute, 2001.
against poverty. Assessing the New Federalism Series, No. B-10.
Washington, DC: Urban Institute, 2000.
47. Loprest, P. Families who left welfare: Who are they and how are
they doing? Assessing the New Federalism Discussion Paper 9902. Washington, DC: Urban Institute, 1999; see also note 42,
Loprest.
64. Brauner, S., and Loprest, P. Where are they now? What states
studies of people who left welfare tell us. Assessing the New Federalism Series, No. A-32. Washington, DC: Urban Institute, May
1999, table 8.
48. The eight states are Arizona, Florida, Georgia, Illinois, Missouri,
New York, Washington, and Wisconsin. See Acs, G., and Loprest,
P. Initial synthesis report of the findings from ASPEs leaver grants.
Washington, DC: Department of Health and Human Services,
Office of the Assistant Secretary for Planning and Evaluation,
January 2001. Note that leaver studies typically report employment rates for single-parent families rather than family employment rates that include employment among spouses and partners.
65. Sorensen, E., and Halpern, A. Child support enforcement is working better than we think. New Federalism: Issues and Options for
States, No. A-31. Washington, DC: Urban Institute, 1999.
49. See note 47, Loprest, and note 42, Loprest. As Loprest points
out, however, we can only estimate how monthly earnings translate into annual income. Some adults may work fewer hours over
the year or work only a portion of the months in a year, leading
to lower annual income than the monthly levels suggest.
50. This is due in part to the fact that, unlike the Urban Institute
study, the states included only the earnings of the former recipients themselves, excluding earnings from other adults in the
household.
140
51. Cancian, M., Haveman, R., Kaplan, T., and Wolfe, B. Post-exit
earnings and benefit receipt among those who left AFDC in Wisconsin. Madison, WI: Institute for Research on Poverty, 1998.
63. About 70% of poor children eligible for child support were not
receiving it in 1996. See note 62, Sorensen and Zibman, p. 4.
66. Between 1993 and 1999, child support for single mothers in the
bottom fifth of the income distribution increased, on average,
only from $253 to $365 (both figures measured in 1999 dollars).
See note 41, Haskins.
67. Sorensen, E., and Zibman, C. Poor dads who dont pay child support: Deadbeats or disadvantaged? Assessing the New Federalism
Series, B-30. Washington, DC: Urban Institute, 2001.
68. Meyers, M. How welfare offices undermine welfare reform. The
American Prospect (June 19July 3, 2000) 11(15):4045.
69. Greenstein, R., and Guyer, J. Supporting work through Medicaid
and food stamps. In The new world of welfare. R. Blank and R.
Haskins, eds. Washington, DC: Brookings Institute Press, 2001,
pp. 33568.
Volume 12, Number 1
70. Lowe, E.T. A status report on hunger and homelessness in Americas cities 2000: A 25-city survey. Washington, DC: U.S. Conference of Mayors, December 2000.
75. For examples of states pursuing this strategy, see Mitchell, A.,
Stoney, L., and Dichter, H. Financing child care in the United
States. Philadelphia: Ewing Marion Kauffman Foundation, 2001.
71. As long as the family has an employed adult. See Greenberg, M.,
and Savner, S. The final TANF regulations. Washington, DC:
Center for Law and Social Policy, May 1999.
76. In 1995, on average, households with incomes below 50% of the

area median income spent more than half of their income on
housing or lived in severely substandard housing. See U.S.
Department of Housing and Urban Development. Rental housing assistancethe crisis continues: The 1997 report to Congress on
worst case housing needs. Washington, DC: DHUD, 1998.
72. For descriptions of these and many other strategies states use to
support families, see Sweeney, E., Schott, L., Lazere, E., et al.
Windows of opportunity: Strategies to support families receiving welfare and other low-income families in the next stage of welfare
reform. Washington, DC: Center on Budget and Policy Priorities,
January 2000.
73. Massing, M. Ending poverty as we know it. The American
Prospect (June 19July 3, 2000) 11(15):3038.
74. Child care subsidies are funded either as direct TANF or MOE
expenditures or through allowable transfers from the TANF block
grant to the Child Care Development Fund. See the article by
Greenberg and colleagues in this journal issue for further details.
77. In six states (Connecticut, Kentucky, Maryland, Minnesota, New

Jersey, and North Carolina), the programs are statewide, but in
California, the programs are located in just two counties: Los
Angeles and San Mateo. In Connecticut, the program is limited
to working families that lose TANF benefits due to the states 21month time limit. See Sard, B., and Lubell, J. The increasing use
of TANF and state matching funds to provide housing assistance to
families moving from welfare to work. Washington, DC: Center on
Budget and Policy Priorities, February 2000.
141
Zedlewski
APPENDICES
Appendix 1
State TANF Policies Affecting Family Economic Resources (as of 1999)
Maximum
income
disregards
relative to
povertya
Maximum
benefit
levels
relative to
povertyb
Combined
disregards
and benefit
levels
relative to
povertyc
Unrestricted
asset
levelsd
Restricted
asset
levelse
Vehicle
exemption
limits
18.3%
14.7%
33.0%
$2,000
One vehicle per driver
Alaska
143.6%
82.5%
179.4%
$1,000
Arizona
52.4%
31.0%
52.4%
$2,000
$9,000
One vehicle per household
Arkansas
39.9%
18.2%
58.1%
$3,000
California
129.4%
54.6%
155.4%
$2,000
$5,000
$4,650
Colorado
45.7%
31.9%
60.0%
$2,000
Determined by county
103.4%
48.5%
152.0%
$3,000
No limit
$9,500
District of Columbia
42.8%
33.9%
59.8%
$2,000
$1,500
Delaware
80.5%
30.2%
110.8%
$1,000
$5,000
$4,650
Florida
72.1%
27.1%
90.5%
$2,000
$5,000
$8,500
Georgia
46.0%
25.0%
60.3%
$1,000
$5,000
$4,650
Hawaii
146.7%
63.7%
146.7%
$5,000
Idaho
56.9%
24.7%
56.9%
$2,000
$4,650
Illinois
101.2%
33.7%
121.0%
$2,000
No limit
Indiana
33.8%
25.7%
42.3%
$1,000
$5,000
Iowa
76.2%
38.1%
76.2%
$2,000
All deposits and interest
$3,916
Kansas
72.0%
38.4%
87.3%
$2,000
Kentucky
55.1%
23.4%
55.1%
$2,000
$5,000
Louisiana
27.7%
17.0%
44.7%
$2,000
$6,000
$10,000
Maine
91.5%
41.2%
103.3%
$2,000
$10,000
Maryland
54.9%
35.7%
61.6%
$2,000
Massachusetts
93.4%
50.5%
120.1%
$2,500
$5,000
Michigan
69.2%
41.0%
69.2%
$3,000
Minnesota
112.9%
47.6%
130.0%
$2,000
$7,500
Mississippi
40.9%
15.2%
49.0%
$2,000
$4,650
Missouri
34.2%
26.1%
34.2%
$1,000
No limit
Montana
71.3%
41.9%
73.1%
$3,000
Nebraska
59.7%
47.8%
59.8%
$4,000
State
Alabama
Connecticut
142
Volume 12, Number 1
Maximum
income
disregards
relative to
povertya
Maximum
benefit
levels
relative to
povertyb
Combined
disregards
and benefit
levels
relative to
povertyc
Unrestricted
asset
levelsd
Restricted
asset
levelse
Vehicle
exemption
limits
Nevada
38.9%
31.1%
38.9%
$2,000
New Hampshire
98.3%
49.2%
116.8%
$1,000
No limit
New Jersey
75.8%
37.9%
85.3%
$2,000
$9,500
New Mexico
34.8%
43.7%
67.9%
$3,500
No limit
New York
95.4%
51.6%
118.6%
$2,000
No limit
$4,650
North Carolina
56.7%
24.3%
56.7%
$3,000
One vehicle per adult
North Dakota
80.6%
66.2%
93.0%
$5,000
Ohio
86.9%
32.4%
102.3%
No limit
Unlimited
Oklahoma
62.9%
26.1%
62.9%
$1,000
$2,000
$5,000
Oregon
49.2%
41.1%
64.6%
$2,500
No limit
$10,000
Pennsylvania
72.1%
36.0%
77.8%
$1,000
Rhode Island
114.3%
49.5%
114.3%
$1,000
$4,650
South Carolina
58.6%
18.0%
76.5%
$2,500
$10,000
South Dakota
56.0%
38.4%
56.1%
$2,000
$4,650
Tennessee
84.8%
16.5%
84.8%
$2,000
$5,000
$4,600
Texas
24.8%
16.8%
24.9%
$2,000
No limit
$4,650
Utah
59.7%
40.3%
59.8%
$2,000
$8,000
Vermont
87.5%
59.1%
87.5%
$1,000
90% of earnings
Virginia
103.4%
26.0%
129.4%
$1,000
$5,000
$7,500
Washington
48.8%
48.8%
73.2%
$1,000
No limit
$5,000
West Virginia
45.0%
27.1%
45.1%
$2,000
Wisconsin
119.0%
56.1%
175.1%
$2,500
$10,000
Wyoming
48.3%
30.4%
48.3%
$2,500
$12,000
State
Sources: Based on data from the Welfare Rules Database, the Urban Institute, July
2000. See also Rowe, G. State TANF policies as of July 1999. Assessing the New
Federalism: Welfare Rules Databook. Washington, DC: Urban Institute, 2000.
a Figures based on policies for earned income disregards after one year on welfare
to reflect the longer-run work incentive. (In many states the earned income disregards are more generous in the first through the fourth months on cash assis-
tance.) Amount of disregard expressed as a percentage relative to the poverty

line.
b Amount of benefit expressed as a percentage relative to the poverty line.
c Maximum earnings plus benefits that could be received in 1999.
d Unrestricted assets may be used for any purpose.
e Restricted assets may be set aside for particular purposes, such as home ownership, education, and business capitalization.
143
Zedlewski
Appendix 2
Innovative Uses of TANF Funds for Programs
to Support Working Families
Program name:
First Wheels Program
State:
Tennessee
Coverage:
Statewide
Program goals:
Enable low-income parents to get to work
Service population: Participants in Families First, Tennessees

TANF program
Funding sources:
TANF
Description: This program addresses the mismatch between where

most low-income people live and where most jobs are located. In
addition to offering grants of up to $6 a day to participants in Families First for their transportation needs, the state established First
Wheels, a program that offers low-income families no-interest
loans to purchase cars. Participants repay the loans based on their
income level, and repaid funds are re-loaned to others. The program
shields participants from damage to their credit history. Also, the
cars are inspected and approved before purchase. Both the onetime grants and the First Wheels program were designed and implemented through the combined efforts of a coalition of low-income
and welfare recipient advocates working with the Tennessee
Department of Human Services.
Results: Since the First Wheels program began in 1999, 168 loans
have been made, and the payback rate has been over 98%. No evaluation has been conducted for this program.
For further information: See the Tennessee Department of Human
Services Web site at http://www.state.tn.us/humanserv/.
144
Volume 12, Number 1
Program name:
Work First Housing Assistance Program
Program name:
School Clothing Allowance
State:
New Jersey
State:
West Virginia
Coverage:
Asbury Park, Camden, Elizabeth,

New Brunswick, and Trenton
Coverage:
Statewide
Program goals:
Program goals:
Increase residential stability for lowincome families
Allow poor parents to purchase new clothing

for their children
Service population: Families leaving TANF for paid

employment
Service population: Families at or below the federal poverty

guidelines and with less than $2,000 in
assets
Funding sources:
Funding sources:
TANF
TANF
Description: This program offers rent subsidies to families for three

years after leaving TANF. In the first year, 55% of the rent amount is
subsidized; in the second year, 45%; and in the third year, 35%. The
amount of money by which rent subsidies decrease from one year
to the next is used to create an interest-bearing savings account for
participating families. The program also offers services to enable
low-income families to purchase a home by providing counseling to
potential homeowners and mortgages with no down payments.
Description: West Virginias clothing allowance program provides

poor families with vouchers of $150 per child for the purchase of new
clothes each fall. The program has been in operation since the 1970s,
but prior to 1998 it was limited to families receiving cash assistance.
Beginning in 1998, however, TANF funds were used to greatly expand
the program so that it now reaches a much larger number of poor
families. Because the School Clothing Allowance does not count as
TANF assistance, these benefits are not time-limited.
Results: Since 1998, the program has provided secure housing for
350 families transitioning from welfare to work. As of October
2001, about half of the families had left the program, moving either
into a home-ownership program sponsored by the Department of
Housing and Urban Development or into market-rate housing.
Results: In 2000, the School Clothing Allowance program helped

nearly 55,000 poor children, mostly from working poor families.
About 75% of those served were not receiving TANF cash assistance.
For further information: See the West Virginia Department of Health
and Human Resources Web site at http://www.wvdhhr.org/ofs/.
For further information: See the New Jersey Department of Human

Services Web site at http://www.state.nj.us/humanservices/
dfd/post-tanf.html.
145
Welfare Reform, Fertility,

and Father Involvement
Sara S. McLanahan
Marcia J. Carlson
SUMMARY
Recognizing that most poor families are singleparent families, the federal welfare reform law of
1996 emphasized the responsibility of both parents to support their children. In addition to
strengthening the child support enforcement
system, the law included several provisions
designed to decrease childbearing outside of
marriage and to promote two-parent families.
This article focuses on the important role that
fathers play in childrens lives and how public
policies have affected childbearing and father
involvement. Key observations are:
Compared with children living with both biological parents, children in father-absent families
often have fewer economic and socioemotional
resources from their parents, and do not fare as
well on many outcome measures.
Efforts to reduce the rising number of fatherabsent families by focusing on preventing
unwanted pregnancy among unmarried
women, especially teen girls, have met with
some success; those programs seeking to alter
adolescents life opportunities in addition to
providing education or family planning services appear to hold the most promise.
Efforts to encourage greater father involve-
ment by focusing almost exclusively on

increasing absent parents child support payments reap only minimal benefits for poor
children because their absent parents often
have few resources and little incentive to make
support payments.
To date, efforts to increase the emotional
involvement of unmarried fathers with their
children have produced disappointing results,
but new research suggests that such programs
can make a difference when targeting fathers
at the time of a childs birth.
Many children spend some time living away from
their fathers, deprived of the financial and emotional resources they can provide. Because of the
importance of fathers to child well-being, the
authors conclude that new directions in research
and public policies are needed to encourage
greater father involvement across the wide diversity of family arrangements in society today.
Sara S. McLanahan, Ph.D., is professor of sociology
and public affairs and director of the BendheimThoman Center for Research on Child Wellbeing at
Princeton University in Princeton, NJ.
Marcia J.Carlson, Ph.D., is an assistant professor of
social work at Columbia University in New York, NY.
147
McLanahan and Carlson
he Personal Responsibility and Work

Opportunity Reconciliation Act of 1996
represented a historic shift in U.S. policy
toward poor families and children.1 In addition to requiring that low-income parents assume
greater responsibility for their own economic wellbeing through increased work, the reform legislation
included provisions to discourage births outside of
marriage, to promote and strengthen two-parent families, and to encourage father involvement (at least with
respect to financial support). These provisions reflect
and contribute toa growing awareness of the importance of fathers for children.
Until recently, discussions about welfare policy have
largely excluded fathers, except with respect to their
frequent failure to pay child support. Despite rising
concerns since the 1980s about the negative consequences of out-of-wedlock childbearing and singleparenthood (particularly for children, but also for
society), most policy and research about families on
welfare have focused only on single mothers. However, recent research on fatherhood has pointed to the
range of contributions that fathers can make in their
childrens lives,2 as well as to the barriers that some
fathers face in providing economic and emotional support for their children.
This article draws on recent research to examine the
role of fathers in childrens lives and how welfare policy may affect father involvement. The first section
reviews demographic trends affecting low-income families and outlines the evidence concerning the effects of
father involvement on fertility and child well-being.
Policies aimed at decreasing nonmarital fertility and
increasing father involvement are described, along with
suggestions for ways that programs can better address
the needs of disadvantaged fathers and families to promote child well-being.
Recent Trends and Effects on Children

Several major demographic trends in the latter half of
the twentieth century have affected the composition of
families in the United States, especially low-income
families. In particular, declining marriage rates, increasing divorce rates, and increasing rates of births to
unmarried women (see Figure 1) have combined to
148
increase the likelihood that children will spend time living away from their fathers. Although many unmarried
parents work together to raise their children by cohabiting or maintaining frequent contact, father involvement for most low-income families in this situation is
not necessarily stable.
Rising Number of Nontraditional Families

Fewer children today spend their entire childhood in
homes with two married parents than children did in
the middle of the twentieth century. A major trend contributing to changes in family composition is the overall decline in marriage rates. This decline is driven by a
combination of factors: people waiting longer to get
married the first time; not marrying at all; or not remarrying after divorce. Between 1960 and 1998, the median age at marriage for women rose from 20.3 to 25.0
years, and for men, from 22.8 to 26.7 years.3 Meanwhile, the rate of marriage among unmarried women
declined from 73.5 per 1,000 in 1960 to 49.7 in 1996.4
At the same time, the decline in the marriage rate has
been largely offset by a dramatic rise in nonmarital cohabitation that has blurred the boundaries of marriage.
Although the mean age at marriage has risen, when
cohabitation is considered along with marriage, the average age at union formation (either marriage or cohabitation) has remained relatively constant.5 In 1995, 49% of
women ages 30 to 34 had cohabited at some time in their
lives, and the proportion of persons entering first marriages who had previously cohabited was 53%.6 For some
couples, cohabitation may serve as a precursor to marriage, whereas for othersparticularly for those with low
educational attainment and earnings, who are at greatest
risk of receiving welfarecohabitation is more likely to
serve as a substitute for marriage.7,8
The second major trend affecting American families is
the increase in marital instability. Divorce rates more
than doubled during the last half of the twentieth century, increasing from 9.2 to 22.8 divorces per 1,000
married women (age 15 years and older) between
1960 and 1979.4 Over half of all marriages begun in
the 1980s were projected to end in divorce.9 Since
1980, the divorce rate has leveled off, but at a level
much higher than during (and before) the 1960s (see
Figure 1). As of 1996, the divorce rate was still 19.5
divorces per 1,000 married women.
Volume 12, Number 1
Reform, Fertility, and Father Involvement
Figure 1
Marriage, Divorce, and Nonmarital Birth Rates
90
80
60
50
40
30
20
0
1960
10
Note: Nonmarital birth rate is per 1,000 unmarried women ages 15 to 44, marriage rate is per
1,000 unmarried women age 15 and over, and
divorce rate is per 1,000 married women age
15 and over.
70
Number per 1,000
Sources: Clarke, S.C. Advance report of final

divorce statistics, 1989 and 1990. Monthly Vital
Statistics Report. Vol. 43, No. 9(S). Hyattsville,
MD: National Center for Health Statistics, March
22, 1995, p. 9, table 1; Clarke, S.C. Advance
report of final marriage statistics, 1989 and
1990. Monthly Vital Statistics Report. Vol. 43,
No. 12(S). Hyattsville, MD: National Center for
Health Statistics, July 14, 1995, p. 7, table 1;
Ventura, S.J., and Bachrach, C.A. Nonmarital
childbearing in the United States, 194099.
National Vital Statistics Report. Vol. 48, No. 16.
Hyattsville, MD: National Center for Health Statistics, October 18, 2000, p. 17, table 1; U.S.
Census Bureau. Statistical abstract of the United States, 2000: The national data book. Washington, DC: U.S. Census Bureau, 2000, table
144.
1965
1970
1975
1980
1985
1990
1995
2000
Year
KEY:
Marriage rate
The third trend is the increased proportion of births

occurring outside of marriage. Between 1960 and
1999, nonmarital births as a proportion of all births
rose from 5% to 33%.10 This trend is driven by the delay
in marriage, the decline in births among married couples relative to births among unmarried couples, and
the increase in childbearing among unmarried women.
The birth rate among unmarried women rose from
21.6 to 46.9 per 1,000 between 1960 and 1994, and
has since declined slightly. Although nonmarital birth
rates for young women ages 15 to 19 also have risen
steadily since 1940, as a proportion of all nonmarital
births, teen births have actually declined. They
accounted for 50% of all unmarried births in 1970, but
only 31% in 1997 (due in part to the rising average age
of the never-married population).11
Taken together, these demographic trends have
increased the likelihood that, before reaching adult-
Nonmarital birth rate
Divorce rate
hood, children will live apart from at least one of their

biological parents (usually the father). This could occur
within multiple family contexts: The child could live
alone with the mother, with the mother and her new
cohabiting partner, or with the mother and a stepfather.12 Single-parent families (defined as unmarried
mothers or fathers and their children living alone or
with the parents cohabiting partner) represented 9% of
all families with children in 1960, and 27% in 1998.13
Of children born in the 1980s, it is estimated that
about half will spend some time in a single-parent family before they reach age 18.14
The overall rise in the number of single-parent families
concerns researchers and policymakers alike, particularly because of the lower parental and economic
resources generally found in such families. Yet recent
research has highlighted the fact that single-parent
families represent neither a homogenous nor a static
149
group. Traditional family formation typically followed a linear course: dating followed by marriage,
sexual activity, and then childbearing. But today,
cohabitation, intercourse, and parenthood all occur
rather frequently outside of marriage, yielding a range
of complex and diverse family arrangements.
The composition of single-parent families has changed
dramatically in the past two decades. In 1976, only 17%
of single mothers were never married, but by 1997, the
proportion of never-married single mothers had grown
to 46%.15 In addition, rising cohabitation rates portend
that living arrangements may serve as a more important
criterion than marital status for determining family
structure.14 According to one estimate, 15% of families
classified as single-mother families by marital status in
1987 included a cohabiting male.16 Also, 41% of all
nonmarital births in the early 1990s occurred to cohabiting couples.6 Such statistics imply that many single
mothers are not rearing their children alone and that
the family history and current circumstances of unmarried mothers and their children vary greatly, calling into
question the extent to which such families can be
appropriately characterized as father-absent.
In combination, these trends suggest the emergence of

a new family typethe fragile family, comprised of
unmarried parents who are working together to raise
their children either by cohabiting or maintaining frequent contact.17 Such families are deemed fragile
because of the multiple risks associated with nonmarital childbearing, including poverty, and to signify the
vulnerability of the parents relationship. Union dissolution rates are much higher among cohabiting couples
than among married couples; this is true particularly in
the United States, but also in Western European countries, where cohabitation is even more prevalent.18
Family Structure and Child Well-Being

Growing family diversity has raised concerns about
child well-being because not all family experiences are
equally beneficial for children. A multitude of studies
have documented that children in single-parent families do not fare as well on a range of outcome measures
as children living with both biological parents.19
Although the mechanisms by which family structure
affects childrens well-being have not been fully determined, it is clear that children in female-headed families are often deprived of two types of resources that a
Box 1
The Fragile Families and Child Wellbeing Study
The Fragile Families and Child Wellbeing Study was designed by
researchers at Princeton University and Columbia University using
an innovative, integrated framework to provide information about
three areas of great interest to policymakers and community leaders: nonmarital childbearing, welfare reform, and the role of
fathers. The study follows a birth cohort of approximately 3,700
children born to unmarried parents in 20 U.S. cities with populations over 200,000. New mothers are interviewed in person at the
hospital within 48 hours of giving birth, and fathers are interviewed in person either at the hospital or as soon as possible
thereafter. Follow-up interviews will be conducted when the child
is 12, 30, and 48 months old. The study is representative of nonmarital births within each city, and the full sample is representative of all nonmarital births to parents residing in large cities
nationwide. Also, married parents (a total of about 1,200 couples)
are being interviewed in each of the cities for comparison. All
income groups are included, but many unmarried parents are lowincome. Baseline interviews in all 20 cities were completed in the
fall of 2000 and show that 41% of the unmarried mothers in the
study had incomes below poverty level, and 39% were receiving
welfare. The 12-month follow-up survey was expected to be completed by the end of 2001.
For more information, see the Fragile Families and Child Wellbeing Study
Web site at http://crcw.princeton.edu/fragilefamilies.
150
Volume 12, Number 1
Children in female-headed families are often deprived of two

types of resources that a father might provideeconomic
and socioemotional.
father might provideeconomic and socioemotional.
The economic consequences can be quantified most
easily: Families with no male present are much more
likely to be poor, with adverse effects on child development and well-being.20 In 1998, for example, the
poverty rate for female-headed families with children
was 39.2%, compared with 7.8% for male-present families with children.21 Children in female-headed families
are also disadvantaged because they are less likely to
have highly involved fathers who provide attention and
emotional support. Nonresident fathers see their children less often than residential fathers do, and lack of
interaction makes it less likely that a father and child
will develop a close relationship.22,23 As described further in the next section, a fathers participation in his
childs lifeboth financially and emotionallyoffers
important benefits.
Academic publications and the popular press have both
focused significant attention on the causes and consequences of the rise in single-parent families and the
need to collect child support from noncustodial
fathers. Spurred by a growing awareness of single-parenthood, in the 1970s and 1980s scholars began to
estimate the income and capabilities of noncustodial
fathers to pay child support. These analyses highlighted the dearth of firsthand information about lowincome men and their families, and showed that
national survey datathe resource social scientists use
most frequently to study individuals and familiesseriously underrepresented fathers who do not live with
their children.24 As a result, information on the characteristics of fragile families is relatively new and just
beginning to be described in the research literature.
Building on the early studies of noncustodial fathers
earnings and relationships, the Fragile Families and
Child Wellbeing Study was launched in January 1997
to investigate the conditions and capabilities of new
unmarried parents and the consequences for their children (see Box 1).
Data from the baseline interviews in the Fragile Families Study reveal two key findings about unmarried par-
ents with a newborn child. First, most unmarried couples are closely connected to each otherand invested
in their new babyaround the time of the childs
birth. More than 80% of mothers reported that they
were romantically involved with the babys father at the
time the baby was born (see Figure 2). Most fathers
were involved during the pregnancy and around the
time of birth: More than three-fourths helped the
mother during the pregnancy, and/or visited the
mother in the hospital. Also, nearly all (99%) of the
fathers in the study expressed a desire to be involved in
raising their child(ren), and 93% of mothers said that
they want the father to be involved. Even among the
mothers who are not romantically involved with the
father at the time of birth, fully two-thirds indicated
that they want the father involved in raising their child.
Figure 2
Romantic Involvement of Unmarried Parents
Little or no contact (9.2%)
Cohabiting (49.7%)
Friends (7.8%)
Visiting (33.3%)
Source: Data tabulated from the Fragile Families and Child Wellbeing Study, based
on a sample of 3,712 unmarried mothers at the time of a new childs birth.
151
Families Study, but other research has shown that

fathers who do not live with their children and who
never married the childs mother have lower involvement than divorced fathers, and that fatherchild contact typically diminishes with time after unmarried
parents separate.23 For these reasons, children in fragile
families whose parents separate are likely to be at higher risk of growing apart from their fathers over time.
Stone/Nick Vedros & Assoc.
Fathers as Resources for Children
Although new parents may have positive expectations

for their future, another major finding from the Fragile Families Study is that many parents lack the skills
and capabilities necessary for a stable family life. Some
37% of mothers and 34% of fathers in the study lacked
a high school degree, and only 4% of mothers and
fathers had a college degree. During the week before
the childs birth, 19% of fathers were idle (neither
working nor in school), and 39% of mothers had
received income from welfare at some time in the past
year. These figures underscore the precarious socioeconomic circumstances and the barriers to long-term
family stability that many of these unmarried parents
face. Thus, though the birth of a child may represent a
magic moment of high attachment and expectations
for unmarried couples, many grapple with an uncertain
future. The fathers involvement may attenuate over
time, depriving their children of an important and irreplaceable resource. Information about fathers longterm involvement is not yet available from the Fragile
152
The consequences of not having a father have been a

source of long-standing concern to society, but the
focus of research on fathers has evolved as the larger
cultural meaning of fatherhood has changed over time.
Only in the last several decades have scholars begun to
examine father involvement more broadly. Early studies focused on the effects of father absence, defined as
the father not living with the child. In this deficit
model, children in mother-only (or father-absent)
families were compared to children in two-parent
(father-present) families without directly measuring
what fatherswhether living with their children or
notwere actually contributing to their childrens
lives. As noted earlier, research shows that children in
single-parent families experience greater adverse outcomes compared with their counterparts who live with
both biological parents; however, most of these studies
focus on children of divorced fathers only.19
In the 1980s, with the emergence of a new fatherhood model (particularly among the middle class) in
which there were greater expectations for fathers emotional investment and active participation in parenting,
studies began to investigate the potential positive
effects of father involvement. The first studies in this
area focused on fathers financial support and found
that the payment of child support is positively associated with childrens well-being.25 For example, one
Urban Institute study found that $1,000 in child support was associated with higher grades and fewer
school problems among children, and that child support income had a larger effect on childrens wellbeing than an identical amount of ordinary income.26
A growing literature in sociology and child development has investigated the effects of fathers nonmonetary involvement as well, such as participating in shared
Volume 12, Number 1
A recent meta-analysis ... found evidence that increased father

involvement produced significant gains in childrens academic
achievement and reductions in behavioral problems.
activities with the child and developing a close, highquality relationship with the child. Positive father
involvement, particularly by fathers who live with their
children, has been linked to less-frequent child and
adolescent behavioral problems, including delinquency, substance use, anxiety, and depression.27 However,
these effects vary in size and significance and are not
always large relative to other important influences on
childrens well-being.
The benefits for children of involvement by fathers
who are not living in the same household are even less
apparent, perhaps because the quality of father involvement has not been accurately measured.28 Most studies
of nonresident father involvement have focused on the
quantity (frequency) of fatherchild contact, and frequent interaction with fathers has been found to have
little if any beneficial effect for children.29 Yet the quality of the relationship inherently reflects the strength of
the affective bond between father and child, and feeling loved and cared for by parents increases childrens
emotional security, sense of worth, and self-confidence,
which facilitate positive development.28 Given the evidence in psychological research that fathers can positively affect child development,30 the lack of clear
findings in other social science literature underscores
the fact that the quantity of father involvement may be
far less important for a childs well-being than the quality of involvement.31 A recent meta-analysis that used
more nuanced measures of nonresidential father
involvementsuch as fatherchild closeness and
fathers authoritative parentingfound evidence that
increased father involvement produced significant
gains in childrens academic achievement and reductions in behavioral problems.28 Future research using
more refined measures of high-quality father involvement is likely to corroborate these results.
Not surprisingly, fathers who do not live with their
children see them less often, which decreases the likelihood that the father and child will develop a close relationship.23,32 Also, fathers who do not share the childs
household are less likely to contribute financial
resources to support their child, as they have less ability than a father living with the child has to monitor the
allocation of resources by the mother.33 Particularly following divorce, absent fathers may become less altruistic toward their children over time.34 Divorced parents
also may be less able to reinforce one another in child
rearing, further diminishing the fathers role.35
Although these findings refer to formerly married couples, the consequences are likely similar for unmarried
couples following a separation.
Despite the potential financial and emotional benefits of
father involvement in general, father involvement may,
in fact, be detrimental for children and their mothers in
those instances when the father is prone to violence or
has mental health or substance abuse problems.36
Recent research shows that approximately 15% of
women on welfare in one city reported being severely
physically abused by a husband or partner in the last
year,37 a rate that is comparable to rates reported in
other studies of welfare recipients.38 About 4% of new
mothers in the Fragile Families Study reported that the
father sometimes or often hit or slapped them
within the last month (or the last month they were
together, for couples no longer romantically involved).
This percentage is somewhat lower than those reported
in other studies of welfare mothers, but violence is likely lower in the month preceding a childs birth. Also,
the mothers in the Fragile Families sample are somewhat more economically advantaged than a sample of
mothers on welfare, and the risk of violence toward
women is lower among families with more income.39
Domestic violence is a very serious problem for the
children and mothers affected by it, and violent behavior may be underreported in surveys. Nevertheless,
most fathers are not violent or potentially dangerous,
and for most children, greater father involvement likely offers important benefits. The challenge is to devise
programs that encourage positive father involvement,
but that include adequate safeguards for the minority
of children and mothers who may be at risk.
153
Though pregnancy prevention programs have met with some

success, they have not fundamentally abated the high levels of
nonmarital fertility and the formation of father-absent families.
Policies Designed to Promote Father
Involvement
As family demographics and the social environment
have changed, public policy also has evolved in an
attempt to mitigate the consequences of family instability and, in some cases, to reshape the demographic
trends themselves. Most recently, the 1996 federal welfare reform law gave new emphasis to two primary categories of programmatic interventions intended to
promote father involvement: 1) programs designed to
discourage nonmarital fertility and thus decrease the
formation of father-absent families; and 2) programs
intended to increase nonresident fathers support for
and involvement with their children.
Programs to Discourage the Formation of FatherAbsent Families

Efforts to reduce the rising number of father-absent
families have focused primarily on preventing unwanted pregnancy among unmarried women, especially
teenage girls. This approach is guided by the awareness
that when a pregnancy is unintended, the father is less
likely to live with the child and provide positive parenting.40 In contrast, when a pregnancy is intended
and births are spaced appropriately, better maternal
and child health outcomes are likely, and assurance is
greater that the child will be loved and nurtured by
both the mother and the father. Most births to unmarried couples, however, are unintended. Of births
among never-married women in 1994, 58% were the
result of unintended pregnancies.41 Therefore, reducing the incidence of unintended pregnancy among
unmarried couples represents a promising strategy to
reduce the likelihood that a child will grow up without
a fathers involvement in his or her life. Pregnancy prevention efforts fall into three main categories: family
planning, teen programs, and family caps.
Family Planning
Federal family-planning funding (particularly through
Title X of the Public Health Services Act) has been
used to provide contraceptives to low-income women
154
in hospitals, community-based clinics and health centers, and private organizations such as Planned Parenthood. The impact of federal family-planning programs
has not been unambiguously documented, in part
because most studies are complicated by methodological problems such as failure to control for levels of sexual activity or to account for the increasing availability
of contraception (such as condoms) outside of familyplanning clinics.40 Yet there is some evidence that family-planning programs have reduced nonmarital
pregnancies and/or birth rates, and that they are costeffective because they reduce medical expenses.42
Teen Programs
Many efforts to reduce nonmarital childbearing have
focused on teenagers, even though births to women
under age 20 account for less than one-third of all
births outside of marriage.43 However, 35% of births to
unmarried women over age 20 were preceded by a
teenage birth, and teen births account for 49% of all
first births outside of marriage.44 Further, teen childbearing is associated with a high probability of receiving welfare, and the majority of mothers on welfare had
their first child as a teenager.45 Therefore, delaying adolescent pregnancy is an important strategy for improving outcomes for children.
Pregnancy prevention programs for teens have disparate goalssome aim to reduce sexual activity altogether by promoting abstinence, whereas others
encourage safe sex by increasing sex education and
availability of contraceptives. In addition, a growing
number of programs are targeting males with messages
about abstinence and statutory rape. Although rigorous research has not yet proven the effectiveness of
either abstinence or education programs,46 teen pregnancy and birth rates declined in the 1990s (see Figure
3), and both abstinence and contraception appear to
have contributed to the declines.47 Interestingly, programs that seek to alter adolescents life opportunities,
such as early childhood education and youth development programs, appear to hold greater promise than
education or service programs alone.48
Volume 12, Number 1
Figure 3
Teen Pregnancy and Birth Rates 19761997
120
Birth rate
Sources: Ventura, S.J., Mosher, W.D., Curtin,

S.C., et al. Trends in pregnancies and pregnancy rates by outcome: Estimates for the United
States, 197696. Vital Health Statistics Report,
series 21, no. 56. Hyattsville, MD: National Center for Health Statistics, January 2000, p.
2829, table 3; Ventura, S.J., Mosher, W.D.,
Curtin, S.C., et al. Trends in pregnancy rates for
the United States, 197697: An update. Vital
Health Statistics Report, Vol. 49, No. 4.
Hyattsville, MD: National Center for Health Statistics, June 6, 2001, p. 5, table 1.
100
80
60
40
20
1976 _
_
1978 _
_
_
1980
_
_
1982
_
_
1984
_
_
1986
_
_
1988
_
_
1990
_
_
1992
_
_
1994
_
_
1996
_
_
1998
Pregnancy rate
Number per 1,000 women ages 15 to 19
KEY:
Family Caps
The family cap (or child exclusion) policy limits the
monthly welfare benefit a mother can receive regardless of whether she has additional children, based on
the theory that more generous welfare benefits are likely to increase illegitimacy and reduce incentives for
marriage.33,34 In the 1970s and 1980s, researchers
began to examine whether welfare benefits were, in
fact, largely responsible for the rise in births outside of
marriage. Charles Murrays Losing Ground, published
in 1984, was the most notable treatise in support of
this theory.49 However, empirical research suggests that
welfares effects on marriage and fertility are relatively
small compared with other factors affecting nonmarital
childbearing.50 Nonetheless, those who believe welfare
provides incentives for nonmarital childbearing expected that family caps would decrease childbearing among
welfare mothers, and hence reduce the number of children with absent fathers.
Family caps were first implemented through waivers
granted to states in the early 1990s, and the federal
Year
welfare reform law of 1996 permitted all states to

impose a family cap without federal approval. New Jersey was the first state to adopt a family cap, in 1992,
and as of 1999, 23 states had established some sort of
family cap policy.51 An evaluation of the New Jersey
program in 1998 used an experimental designthe
most rigorous test of program effectsand showed
that births outside of marriage were significantly lower
among the experimental group compared to the control group, but only among new welfare recipients.
The decline in birth rate was accompanied by an initial
rise in abortions that subsequently dissipated.52 However, in Arkansasthe only other state to evaluate its
family cap program using an experimental designno
statistically significant impact of the family cap was
noted for nonmarital births.53
Overall, though pregnancy prevention programs have
met with some success, they have not fundamentally
abated the high levels of nonmarital fertility and the
formation of father-absent families. This is because
nonmarital fertility has risen for reasons that reflect
155
Rising paternity establishment rates, child support orders,

and collections signal that the system is increasingly successful
... but for children on welfare, the benefits of an effective CSE
system are minimal.
larger cultural shifts in attitudes, values, and practices
not simply because couples lack information about sex
or access to contraceptive technology. As described in
the section on demographic trends, consensual unions
other than marriage have become more accepted and
prevalent, increasing the likelihood that children will
be born outside of marriage.
Because father absence is the defining characteristic
of most single-parent families, public policy has
attempted to compensate for the resources that children lose when the father is not in the household. The
most obvious resource deficit is economicwithout
fathers income, female-headed families are much more
likely to be poor. Initially, policies were designed to
compensate for the loss of the fathers income directly
with cash assistance and in-kind benefits such as food
stamps and Medicaid. Then, as single mothers increasingly were women who were separated and divorced
from their partners, as opposed to being widowed, policymakers began to consider seeking resources from
fathers. Programs were initiated to collect child support from unmarried fathers and, more recently, to
increase their earnings so that they can pay child support. Only recently has policy attention to fathers
broadened beyond financial support to incorporate
nonmonetary investments in children as well.
Also from the beginning, the CSE system focused on

never-married fathers because these men are the most
likely to have children on welfare.54 Legislative reforms
in the last two decades, and most recently the federal
welfare reform law of 1996, have strengthened CSEs
overall effectiveness (see Figure 4). During the past
two decades, paternity establishmenta prerequisite
to the formal establishment of child supporthas
increased dramatically from 19% of nonmarital births
in 1979 to 52% in 1996. In addition, the use of
increasingly aggressive enforcement tools, such as universal wage withholding, revocation of drivers, recreational, and professional licenses, and interception of
tax refunds, has increased payments dramatically. Child
support collections facilitated by CSE more than doubled (in nominal dollars) between 1990 and 1998,
from $6.0 billion to $14.3 billion.55 Still, it should be
noted that child support is collected for only a relatively small percentage of families served by the CSE system. Collections were made on behalf of only 14% of
families on welfare (now Temporary Assistance for
Needy Families, or TANF) and 28% of nonwelfare
families during 1998.56

The first program that could be considered a fatherhood program is the federal Child Support Enforcement (CSE) system. Beginning in 1975, federal
matching funds were provided to states to establish
paternity and support awards and to collect child support payments on behalf of single-parent families. The
program was initially designed to recoup money from
noncustodial fathers to offset expenditures in the Aid
to Families with Dependent Children (AFDC) program, but in 1980, CSE was broadened to serve all
children eligible for support regardless of family
income or welfare status. From its inception, the CSE
Policymakers have largely assumed that fathers fail to

pay child support because they do not want to pay (socalled deadbeat dads), not because they are unable
to pay.57 But research indicates that, although most
noncustodial fathers can afford to pay more child support, a sizeable number of fathers are poor and unable
to support their children.58 These are typically the
fathers of children on welfare, and little is known about
these men beyond the fact that they are often unemployed or underemployed and have few resources.59
Clearly, the situations of fathers living apart from their
children vary widely. Recent analyses have found that
as a group, however, nonresident fathers are more like-
Programs to Encourage Greater Father Involvement
156
program was charged solely with enforcing fathers

financial support of their children, while other aspects
of fathers involvement, including visitation and custody, were (and remain) governed by state laws.
Volume 12, Number 1
Figure 4
Increased Child Support Enforcement Efforts (19801998)
3500
KEY:
3000
Support orders
established
2500
AFDC/TANF cases
with collections
Non-AFDC/TANF
cases with collections
Thousands
Paternities established
2000
1500
1000
500
Source: U.S. Census Bureau. Statistical

abstract of the United States, 2000: The
national data book. Washington, DC: U.S.
Census Bureau, 2000, table 632.
0
1980
ly than resident fathers to be young, to have less than

a high school education, to be in poor health, to have
had some involvement with the criminal justice system,
and to have lower hourly wages and fewer work hours
per week.60
Rising paternity establishment rates, child support
orders, and collections signal that the CSE system is
increasingly successful at its primary objectiveensuring that noncustodial fathers provide economic support to their children. But for children on welfare, the
benefits of an effective CSE system are minimal, for
several reasons. First, for the most part, child support
collected on behalf of welfare families goes to repay
welfare expenditures and not to increase family
income.61 (See the article by Greenberg and colleagues
in this journal issue.) Because of this, fathers have little
incentive to pay their obligations, as their children are
not economically better off as a result.54 In addition, to
receive welfare benefits, mothers must provide information to help locate the father, which can lead to conflict and tension between parents. Further, many states
1985
1990
Year
1995
2000
set minimum baseline amounts for child support

orders on the assumption that fathers work full time,
regardless of actual employment status. As a result,
low-income fathers may be forced to pay a much higher proportion of their income in child support than
middle-income fathers pay.62 Finally, child support
orders are not routinely adjusted for changes in the
fathers income, which can lead to fathers accumulating large arrearages that, according to federal law, cannot be forgiven or adjusted in most cases.63
In addition to increasing fathers financial contributions, child support enforcement policy may affect family formation and how fathers relate to their children.
By increasing the costs of living apart from their children, strong child support enforcement may encourage
fathers to avoid having children outside of marriage.
Studies suggest that strong child support enforcement
may be linked to reduced nonmarital childbearing64
and reduced likelihood of marital dissolution.65 Also,
some evidence indicates that requiring fathers to pay
child support increases their involvement with their
157
EyeWire Collection
children.66 At the same time, if CSE programs are successful in stimulating fathers support and involvement to an extent that offsets the detriments of single
parenthood, the incentives increase for mothers to stay
single rather than marry.
Noncustodial Parent Work Programs
In view of the fact that some low-income fathers are
not able to meet their child support obligations, several demonstration projects were undertaken in the
1980s and 1990s to improve fathers labor market outcomes. The primary demonstration in this area was
Parents Fair Share (PFS), a program administered in
seven sites around the country to increase low-income
noncustodial parents employment, earnings, and ability to pay child support.59 The PFS program enrolled
fathers who were unemployed or in a low-wage job
and had fallen behind in their child support payments.
Most fathers in the program had been divorced and
were disconnected from their children.
An evaluation of the PFS program revealed the difficulty and complexity of improving labor market outcomes for low-income men, and the fact that child
158
support and welfare programs are not equipped to

meet the needs of poor fathers. The program relied on
judges to order fathers to enroll in PFS or go to jail;
thus, these men were not necessarily motivated to participate because of their desire to be better fathers.
Nonetheless, although their relationships with their
children varied dramatically, most fathers expressed a
deep desire to be involved in their childrens lives.67
Although PFS did not, on average, increase the frequency of noncustodial fathers visits with their children, some increases were noted in sites with
particularly low initial levels of fatherchild contact.
Similarly, the frequency of contact between fathers and
the custodial mothers varied widely, with nearly threequarters of mothers reporting that the father had no
involvement in decisions about the child. Again, the
PFS program was found to have no effect on the frequency of parents interactions; a slight increase was
noted, however, in the proportion of parents who
reported frequent conflict. Although these findings are
discouraging, the implementation and evaluation of the
Parents Fair Share program have highlighted the mul-
Volume 12, Number 1
Fatherhood programs that begin earlyin the hospital, if

possibleare more likely to be successful than programs that
target fathers after the relationship with the mother has ended.
tiple challenges of supporting low-income fathers and
families, and the need to develop new program models.
In 1997, Congress created the Welfare-to-Work program to support state and community efforts to help
welfare recipients and noncustodial parents move into
unsubsidized jobs.68 Nearly $2 billion in grant funds
were awarded in 1998 and 1999 under this program to
assist hard-to-employ welfare recipients and noncustodial fathers who are unemployed, underemployed, or
having difficulty making child support payments.69 An
interim report from the contracted evaluator for this
new federal program noted that Welfare-to-Work initiatives are seriously attempting to reach and serve
noncustodial parents, and that the state and local staff
working on these projects say the grants have encouraged a more serious focus on fathers.70
New Fatherhood Programs
Until very recently, poor noncustodial fathers of children on welfare were largely ignored by social policymakers and disconnected from resources that might
help them become more involved in their childrens
lives.59 The child support system has operated solely as
an enforcement agency collecting money from fathers
(and punishing those who fail to pay) rather than as a
social service organization attempting to balance
responsibility with appropriate services and supports
(and providing incentives to pay). This is changing as
the confluence of three factorsdemographic changes
that have increased the number of fragile families;
growing awareness of the difficulties faced by lowincome fathers and families; and greater understanding
of the benefits to children of father involvementhas
led to the development of programs that more effectively promote fathers financial and emotional involvement with their children.
Representing an important first step toward developing such programs, in March 2000 the U.S. Department of Health and Human Services approved 10 state
demonstration projects to improve the opportunities
of young, unmarried fathers to support their children
both financially and emotionally.71 These new pro-
grams serve both divorced fathers and new fathers in

fragile families. They have varied emphases, but they
generally are designed to improve fathers parenting
skills and employment capabilities, and to ensure that
fathers have access to their children. Initial assessments
of these new programs have found that enrolling
fathers and sustaining their participation over time
present particular challenges. More rigorous evaluations have yet to determine the nature and magnitude
of the impacts across various program types.
Meanwhile, data from the Fragile Families Study support the current direction of public policy, but highlight the importance of careful program design and
implementation. Fatherhood programs are more likely
to make a difference if they are targeted to the right
men and if they are timed correctly. Practitioners who
run employment programs for disadvantaged men say
that participants motivation has an important effect on
whether the program will work or not. Fathers are
most likely to be highly motivated and to take advantage of the services offered by fatherhood programs
around the time of a new childs birth, when the father
is likely to be romantically involved with the babys
mother. Thus, fatherhood programs that begin early
in the hospital, if possibleare more likely to be successful than programs that target fathers after the
relationship with the mother has ended. Also, programs that treat fathers not only as individuals (recognizing their personal strengths, limitations, and needs)
but also as part of families (recognizing their familial
commitments, responsibilities, and supports) have a
greater chance at success. Finally, programs that
address the multiple needs of both mothers and fathers
across multiple family circumstances (married or
unmarried, living together or living apart) hold the
greatest promise. Such needs might include expanding
labor market skills and capabilities, developing parenting and relationship skills, and overcoming violence,
substance abuse, or mental health problems. (See the
appendix following this article for three examples of
programs using TANF funds to focus on teen males
and fathers.)
159
Conclusion
Public policy, supported by sound research, can

improve the likelihood that fathers will be involved with
their children, both by discouraging the formation of
father-absent families in the first place, and by increasing incentives and supports for positive father involvement. For example, programs designed to reduce the
rising number of father-absent families by focusing on
preventing unwanted pregnancies, especially among
teens, appear to be most successful when they seek to
alter adolescents life opportunities in addition to providing family planning education or services. Also,
although early efforts to encourage father involvement
yielded disappointing results, newer programs that are
better targeted and timed to the birth of a child appear
to hold greater promise for improving the circumstances of low-income fathers and families.
This article has highlighted the changing composition

of families in the United States, particularly the fact
that many children will spend some time living away
from their father during childhood. Because fathers
offer important financial and emotional resources to
children, it is important to encourage greater father
involvement, especially among fathers who do not live
with their children. Recent trends and concern for such
children have stimulated a variety of new public policies
and programs to promote fathers involvement with
their children, both financially and emotionally.
The authors wish to thank Margie K. Shields (issue

editor), Kristin A. Moore (issue editorial advisor),
and Sandra Hofferth for very helpful comments, and
Cary Bodenheimer and Regina Leidy for assistance
with manuscript preparation.
In addition, failure to examine a wide diversity of family arrangements undermines the capacity of research
to contribute to public policy. Despite a burgeoning
literature on the effects of father involvement for children generally, the fathers of children born outside of
marriage (particularly coresident but unmarried
fathers) have been largely unstudied. Most of the
research on fathers living apart from their children has
focused on divorced fathers (who are often middle
class) and whether or how they remain involved in
their childrens lives after the divorce. Given the growing diversity of family composition, this represents a
striking shortcoming in the literature. New directions
in research and public policies are needed to encourage
greater father involvement across the wide diversity of
family arrangements in society today.
160
Volume 12, Number 1
ENDNOTES
1. In the 1980s and early 1990s, states experimented (through waivers
and demonstrations) with many welfare and child support provisions
that were subsequently enacted as part of the Personal Responsibility
and Work Opportunity Reconciliation Act of 1996.
2. See, for example, Federal Interagency Forum on Child and Family
Statistics. Nurturing fatherhood: Improving data and research on male
fertility, family formation and fatherhood. Washington, DC: Federal
Interagency Forum on Child and Family Statistics, 1998; Lamb, M.,
ed. The role of the father in child development. New York: John Wiley
& Sons, 1997; Marsiglio, W. Fathers diverse life course patterns and
roles. In Fatherhood: Contemporary theory, research, and social policy.
W. Marsiglio, ed. Thousand Oaks, CA: Sage Publications, 1995,
pp. 78101.
3. U.S. Bureau of the Census. Marital status and living arrangements:
March 1998 (update). Current Population Reports, Series P20-514.
Washington, DC: U.S. Bureau of the Census, January 7, 1999, and
earlier reports. Available online at http://www.census.gov/population/
www/socdemo/ms-la.html.
4. Clarke, S.C. Advance report of final marriage statistics, 1989 and
1990. Monthly Vital Statistics Reports. Vol. 43, no. 12S. Hyattsville,
MD: National Center for Health Statistics, Center for Disease Control and Prevention, 1995; U.S. Census Bureau. Statistical abstract of
the United States, 2000: The national data book. Washington, DC:
U.S. Census Bureau, 2000.
5. Bumpass, L., Cherlin, A., and Sweet, J. The role of cohabitation in
declining rates of marriage. Journal of Marriage and the Family
(1991) 53:91327; Manning, W.D., and Smock, P.J. Why marry?
Race and the transition to marriage among cohabitors. Demography
(1995) 32(4):50920.
6. Bumpass, L., and Lu, H.H. Trends in cohabitation and implications
for childrens family contexts. Population Studies (2000) 54:2941.
7. Manning, W.D., and Lichter, D.T. Parental cohabitation and childrens economic well-being. Journal of Marriage and the Family
(1996) 58:9981010; McLanahan, S., and Casper, L. Growing diversity and inequality in the American family. In State of the union:
America in the 1990s. Vol. 2. R. Farley, ed. New York: Russell Sage
Foundation, 1995.
8. Some argue that because the new rules under welfare reform increase
a couples economic security without penalizing cohabitation, welfare
is encouraging family formation but not marriage itself. See Mincy,
R.B., and Dupree, A. T. Welfare, child support and family formation.
Children and Youth Services Review (2001) 23(6/7):577601.
9. Castro-Martin, T., and Bumpass, L. Recent trends and differentials in
marital disruption. Demography (1989) 25(1):3751.
10. Ventura, S.J., and Bachrach, C.A. Nonmarital childbearing in the
United States, 1940-99. National Vital Statistics Reports. Vol. 48, no.
16 (revised). Hyattsville, MD: National Center for Health Statistics,
October 18, 2000.
11. Also, only half of all nonmarital births are now first births. Ventura, S.,
Bachrach, C., Hill, L., et al. The demography of out-of-wedlock
childbearing. In Report to Congress on out-of-wedlock childbearing.
Hyattsville, MD: U.S. Government Printing Office, 1995, pp. 3134;
Ventura, S.J., Martin, J.A., Curtin, S.C., and Mathews, T.J. Births:
Final data for 1997. National Vital Statistics Reports. Vol. 47, no. 18.
Hyattsville, MD: National Center for Health Statistics, 1999.
12. Children also could live with their fathers in comparable situations.
The number of single-father families has risen in recent years,

although single-mother families still represent four-fifths of all oneparent families with children (as of 1998). See U.S. Bureau of the
Census. Household and family characteristics: March 1998 (update).
Current Population Reports, P20-515. Washington, DC: U.S. Bureau
of the Census, December 11, 1998, and earlier reports. Available
online at http://www.census.gov/population/www/socdemo/
hh-fam.html.
13. See note 12, U.S. Bureau of the Census.
14. Bumpass, L., and Raley, K. Redefining single-parent families: Cohabitation and changing family reality. Demography (1995) 32(1):97109;
and see note 9, Castro-Martin and Bumpass.
15. Sorensen, E., and Halpern, A. Child support enforcement is working
better than we think. New federalism: Issues and options for states, no.
A-31. Washington, DC: Urban Institute, March 1999.
16. Folk, K.F. Single mothers in various living arrangements: Differences
in economic and time resources. American Journal of Economics and
Sociology (1996) 55(3):27792.
17. Mincy, R., and Pouncy, H. Paternalism, child support enforcement,
and fragile families. In The new paternalism. L. Mead, ed. Washington, DC: Brookings Institution Press, 1997, pp. 13060; and
Sorensen, E., Mincy, R., and Halpern, A. Redirecting welfare policy
toward building strong families. Strengthening Families Series, no. 3.
Washington, DC: Urban Institute, March 2000.
18. Kiernan, K. Cohabitation in Western Europe. Population Trends
(1999) 96:2532.
1994.
20. Duncan, G.J., and Brooks-Gunn, J., eds. Consequences of growing up
poor. New York: Russell Sage Foundation, 1997.
2000 green book: Background material and data on programs within
the jurisdiction of the Committee on Ways and Means. Washington,
DC: U.S. Government Printing Office, 2001, table H-6, p. 1289.
22. Shapiro, A., and Lambert, J.D. Longitudinal effects of divorce on the
quality of the fatherchild relationship and on fathers psychological
well-being. Journal of Marriage and the Family (1999) 61:397408.
23. Seltzer, J. Relationships between fathers and children who live apart:
The fathers role after separation. Journal of Marriage and the Family
(1991) 53:79101.
24. This underrepresentation occurs because low-income men are difficult
to locate or because they either intentionally or unknowingly misreport
their parental status. See Garfinkel, I., McLanahan, S., and Hanson, T.
A patchwork portrait of nonresident fathers. In Fathers under fire: The
revolution in child support enforcement. I. Garfinkel, S. McLanahan, D.
Meyer, and J. Seltzer, eds. New York: Russell Sage Foundation, 1998,
pp. 3160.
25. Furstenberg, F., Morgan, P., and Allison, P. Paternal participation and
childrens well-being after marital dissolution. American Sociological
Review (1987) 52:695701.
26. McLanahan, S., Seltzer, J., Hanson, T., and Thomson, E. Child support enforcement and child well-being: Greater security or greater
conflict? In Child support and child well-being. I. Garfinkel, S.S.
161
McLanahan, and P.K. Robins, eds. Washington, DC: Urban Institute

Press, 1994.
27. Amato, P.R., and Rivera, F. Paternal involvement and childrens
behavior problems. Journal of Marriage and the Family (1999)
61:37284; Floyd, K., and Morman, M.T. Affection received from
fathers as a predictor of mens affection with their own sons: Tests of
the modeling and compensation hypotheses. Communication Monographs (December 2000) 67(4):34761; Harris, K.M., Furstenberg
Jr., F.F., and Marmer, J.K. Paternal involvement with adolescents in
intact families: The influence of fathers over the life course. Demography (1998) 35(2):20116; Harris, K.M., and Marmer, J.K. Poverty,
paternal involvement, and adolescent well-being. Journal of Family
Issues (1996) 17(5):61440; and Zimmerman, M.A., Salem, D.A.,
and Maton, K.I. Family structure and psychosocial correlates among
urban African-American adolescent males. Child Development (1995)
66:15981613.
28. Amato, P.R., and Gilbreth, J.G. Nonresident fathers and childrens
well-being: A meta-analysis. Journal of Marriage and the Family
(1999) 61:55773.
29. See, for example, King, V. Nonresident father involvement and child
well-being: Can dads make a difference? Journal of Family Issues
(1994) 15(1):7896; King, V. Variation in the consequences of nonresident father involvement for childrens well-being. Journal of Marriage and the Family (1994) 56:96372; and Simons, R.L.,
Whitbeck, L.B., Beaman, J., and Conger, R.D. The impact of mothers parenting, involvement by nonresidential fathers, and parental
conflict on the adjustment of adolescent children. Journal of Marriage and the Family (1994) 56:35674.
30. See note 2, Lamb.
31. See note 28, Amato and Gilbreth; and note 29, Simons, et al.
32. Furstenberg Jr., F.F., and Harris, K.M. The disappearing American
father? Divorce and the waning significance of biological fatherhood.
In The changing American family. S.J. South and S.E. Tolnay, eds.
Boulder, CO: Westview Press, 1992, pp. 197223.
33. Weiss, Y., and Willis, R.J. Children as collective goods in divorce settlements. Journal of Labor Economics (1985) 3(3):26892.
34. Becker, G.S. A treatise on the family. Cambridge, MA: Harvard University Press, 1991.
35. Thornton, A. Influence of the marital history of parents on the marital and cohabitational experiences of children. American Journal of
Sociology (1991) 96(4):86894.
36. McLanahan, S., Garfinkel, G., Reichman, N., and Teitler, J. Unwed
parents or fragile families? Implications for welfare and child support
policy. In Out of wedlock: Trends, causes and consequences of nonmarital fertility. L. Wu and B. Wolfe, eds. New York: Russell Sage Foundation, 2001, pp. 20208.
employment of welfare recipients. In Prosperity for all? The economic
boom and African Americans. R. Cherry and W.M. Rodgers, III, eds.
New York: Russell Sage Foundation. 2000, pp. 24578.
38. Raphael, J., and Tolman, R.M. Trapped by poverty and trapped by
abuse: New evidence documenting the relationship between domestic violence and welfare. Executive summary. Chicago: Taylor Institute, and
Ann Arbor, MI: University of Michigan Research Development Center on Poverty, Risk, and Mental Health, 1997.
39. Rennison, C. M., and Welchans, S. Intimate Partner Violence. Bureau
of Justice Statistics Special Report. Washington, DC: U.S. Department of Justice, Office of Justice Programs, May 2000. Available
online at http://www.ojp.usdoj.gov/bjs/pub/pdf/ipv.pdf.
162
40. Brown, S.S., and Eisenberg, L. The best intentions: Unintended pregnancy and the well-being of children and families. Washington, DC:
41. Henshaw, S.K. Unintended pregnancy in the United States. Family
Planning Perspectives (1998) 30(1):2429, 46.
42. Ooms, T. Strategies to reduce nonmarital childbearing. In Report to
Congress on out-of-wedlock childbearing. Hyattsville, MD: U.S. Government Printing Office, 1995, pp. 24165.
43. See note 11, Ventura, et al, 1999.
44. Terry-Humen, E., Manlove, J., and Moore, K.A. Births outside of
marriage: Perceptions vs. reality. Research Brief. Washington, DC:
Child Trends, April 2001.
45. Wertheimer, R., and Moore, K. Childbearing by teens: Links to welfare
reform. New federalism: Issues and options for states, no. A-24.
Washington, DC: Urban Institute, August 1998.
46. Kirby, D. No easy answers: Research findings on programs to reduce
teen pregnancy. Washington, DC: National Campaign to Prevent
Teen Pregnancy, 1997.
47. Child Trends. Trends in sexual activity and contraceptive use among
teens. Research Brief. Washington, DC: Child Trends, 2000.
48. Kirby, D. Emerging answers: New research findings on programs to
reduce teen pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy, May 2001. See also Moore, K.A., Miller, B.C.,
Morrison, D.R., and Glei, D.A. Adolescent sex, contraception, and
childbearing: A review of recent research. Washington, DC: Child
Trends, 1995.
49. Murray, Charles. Losing ground: American social policy, 19501980.
New York: Basic Books, 1984.
50. Moffitt, R.A. The effect of welfare on marriage and fertility. In Welfare, the family, and reproductive behavior. R.A. Moffitt, ed. Washington, DC: National Academy Press, 1998, pp. 5097.
51. Stark, S., and Levin-Epstein, J. Excluded children: Family cap in a
new era. Washington, DC: Center for Law and Social Policy, 1999.
52. Camasso, M.J., Harvey, C., Jagannathan, R., and Killingsworth, M. A
final report on the impact of New Jerseys Family Development program
experimental-control group analysis. New Brunswick, NJ: Rutgers
University, 1998.
53. Turtorro, C., Benda, B., and Turney, H. Arkansas Welfare Waiver
Demonstration project: Final report. Little Rock: University of
Arkansas at Little Rock School of Social Work, 1997. See note 51, as
cited in Stark and Levin-Epstein.
54. Garfinkel, I., Meyer, D.R., and McLanahan, S. A brief history of child
support policies in the United States. In Fathers under fire: The revolution in child support enforcement. I. Garfinkel, S. McLanahan, D.
Meyer, and J. Seltzer, eds. New York: Russell Sage Foundation, 1998,
pp. 1430.
55. See note 21, U.S. House of Representatives, Committee on Ways and
Means, p. 467.
56. Office of Child Support Enforcement. Child Support Enforcement
twenty-third annual report to Congress. Washington, DC: U.S.
Department of Health and Human Services, Administration for Children & Families, OCSE, 2000, table 2.
57. Sorensen, E., and Turner, M. Barriers in child support policy: A
review of the literature. Paper prepared for the System Barriers
Roundtable, sponsored by the National Center of Fathers and Families. Philadelphia, PA, May 29, 1996.
58. Sorensen, E. Noncustodial fathers: Can they afford to pay more child
Volume 12, Number 1
support? Washington, DC: Urban Institute, 1995; see also note 24,
Garfinkel, McLanahan, and Hanson.
59. Johnson, E., Levine, A., and Doolittle, F. Fathers fair share: Helping
poor men manage child support and fatherhood. New York: Russell
Sage Foundation, 1999.
60. See note 24, Garfinkel, et al.
61. Historically, states were required to pass through the first $50 per
month in child support to mothers, which at least improved the familys well-being slightly. This requirement was eliminated with the
1996 welfare reform law, but nearly half the states have chosen to
continue or increase the pass-through with state monies. See Turetsky,
V. Realistic child support policies for low income fathers. Kellogg Devolution Initiative Paper. Washington, DC: Center for Law and Social
Policy, March 2000. Available online at
http://www.clasp.org/pubs/childenforce/kellogg.htm.
62. Sorensen, E. Obligating dads: Helping low-income noncustodial fathers
do more for their children. Washington, DC: Urban Institute, 1999.
63. This provision was enacted in the 1986 Bradley amendment. See
note 61, Turetsky.
64. Case, A. The effects of stronger child support enforcement on nonmarital fertility. In Fathers under fire: The revolution in child support
enforcement. I. Garfinkel, S. McLanahan, D. Meyer, and J. Seltzer,
eds. New York: Russell Sage Foundation, 1998, pp. 191215;
Garfinkel, I., Gaylin, D.S., McLanahan, S.S., and Huang, C. The roles
of child support enforcement and welfare in nonmarital childbearing.
Working Paper No. 200006. Princeton, NJ: Princeton University,
Center for Research on Child Well-being, 2000; and Plotnick, R.D.,
Ku, I., Garfinkel, I., and McLanahan, S. The impact of child support
enforcement policy on nonmarital childbearing. Paper presented at
the Annual Meeting of the Population Association of America, Washington, DC, March 2931, 2001.
65. Nixon, L. The effect of child support enforcement on marital dissolution. The Journal of Human Resources (1997) 32(1):15981.
66. Seltzer, J.A., McLanahan, S.S., and Hanson, T.L. Will child support
enforcement increase fatherchild contact and parental conflict after
separation? In Fathers under fire: The revolution in child support
enforcement. I. Garfinkel, S. McLanahan, D. Meyer, and J. Seltzer,
eds. New York: Russell Sage Foundation, 1998, pp. 157190; see also
note 23, Seltzer.
67. Knox, V., and Redcross, C. Parenting and providing: The impact of
Parents Fair Share on paternal involvement. New York: Manpower
Demonstration Research Corp., 2000.
68. For more information about the Welfare-to-Work program, see the
U.S. Department of Labor Web site at
http://wtw.doleta.gov/resources/fact-noncustodial.asp.
69. See note 21, U.S. House of Representatives, Committee on Ways and
Means, pp. 44243.
70. Nightingale, D.S. Program structure and service delivery in eleven Welfare-to-Work grant programs. Princeton, NJ: Mathematica Policy
Research, January 2001.
71. U.S. Department of Health and Human Services. HHS awards child
support waivers to help promote responsible fatherhood. Press release.
Hyattsville, MD: DHHS, March 29, 2000.
163
APPENDIX
Innovative Uses of TANF Funds for Programs

for Teen Males and Fathers
Program name:
Adolescent Health and Youth Development
State:
Georgia

Coverage:
34 counties
Program goals:
Improve teen health and skills;

reduce teen pregnancy
Service population: Teens age 19 and under

Funding sources:
TANF and other state funds
Description: Georgia is spending TANF funds on the Adolescent

Health and Youth Development (AHYD) initiative, a comprehensive
pregnancy prevention program based on youth development principles. AHYD offers three types of grants to counties: 1) teen center
grants, for localities to provide comprehensive health services,
including contraception, as well as abstinence-based programs and
youth development services; 2) male involvement program grants,
to reduce pregnancy, promote abstinence, and increase young
fathers involvement with their children; and 3) community involvement grants, to foster partnerships between organizations and fill
gaps in service. For example, the program located at Teen Headquarters, a community center in Bibb County, not only provides a
male-only health clinic, it also offers holistic programming related
to the many pressures specifically confronting young men, such as
gang involvement, drug abuse, and the emotional and physical risks
of sexual involvement. Young men in the community have a place to
go where they feel they belong, can participate in structured and
unstructured recreational time, and find adult and peer support.
Results: AHYD services are reaching thousands of young people. In
1998, AHYD male involvement programs served a little more than
2,000 youths. By fiscal year 2000, that number had climbed to
8,664 adolescents in 13 programs.
For further information: See the programs Web site at
http://www.ph.dhr.state.ga.us/programs/adolescent/index.shtml.
164
Volume 12, Number 1
Program name:
Step-Up Young Father Program
State:
Arizona
Coverage:
Phoenix
Program goals:
Help young, low-income fathers meet their

parental responsibilities
Program name:
Wisconsin Works Child Support

Demonstration
State:
Wisconsin
Coverage:
Statewide
Program goals:
Increase child support collections from

fathers
Service population: Low-income fathers, age 22 and under

Service population: TANF families with a nonresident parent
Funding sources:

Funding sources:
Description: While many new programs are focusing on the importance of fathers in childrens lives, the Step-Up program is particularly comprehensive. Counseling and case management services
provide the foundation of services for young fathers, but the scope
of program activities is much more diverse. This collaborative program works with volunteer mentors, Gateway Community College,
and the Phoenix Job Training Partnership, to offer educational services and job training. Other services offered include legal assistance
in paternity establishment and services to meet basic needs, such as
housing and clothing. Additionally, the citys Parks and Recreation
Department sponsors an annual Family Camp event for the young
men and their families.
Results: Evaluation results showed improvements in the mens
employment at full-time jobs, increases in hourly wages, and
greater educational attainments. Other promising results include
improved relationships with spouses and children, lower levels of
gang involvement, and lower levels of substance abuse.
For further information: See the programs Web site at
http://www.ci.phoenix.az.us/YOUTH/stepup.html.
Description: In the fall of 1997, as Wisconsins new welfare system was taking hold, the state began testing the idea that fathers
would be more likely to pay child support if they knew their children would receive all the money. Generally speaking, only about
30% of single mothers nationwide ever receive child support. In an
effort to improve that rate, Wisconsin passes through every dollar of child support paid by the father directly to the mother receiving welfare. The family not only keeps 100% of the child support
but also retains their monthly cash grant at the same level, so there
is no penalty for these families. Wisconsin is unique among
states; many states keep all of the child support fathers pay if their
children are receiving cash assistance, or pass through a small
amount such as $50, the amount of the required pass-through
under the earlier AFDC program.
Results: The program has achieved some success in encouraging
fathers to pay child support. When compared to a control group of
families still under old AFDC rules, families in the new program
received more child support, as much as 50% more, which translates into an additional $1,500 a year. Fathers of children under the
new program are more likely to pay support and, on average, are
likely to pay more than fathers of children in the control group.
For further information: See Meyer, D.R., and Cancian, M. W-2 Child
Support Demonstration evaluation. Phase 1: Final report. Madison, WI:
University of Wisconsin Institute for Research on Poverty, April 2001.
Available online at http://www.ssc.wisc.edu/irp/csde/phase1-tocs.htm.
165
Welfare Reform and Parenting:

Reasonable Expectations
P. Lindsay Chase-Lansdale
Laura D. Pittman
SUMMARY
Although the primary goals of federal welfare

reform legislation were to move welfare mothers into the workforce and reduce births outside
of marriage, promotion of responsible parenting was also an important underlying theme.
Parenting is a complex, multifaceted phenomenon, however, encompassing a wide range of
functions related to nurturing, discipline, stimulation, values, activities, and routines. This
article provides a framework for assessing the
impact of welfare reform on various dimensions
of parenting, with the following key findings:
Many aspects of life affect parenting and child
development, such as parent characteristics,
child characteristics, family economic
resources, family structure, parental mental
health, marital or partner relationships, and
the quality of parents kin and social networks.
About two-thirds of states are using federal
welfare funds to promote better parenting
through programs such as home visits to new
parents and parenting classes, but virtually no
state parenting programs have been evaluated.
Welfare reform appears to have limited effects
on parenting. The only dimension of parenting significantly affected by some welfare
demonstration programs was parents choice

of child care settings and extracurricular activities for their children.
The programs with the greatest positive
impact on parenting were those with more
generous work supports and more flexible
work requirements. Not only did these programs lead to different choices concerning
child care and activities for preschool and
school-age children, but they also resulted in
more stable marriages and less violence
between partners, which also could lead to
improved parenting.
The authors conclude that many important
aspects of the connection between welfare
reform and parenting have yet to be examined,
and that further research is needed to identify
the ways states welfare programs can promote
better parenting.
P. Lindsay Chase-Lansdale, Ph.D., is professor of developmental psychology at the School of Education and Social
Policy Program in Human Development and Social
Policy, and faculty fellow at the Institute for Policy
Research at Northwestern University in Evanston, IL.
Laura D. Pittman, Ph.D., is a research scientist at the
Institute for Policy Research at Northwestern University in Evanston, IL.
167
Chase-Lansdale and Pittman
elfare reform in the 1990sbeginning

with state waivers and culminating in
the federal welfare reform law enacted
in 1996represents the most wideranging change in policies for low-income families since
the first federal welfare program began in 1935. (See
the article by Greenberg and colleagues in this journal
issue.) Although the primary goals of reform were to
move welfare mothers into the workforce and reduce
births outside of marriage, promotion of responsible
parenting is an underlying theme in many of the laws
provisions.1
The law does not specify what constitutes responsible
parenthood, but about two-thirds of states are using
federal welfare funds to promote better parenting practices as part of their new welfare programs.2 In addition,
policymakers argue that parental employment, family
income, and family structure all have significant implications for parenting. For example, proponents of welfare reform believe that if mothers leave welfare for
employment, they will be better parents because they
will provide better routines and serve as better role
models for their children. Similarly, those who view welfare reform as a pathway out of poverty argue that
increased economic resources will lead to better parenting. Finally, those who champion welfare reform as a
way to promote marriage believe that married parents
are more effective than single parents.
In policy debates over welfare reform, parenting tends
to be discussed simplistically. A more thoughtful discussion requires a better understanding of what shapes parenting. In reality, parenting is a complex, multifaceted
phenomenon that ignites controversy in scientific and
policy circles but is also a very personal, private experience. This article provides a framework for assessing the
impact of welfare reform on parenting. The first section
reviews the scientific literature describing the multiple
dimensions of parenting and discusses how each dimension is linked to child development. The second section
presents a model that synthesizes research evidence
about key factors believed to influence parenting. The
third section summarizes what has been learned to date
about the direct and indirect effects of welfare reform
on parenting. A final section discusses the limitations of
current studies, and the areas where further research
168
would be most helpful in understanding how states

welfare reform programs could more effectively support
positive parenting practices.
Dimensions of Parenting and Links to

Child Outcomes
Extensive research supports the widespread belief that
parenting does, in fact, have an important influence on
developmental outcomes for children and adolescents.3
Box 1 displays six key dimensions of parenting that are
interrelated yet distinct. Each dimension affects parenting styles, and therefore child outcomes, in unique ways.
Box 1
Dimensions of Parenting
Warmth and responsiveness: Parents express love and
affection and are responsive to their childs needs and
requests.
Control and discipline: Parents outline specific rules and
expectations and enforce them consistently. As a child
grows and matures, parents continue to set limits, but allow
for more autonomy in the childs world.
Cognitive stimulation: Parents ensure that their child has
materials that are stimulating; parents are verbally engaging and actively teach their children key concepts.
Modeling of attitudes, values, and behaviors: Parents discuss their values, convey their attitudes, and act toward
their child and others in the way they want their child to act.
Gatekeeping: Parents serve as links to the environment,
influencing which family and friends their child interacts
with and what outside activities and programs they become
involved in. Parents become involved in school and other
community activities to maintain connection with the child
and outside influences.
Family routines and traditions: Parents create a daily routine as well as family traditions that help structure a childs
expectations for the day as well as promote knowledge of
cultural and family heritage.
Volume 12, Number 1
Welfare Reform and Parenting
Although the primary goals of reform were to move welfare

mothers into the workforce and reduce births outside of
marriage, promotion of responsible parenting is an underlying theme in
many of the laws provisions.
Warmth and Responsiveness
Infants and toddlers whose parents are consistently
warm and responsive develop secure attachments to
their parents and rely on parents or attachment figures as a secure base from which to explore the
environment.4 These secure attachments at a young
age provide an important emotional foundation for
later development and help promote healthy peer relations and fewer problem behaviors in childhood and
adolescence.5 Children and adolescents whose parents
are warm and supportive also have higher levels of selfesteem and social competence as well as lower levels of
depression, anxiety, and problem behaviors.6,7
and school achievement in early and middle childhood

when parents are more verbally engaging and provide
more cognitively stimulating toys, activities, and interactions.15,16 During the school years, parents active
involvement with their childrens homework and
schools is related to better academic achievement.17
Modeling
Whether consciously or unconsciously, parents model
behavior for their children every day. Modeling encompasses a broad array of experiences, and children can learn
both positive and negative behaviors through witnessing
adults responses to a variety of situations.18 For example,
when a child witnesses domestic violence or excessively
harsh punishment of a sibling, that child is more likely to
act aggressively toward others. In contrast, when children
see their parents displaying considerate, empathic, or
moral behavior, they learn to consider what is right and
wrong and how others will respond to their actions.19
Control and Discipline

Parents discipline style and level of supervision are key
to childrens healthy development. Parents need to take
childrens age and developmental level into account
when deciding how to approach discipline and supervision. Younger children need more structure and
boundaries to keep them safe, whereas older children
should be granted more autonomy as they mature and
transition into adulthood. Parents who provide firm,
consistent limits and rules teach children what type of
behavior is socially acceptable and how to regulate their
own behavior. These children fare better in school and
with their peers.8,9 Parents who know where their
teenagers are and set limits such as curfews, but who
also grant some autonomy, have adolescents with fewer
behavior problems, including lower levels of drug and
alcohol use, school suspensions and expulsions, and
police involvement.10,11 Authoritative parentsthose
who are warm but firmhave children who do better
on virtually all developmental outcomes compared with
children whose parents are low on either warmth or
control.12,13 These links have been found across different countries and socioeconomic and cultural groups.14
Gatekeeping
Parents serve as gatekeepers to the world for their children. They decide what neighborhood the family will
live in, and they provide access to community organizations, such as child care settings, parks, and playgrounds,
as well as opportunities to interact with neighbors, peers,
and kin.20 As children grow up, parents continue to
influence who their childrens friends are,21 as well as
whether children will be involved in extracurricular activities and after-school programs.22 These gatekeeping
activities are related to childrens social and academic
competence, as participation in activities outside the
home promotes more positive peer interactions, greater
feelings of self-worth, and advances in learning.23,24 Adolescents are more likely to take initiative outside the family than are young children, but many adolescent
decisions are still influenced by family life.
Cognitive Stimulation
The stimulation provided by parents is especially
important for infants, toddlers, and young children.
Children show higher levels of cognitive functioning
Family Routines and Traditions

Parents also influence their children by providing
structure in their daily lives. Children who come from
families who have a regular, predictable routine and
169
time together are believed to do better than those

whose family life is less organized,25,26 although
research on this dimension is limited. Parents also
affect their children by initiating and maintaining
familial, religious, and cultural traditions. When parents engender an understanding of the familys cultural and ethnic heritage, children develop a stronger
sense of ethnic identity,27 which is associated with both
cognitive and social competence.2830
Determinants of Parenting
To understand how welfare reform might affect parenting, an overview of the key factors that influence
parenting is essential. Figure 1 presents a model of
these factors, illustrating the complexity of the connections between these factors and parenting, and the relationships among the factors themselves. The model
shows that parenting is affected by many aspects of life:
parent characteristics (such as personality, cognitive
abilities, and level of education); child characteristics
and the shared genes between parents and children;
family economic resources (including income from
work and/or welfare); and family structure and size.
These four determinants are seen as influencing the
three othersparental mental health, marital or partner relationships, and the quality of parents kin and
social networkswhich in turn influence parenting
and child development.
Parent Characteristics
The first determinant in the model is the parents personal characteristics, which potentially influence every
other determinant as well as parenting. Parents characteristics encompass a wide variety of attributes,
including the parents own experience of being parented, age and education level, cognitive ability, personality, and other traits. Parents tend to model their own
parents child-rearing practices, including both supportive or more harsh parenting.31,32 Parenting patterns also are influenced by parents mental frameworks
or internal working models, which are based on and
adapted from experiences with their own parents.5 In
addition, parents with higher intellectual ability and
levels of education, and those with positive personalities, tend to display more effective parenting.33,34 For
example, such parents tend to be more responsive to
childrens emotional needs, engage their children
more, provide a more cognitively stimulating environment, and explain their punishments.
Figure 1
Determinants of Parenting Model
Child
Characteristics
Parent
Characteristics
Distal Influences
on Parenting
Proximal Influences
on Parenting
Family economic
resources
Parental mental and

physical health
Family
structure/size
Marital/partner
relationship quality
Parenting
Child
Outcomes
Kin and social

networks
170
Volume 12, Number 1
Single mothers tend to have less positive interactions with their

children and provide less firm and consistent discipline as
compared with mothers in two-parent households ... primarily
because single mothers have more stressors in their lives.
Child Characteristics
Children influence their parents through their personalities, temperaments, and special needs.35,36 If a child has
a difficult temperament or personalitycrying or complaining much of the timethe parent is more likely to
respond negatively, providing harsher punishment and
fewer positive interactions than if a child is more easygoing.34,37 This pattern could cause developmental
problems. The childs innate characteristics, however,
do not set in stone how the child will develop.
Although the debate over nature versus nurture has
been an intense focus of scientific research for most of
the past century, current research shows that childrens
development is shaped by a complex interaction of
genetic and environmental influences.38 Thus, although
a child with a difficult temperament presents a challenge
and a risk, effective parenting can shape such children to
become more socially competent and pleasant.39 In
other words, children do indeed influence how parents
respond to them, but parents also have a role in shaping childrens behavior and development.
Economic Resources
Family economic resources, including income from
employment as well as other sources, such as welfare,
influence parenting both directly and indirectly. Like
many parents who are single, those facing economic
hardships tend to be less effective because they have
more stressors in their lives and thus are likely to have
greater psychological distress.40,41 Low-income parents
have been found to use less effective parenting strategies, including less warmth, harsher discipline, and less
stimulating home environments.42,43
Family Structure and Size
Family structure and economic resources have interrelated effects on parenting, as households with one parent,
rather than two, have fewer economic and emotional
resources. Single mothers tend to have less positive interactions with their children and provide less firm and consistent discipline as compared with mothers in two-parent
households.44,45 This is primarily because single mothers
have more stressors in their lives, poorer mental health,

and no spousal or partner support. In addition, children
with more siblings have fewer financial resources available
to them as well as less time with and attention from parents.46 Depending on their ages, however, older siblings
may help care for younger siblings.47
Parents Mental and Physical Health
Research evidence is very strong regarding the link
between parental mental health and the ability to parent effectively. Parents who are experiencing psychological distress, whether diagnosed as a mental disorder
or not, are less likely to be warm and supportive of
their children and are less effective in monitoring their
children or providing consistent discipline.48,49 The
association between parental physical health and parenting is less well established, with some indication that
disability or poor physical health may lead to more
family stress and parental depression, which in turn
may lead to less effective parenting.50
Quality of Marital/Partner Relationships
Similarly, the quality of the parents relationship with a
spouse or partner as well as the size and supportiveness of
kin networks have direct links to parenting effectiveness.
The marital relationship is often considered the cornerstone of good family functioning, and a positive marital
relationship is associated with more positive parenting.
Conversely, chronic parental conflict is considered to
have serious adverse effects on child development, causing more negative parentchild relationships and less
consistent and effective discipline and monitoring of children.5153 The most extreme form of parental conflict is
domestic violence, a phenomenon associated with poor
mental health and compromised parenting.54
Kin and Social Networks
Mothers with strong networks of positive social support from friends and extended family are more effective parents.55,56 However, if interactions with kin
generate conflict, or if kin are excessively demanding of
parents time and energy, such negative relationships
171
Parenting and Welfare Reform

The federal welfare reform law of 1996 was couched in
broad language: The promotion of responsible fatherhood and motherhood is integral to successful child
rearing and the well-being of children.1 As a result,
the scope of programs implemented under welfare
reform is wide-ranging, and nearly every factor illustrated in the Determinants of Parenting Model (Figure
1) could be affected. The challenge is to specify how
the changes under welfare reform are affecting parenting, either directly or indirectly, based on findings from
relevant research.
Stone/David Hanover
Direct Effects on Parenting
can lead to higher levels of stress and depressed

mood,57,58 both of which are linked to more disruptive
parenting. Determinants of parental mental health,
quality of marital/partner relationships, and kin and
social networks are interrelated and influence one
another as well as affect parenting.59
Although this model of the determinants of parenting is
based on the most current developmental literature, it has
limitations. Much of the research reflects findings from
samples that were not randomly selected and, in many
instances, included primarily white, middle-class families.
In addition, no definitive statement can be made that
these determinants cause the differences in parenting. For
example, it is not possible to hold everything constant in
a family to evaluate whether one factor influences parenting independent of other factors, or to randomly assign
children to parents. Moreover, while this model may be
relevant to most families, further research is needed to
determine the extent to which the model applies to ethnically and culturally diverse, low-income families.60,61
Finally, some areas of research rely heavily on maternal
self-report, which may be biased by the mothers view of
the world or her own psychological health. Nevertheless,
the model provides a useful framework for analyzing how
welfare reform may affect parenting.
172
Many states are using welfare funds to implement initiatives designed to improve parenting directly, such as providing home visits to new parents or requiring certain
welfare recipients to attend parenting classes.2 (See the
appendix at the end of this article for three examples of
welfare-supported programs that are intended to help
improve parenting.) Assessments of these initiatives are
quite limited. In the scientific fields, however, several
theoretically based programs to improve parenting have
been developed and evaluated. Studies have shown that
these interventions are most successful when children
are having socioemotional or developmental problems.34
In such situations, parents are motivated to change their
parenting practices to help their children. Furthermore,
the interventions tend to be provided in small groups led
by professionals with extensive training in child development and expertise in parenting and family issues. Such
interventions can result in long-lasting improvements in
parents knowledge and attitudes, use of appropriate discipline, and parentchild interactions.62
In contrast, parenting programs that target low-income
families tend to be broad in scope and have multiple goals,
such as promoting job readiness and increasing access to
social services. These programs generally have only modest effects on parenting that lessen with time.63,64 Several
factors influence the effectiveness of such programs. Better outcomes are more likely when professionals, rather
than paraprofessionals, meet with the participants, when
families can expect more frequent contact, and when parents participate more fully in the intervention (for example, they miss fewer meetings and follow through on tasks
expected to be done outside of meetings).65 Effects are
Volume 12, Number 1
If maternal employment decreases psychological distress, this

could result in increased parental warmth, responsiveness, and
appropriate control ... [but] if maternal employment increases mental
health problems, then parenting would be more harsh, less
responsive, and less consistent.
lessened when the parenting intervention is secondary to
other areas of focus in a program, such as parental education or employment.66 Thus, to improve the effectiveness
of welfare-funded parenting programs, such programs
need to be more intensive and led by professionals with a
thorough understanding of developmental and family
issues, and involve parents who have chosen to participate
and are actively engaged in the intervention.
Indirect Effects on Parenting

Programs designed to achieve the primary goals of welfare reformto move welfare mothers into the workforce and reduce births outside of marriageare likely
to influence parenting indirectly, as they affect several of
the key determinants of parenting such as family economic resources, parents mental health, and partner
relationship quality (see Figure 1). In fact, many policymakers expected that by promoting employment, earnings, and marriage, parenting practices would improve.
Most studies focusing on welfare reforms broader goals
have not collected data on the potential impacts on parenting, either positive or negative. However, several
large-scale studies currently under way in various states
include components specifically designed to illuminate
how the welfare reform law has affected families and parenting, and preliminary findings from these studies are
emerging.67 For example, the Three-City Study will provide detailed information about any changes that occur
in parenting practices as families transition on and off the
welfare rolls. (See Box 2.) Some clues are also provided
by the findings from several experiments or demonstrations begun in the early to mid-1990s that implemented
various programs with features similar to current welfare
reform policies.68 (For a detailed description of these
studies, see the article by Zaslow and colleagues in this
journal issue.) What can been gleaned from existing
research about how changes in employment, earnings,
and marriage under welfare reform may affect the various dimensions of parenting is summarized below.
The Connection between Employment

and Parenting
Welfare reform has been galvanized by the opinion that
employed mothers are better parents than mothers on
welfare.69 The positive scenario is that mothers who
move from welfare to work will have better mental
health, leading to improved parenting. The underlying
hypothesis in this scenario is that when mothers find
employment challenging and rewarding, they will provide more cognitive stimulation to their children.70 Policymakers have also contributed considerable rhetoric to
the notion that employed mothers will be better role
models for their children, and that employment per se
will lead to more predictable, organized daily routines at
home. If maternal employment decreases psychological
distress, this could result in increased parental warmth,
responsiveness, and appropriate control.
The counterhypothesis is that mothers who leave welfare
for employment will feel more stressed and have worse
mental health. Mothers in poverty, most of whom are
single, may find it difficult to balance employment, especially full-time work, and child rearing. The tedium and
low levels of complexity in low-wage jobs may be linked
to lower levels of cognitive stimulation at home.71 Time
spent at work, combined with changing, off-hour, and
inflexible work schedules, may mean less supervision and
monitoring of children, the inability to respond to the
erratic demands of parenting (such as teacher meetings
and doctor appointments), and a lower likelihood of
establishing and maintaining predictable family routines.
If maternal employment increases mental health problems, then parenting would be more harsh, less responsive, and less consistent.
Psychological and sociological studies show that
employed mothers have better mental health than stayat-home mothers, but this literature focuses primarily
on the middle class.72 Moreover, it is just as likely that
mothers with good mental health are better able to
173
Box 2
Welfare, Children, and Families: A Three-City Study
The Three-City Study is an ongoing research project in Boston,
Chicago, and San Antonio designed to monitor the impact of welfare reform on the well-being of children and families. Parenting
is a key focus across all three components of the study.
The main component consists of an in-person survey of approximately 2,400 families with children ages 0 to 4, or ages 10 to 14, in
low-income neighborhoods. (About 40% of the families were receiving cash welfare payments when they were interviewed initially.) In
this survey, mothers report on their parenting practices and discipline
style, their family routines, and the cognitive stimulation provided
within the home. For families with young adolescents, the teenager
reports on closeness and conflict within the motherchild and
fatherchild relationship as well as the level of parental monitoring.
The second component consists of an embedded developmental
study (EDS) of families from the main survey with children ages 2
to 4. Supplemental data are gathered about these families
through videotaped assessments of childrens behavior and motherchild interactions, which are then coded for maternal responsiveness and emotional expression, and through observations of
child care settings. More in-depth interviews with the mothers
provide further reports of activities they do with the children as
well as the type of punishment used. In addition, biological fathers
are asked about their parenting styles and the activities they have
shared with their children.
The third component is an ethnographic study of about 215 families residing in the same neighborhoods as the survey families,
seek, find, and keep jobs outside the home as it is that

employment promotes mental health.73 These studies
indicate that the quality of parenting is similar for
employed and nonemployed mothers, but again, this
conclusion is biased toward the middle class and is not
really relevant to mothers who are required by welfare
policies to be employed in low-level jobs that offer little flexibility. Within the limited available literature,
there is some indication that employed single mothers
may provide a more cohesive, stimulating, and organ-
174
using in-depth interviewing and participant observation. Within

the ethnography, interviewers take a broad perspective on parenting, inquiring into a range of processes that affect the transmission of values and expectations from parents to children.
Through loosely structured conversations with the mothers, the
ethnography focuses primarily on the parenting domains key to
understanding the ways in which economically disadvantaged
children from various ethnic and racial groups navigate multiple
contexts such as home, school, and neighborhood. Issues of racial
socialization, parental monitoring strategies, and socialization for
achievement and social mobility are also explored.
The main survey was conducted in 1999. Follow-up interviews
with these families and those in the EDS sample were conducted
approximately 18 months after the initial survey, and will be conducted again at the 5-year mark. Also at the 5-year mark, a second sample of about 1,250 families, including young parents who
are encountering the welfare system for the first time under the
new rules, will be selected and interviewed. Families in the ethnographic study are followed for 12 to 18 months and periodically
thereafter. As of October 2001, six reports had been released summarizing the findings across the various components of the study,
and more were planned.
For more details, see Winston, P. with Angel, R.J., Burton, L.M., Chase-Lansdale,
P.L., et al. Welfare, Children, and Families: A Three-City StudyOverview and
design. Baltimore: John Hopkins University, 1999. See also the project Web site
at http://www.jhu.edu/~welfare/.
ized family environment, and place more emphasis on

the value of education when compared with unemployed single mothers.74,75 However, parenting by
mothers on welfare, compared with parenting by
mothers who are poor but not on welfare, has not been
a central focus of research.
Results from the experimental studies of seven welfareto-work demonstrations show only limited effects of
employment on a mothers mental health and parent-
Volume 12, Number 1
For the most part, research on welfare-to-work programs

suggests that requiring employment neither benefits nor
harms parenting.
ing.76 In two studies (New Chance and Canadas SelfSufficiency Project), mothers who were required to participate experienced higher levels of depression and stress
than those who were not required to participate.77,78 In
another study (New Hope), parents in the experimental
group reported modest increases in time pressure, but
less stress and greater optimism about achieving life
goals than did parents in the control group, perhaps due
in part to the intensive case management provided.79 But
a synthesis of findings across all seven programs found
that mothers who participated in work preparation activities or who were employed showed no differences in
depression, self-esteem, mastery, and financial strain,
compared with mothers in the control groups.80
These experimental studies also found minimal effects
of employment on all dimensions of parenting except
gatekeeping. Across all seven programs, there were
very few differences between mothers in the program
groups and those in the control groups with respect to
warmth, control, cognitive stimulation, family routines, or harsh parenting.73 In an embedded observational study of parenting and child outcomes in the
New Chance demonstration, mothers in the experimental group showed slightly lower levels of harsh
treatment, and slightly higher levels of emotional support, with mixed effects on cognitive stimulation.
These improvements in parenting did not translate into
better child outcomes, however.81 In the New Hope
study, mothers in the experimental group showed
slightly increased supervision and control, but the
effect was modest.82 Although more in-depth measurement of parenting practices might have found greater
differences, and not all dimensions of parenting were
assessed, these studies indicate that employment or
employment preparation did not lead to major
improvements or problems in most dimensions of parenting in these families.
Gatekeeping is the one important exception to this
pattern. Superior gatekeeping was found in four of the
seven demonstrations (New Hope, the Minnesota
Family Investment Program, Floridas Family Transi-
tion Program, and Canadas Self-Sufficiency Project).83

Mothers in the experimental groups were more likely
to enroll their preschool and elementary-school-age
children in formal child care programs than were
mothers in the control groups. In two of these demonstrations, program mothers were also more likely to
enroll their children in after-school programs and
extracurricular activities.84 It makes sense that mothers
who are required to work 20 to 30 hours per week
would increase their use of child care, especially if child
care assistance is provided. But in these demonstrations, the mothers in the experimental groups enrolled
their children in formal child care, which is likely to be
more structured than informal settings and provide
better preparation for success in school. (For further
discussion of this topic, see the article by Fuller and
colleagues in this journal issue.)
For the most part, research on welfare-to-work programs suggests that requiring employment neither
benefits nor harms parenting. The major exception is
that mothers who are required to work in some programs seem to be better gatekeepers and are able to
find potentially enriching opportunities for their preschool and elementary-school-age children. However,
this improved gatekeeping may be related as much to
increased income as to maternal employment, as discussed below.
The Connection between Increased
Economic Resources and Parenting
A second theme in welfare reform policy is the view
that when parents are employed full time, family
incomes should be higher than when the parents are
on welfare, and that higher incomes will lead to better
parenting. Recent analyses show that poverty has
indeed decreased since the federal welfare reform law
was passed, although the booming economy in the
1990s was also a factor, along with many federal and
state initiatives to make work pay.8587 (See the article by Zedlewski in this journal issue.) The connection
between a familys economic resources and parenting,
however, is complex.
175
In the Minnesota demonstration, many mothers in the optional

employment and financial incentives subgroup decided to work
part time or reduce their work hours, and most of the positive effects
on child outcomes occurred within this subgroup.
Considerable psychological and sociological research
shows that poverty is linked to a higher likelihood of
problematic parenting and negative outcomes for children (such as lower levels of cognitive achievement and
increased levels of behavioral problems and delinquency). But few studies have examined whether parenting
improves as families manage to increase their income to
just above the poverty line. Also, families level of
income (especially at or just above the poverty line) and
source of income (welfare versus employment) have
rarely been disentangled in nonexperimental studies.
cantly lower levels of depression and harsh parenting

than did mothers in the control group. Many mothers
in the optional employment and financial incentives
subgroup decided to work part time or reduce their
work hours, and most of the positive effects on child
outcomes occurred within this subgroup, as opposed
to the subgroup required to work 30 hours per week
with the same financial incentives.92 Thus, the positive
pathway between increased income and better parenting seems to occur when mothers can choose the number of hours they work.
However, three recent experimental programs that

increased work supports for poor, employed families
shed light on this question.88 In all three demonstrations (New Hope, the Minnesota Family Investment
Program, and Canadas Self-Sufficiency Project),
mothers in the program groups successfully increased
earnings compared with mothers in the control
groups, although many still had incomes below the
poverty line.82,89,90 Because gatekeeping (that is, use of
child care and after-school programs) was the only
dimension of parenting that improved among the program groups, evaluators concluded that increased
resources, not the experience of employment itself,
caused this result.
With respect to adolescents, however, evidence suggests that those in families with increased economic
resources may, in fact, do worse. Because of fairly low
response rates in this age group, these findings should
be viewed with caution. Nevertheless, in the two evaluations that included data about adolescents (Canadas
Self-Sufficiency Project and Floridas Family Transition
Program), adolescents in the program groups had
lower levels of school achievement; in one study, they
also had higher levels of problem behavior.93,94 The
evaluators speculate that the mothers employment
requirement in these demonstrations may be linked to
worse parenting of adolescents (that is, lower levels of
supervision and monitoring, and less engagement in
school and homework activities), despite the increase
in income. (For other possible explanations, see the
article by Zaslow and colleagues in this journal issue.)
Moreover, preschool and elementary-school-age children of the mothers in the demonstration program
groups generally fared better developmentally than
children of the mothers in the control groups. The lack
of improvement in other dimensions of parenting
(such as warmth, control, cognitive stimulation, routines) indicates that childrens experiences in child care
and after-school programs likely accounted for the positive outcomes.80
Additional insights are provided in the Minnesota
demonstration that included a subgroup of mothers
with optional, primarily part-time, employment, plus
financial work incentives.91 Although the effects were
modest overall, this subgroup of mothers had signifi-
176
In sum, the research suggests that the effects of

employment and work support programs on income in
poor families can have a positive impact on several
domains of parenting and child outcomes. In addition,
evidence from some experimental demonstrations suggests that increased income and employment are both
required for positive impacts on parenting and child
outcomes to occur. Thus, if families economic
resources do not increase in the move from welfare to
work, parenting and child outcomes are not likely to
improve. Furthermore, emerging evidence indicates
that positive impacts are maximized when parents
Volume 12, Number 1
Courtney Bennett
experience increased resources but are not required to

work full time. This finding comes from only one
study, and further research along these lines is needed.
The Connection between Family Structure
and Parenting
Finally, the third theme of welfare reform relevant to
parenting is that married parents are more effective
than single parents. The preamble of the 1996 welfare
reform law states that marriage is an essential institution of a successful society which promotes the interests of children and that the prevention of
out-of-wedlock pregnancy and the reduction of outof-wedlock births are very important government
interests.95 As discussed in the articles by Greenberg
and colleagues and by McLanahan and Carlson in this
journal issue, states were given great latitude under the
law to design programs to meet goals involving family
structure, such as promoting marriage and reducing
nonmarital births. Whether such programs lead to
improved parenting, however, is unclear.
Most of the psychological and sociological research finds
that children show more competent and healthy devel-
opment in married families with two biological parents

than in single-mother families or stepfamilies.45 This is
largely due to higher economic and parenting resources
in married families.96 However, the research literature
primarily uses national samples representing the full
range of incomes, or samples of convenience that favor
the middle class. Research is needed that examines the
role of marriage in parenting within low-income families. For example, the Fragile Families and Child Wellbeing Study examines the links among cohabitation,
diverse forms of father involvement, and parenting.97
(See the article by McLanahan and Carlson in this journal issue.) To date, few nonexperimental studies have
examined whether efforts to promote marriage lead to
better parenting in the low-income population.
Moreover, the experimental demonstrations did not
focus on the relationship between family structure and
parenting, and so have little data to offer on this subject. Findings from the Minnesota study indicate that
program participants with increased income from
employment and work supports are more likely to
marry if single, more likely to stay married if already
177
The existing research suggests that an important way states

welfare reform programs can promote better parenting is
through better work supports.
married, and less likely to experience domestic violence.98 But no data were collected on parenting or
child outcomes in the two-parent, married families;
thus, no links can be established between marital stability, marital quality, and parenting. This is an important direction for future research.
Only recently, a few states have decided to promote
marriage using federal welfare funds.99 Some evidence
from model intervention programs indicates that such
endeavors facilitate positive marriages and lead to better parenting.100,101 But these studies were generally
small, and focused on more affluent rather than lowincome families.102 This is another important area for
future research, possibly finding some common
ground between liberals and conservatives in forging
policies to promote healthy marriages that lead to better parenting.103,104
Similarly, to help reduce out-of-wedlock births, some
states have implemented family cap provisions that prohibit increases in welfare benefits when an additional
child is born. One research study tested the effects of
the family cap provision on nonmarital birth ratios, and
suggested that the caps may be related to a 5% to 9%
decrease in nonmarital birth ratios.105 However, to
date, no research has examined this aspect of welfare
reform on parenting. Family structure, family size, and
adult relationships and their effects on parenting are
important topics for future research.
Summary and Future Directions

Overall, studies of welfare reform show few dramatic
impacts on parenting. Given the complexity of the
Determinants of Parenting Model, this is not surprising. Parenting practices are not easy to change. It may
be unreasonable to expect major changes in parenting
as a result of welfare reform when the primary goal of
reform is to move mothers into the workforce, not to
provide parenting services.
At the same time, many possible connections between
welfare reform and parenting have yet to be examined.
178
Although many states have funded initiatives to

improve parenting through home visits and parenting
classes, such efforts are largely unstudied. But previous
research on similar programs for low-income families
showed few, if any, long-term effects on parenting.
Similarly, research indicates that welfare reform has had
little indirect impact on parenting. One exception is
gatekeepingthat is, employed mothers enroll their
children in more formal child care and after-school care
programs, which may provide better preparation for
success in school.
The indirect effects of welfare reform are difficult to
analyze, however, and there are many gaps in the
research. In terms of dimensions of parenting, welfare
reform research has targeted warmth, control, cognitive stimulation, and, most recently, aspects of gatekeeping. Modeling and family routines are
understudied, and no research has addressed cultural,
religious, and ethnic identity formation. In terms of
determinants of parenting, most welfare reform
research focuses on economic resources (parental
employment, welfare status, and income) and parents
mental health. Very few studies have addressed family
structure, family size, or the quality of the parents relationships with a spouse or partner and other adults and
kin. Ethnic and cultural influences on parenting
deserve a stronger focus in the literature as well.
More research also is needed on the effectiveness of
activities within the welfare office that could have positive effects on parenting. For example, as suggested by
findings from the New Hope study, case management
can serve as an important form of social and functional support for parents by providing useful information,
practical advice, and contacts. The role of intensive case
management as a strategy for improving parenting
deserves greater attention. (See the article by Zedlewski in this journal issue.) In addition, many families
remaining on the welfare rolls face significant barriers
to employment, such as physical disability, substance
dependence, domestic violence, lower cognitive ability,
and serious mental health conditions.106,107 These bar-
Volume 12, Number 1
riers to employment also could result in problematic

parenting. Implementing procedures to screen for such
barriers and assist families in accessing specialized services is likely to result in improvements in parenting and
child outcomes.108 To date, no evaluations have been
conducted to determine the effectiveness of states procedures for identifying these barriers or helping such
families access needed services, but these efforts also
deserve attention in research on welfare reform and
parenting.
Meanwhile, it is important to note the limitations of
the available research. Many studies predate the 1996
federal welfare reform law, so the findings may not fully
apply to current welfare policy environment. For example, the demonstration projects reviewed in this article
provided more generous work supports than are typically provided by todays state policies. Also, the
research generally did not examine such policies as
sanctions and time limits. A small embedded study
within one demonstration examined the impact of
sanctions on parenting and found no negative
effects,109 but initial reports about sanctioning since the
1996 law was enacted suggest that sanctioned families
are among the most disadvantaged and face numerous
barriers.110112 Similarly, only one demonstration project has evaluated the impact of time limits on families.
In this study, families who hit the time limit and had to
leave the welfare system did not seem to fare worse
than families who left for other reasons.94 Clearly, more
research on the effects of sanctions and time limits on
parenting is needed.
Finally, the measures of parenting in most welfare studies are unavoidably lean. In large part, this is due to the
wide scope of the studies, with sample sizes ranging
from approximately 600 to 3,000 families. Within
these design constraints, most studies relied on parents reports of their parenting styles or practices,
rather than more in-depth or observational assessments, which, though more valid, are time-consuming,
difficult to administer, and expensive.113
Nevertheless, the existing research suggests that an

important way states welfare reform programs can
promote better parenting is through better work supports. Findings from the demonstration programs indicate that more generous work supports appear to result
in improved gatekeeping for preschool and school-age
children (though not for adolescents). That is, parents
who accrue more economic resources due to program
financial supports tend to involve their preschool and
school-age children in extracurricular activities and
more formal child care settings. Further, evidence suggests that the most positive outcomes for children
occur when mothers work a reduced number of hours,
made possible by financial incentives. The demonstrations that provided stronger work supports also resulted in more stable marriages and less violent marital and
partner relationships, which are likely to have a positive
influence on parenting.
Welfare reform by no means assures that a family will
live above the poverty line. In fact, many families participating in demonstration projects with generous
supports have not escaped poverty. Moreover,
although work supports may continue to help families
leave welfare, this does not guarantee an improvement
in parenting. The fact that children living in poverty,
especially pervasive poverty, are more likely to experience developmental problems remains an important
issue facing the nation.
The authors are deeply grateful to the Joyce Foundation, the Robert Wood Johnson Foundation, and the
National Institute of Child Health and Human
Development for their support. They also wish to thank
Jeanne Brooks-Gunn, Andrew Cherlin, Rebekah
Levine Coley, Greg Duncan, Martha Moorehouse,
Robert Moffit, and the journals editors for their
insightful comments.
179
ENDNOTES
1. Personal Responsibility and Work Opportunity Reconciliation
Act of 1996 Public Law 104-193, Section 101. 110 Stat. 2110.
August 22, 1996.
2. See summary of states using Temporary Assistance for Needy
Families funds for family support services such as home visits for
new parents, and parenting skills training, on the Welfare Information Network Web site at http://www.welfareinfo.org/
FamilySupportServices.htm.
3. Maccoby, E. E. Parenting and its effects on children: On reading
and misreading behavior genetics. Annual Review of Psychology
(2000) 51:127.
4. Ainsworth, M.S. Infant-mother attachment. American Psychologist (1979) 34:93237.
5. Cassidy, J., and Shaver, P.R., eds. Handbook of attachment: Theory, research, and clinical applications. New York: Guilford Press,
1999.
6. Shumow, L., Vandell, D.L., and Posner, J.K. Harsh, firm, and
permissive parenting in low-income families: Relations to childrens academic achievement and behavioral adjustment. Journal
of Family Issues (1998) 19:483507.
18. Bandura, A. Social learning theory. Englewood Cliffs, NJ: Prentice-Hall, 1977.
19. Eisenberg, N. The caring child. Cambridge, MA: Harvard University Press, 1992.
20. Parke, R.D., Burks, V.M., Carson, J.L., et al. Family-peer relationships: A tripartite model. In Exploring family relationships
with other social contexts. R.D. Parke and S.G. Kellam, eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 1994, pp. 11546.
21. Cooper, C.R., and Cooper Jr., R.G. Links between adolescents
relationships with their parents and peers: Models, evidence, and
mechanisms. In Family-peer relationships: Modes of linkage. R.D.
Parke and G.W. Ladd, eds. Hillsdale, NJ: Lawrence Erlbaum
Associates, 1992, pp. 13558.
22. Furstenberg, F.F., Cook, P.D., Eccles, J., et al. Managing to
make it: Urban families and adolescent success. Chicago: University of Chicago Press, 1999.
7. Weiss, L.H., and Schwarz, J.C. The relationship between parenting types and older adolescents personality, academic achievement, adjustment, and substance use. Child Development (1996)
67:210114.
23. Eccles, J.S., and Barber, B.L. Student council, volunteering, basketball, or marching band: What kind of extracurricular involvement matters? Journal of Adolescent Research (1999) 14:1043.
8. Hartup, W.W. Social relationships and their developmental significance. American Psychologist (1989) 44:12026.
25. Boyce, W.T., Jensen, E.W., James, S.A., and Peacock, J.L. The
family routines inventory: Theoretical origins. Social Science
Medicine (1983) 17:193200.
9. Bornstein, M.H., ed. Handbook of parenting. Mahwah, NJ:

Lawrence Erlbaum Associates, 1995.
10. Capaldi, D.M., and Patterson, G.R. Relation of parental transitions to boys adjustment problems: I. A linear hypothesis: II.
Mothers at risk for transitions and unskilled parenting. Developmental Psychology (1991) 27:489504.
11. Taylor, R.D., and Roberts, D. Kinship support and maternal and
adolescent well-being in economically disadvantaged AfricanAmerican families. Child Development (1995) 66:158597.
12. Maccoby, E.E., and Martin, J.A. Socialization in the context of
the family: Parentchild interaction. In Handbook of child psychology. Vol. 4, Socialization, personality, and social development.
E.M. Hetherington, ed. New York: John Wiley & Sons, 1983,
pp. 1101.
13. Pittman, L.D., and Chase-Lansdale, P.L. African American adolescent girls in impoverished communities: Quality of parenting
and adolescent outcomes. Journal of Research on Adolescence
(June 2001) 11(2):199224.
14. Steinberg, L. We know some things: Parent-adolescent relations
in retrospect and prospect. Journal of Research on Adolescence
(2001) 11:119.
15. Bradley, R.H. Environment and parenting. In Handbook of parenting. Vol 2. M.H. Bornstein, ed. Mahwah, NJ: Lawrence Erlbaum Associates, 1995, pp. 23561.
16. Huttenlocher, J., Levine, S., and Vevea, J. Environmental input
and cognitive growth: A study using time-period comparisons.
180
17. Resnick, M.D., Bearman, P.S., Blum, R.W., et al. Protecting adolescents from harm: Findings from the National Longitudinal
Study on Adolescent Health. Journal of the American Medical
Association (1997) 278:82332.
24. See note 20, Parke, Burks, Carson, et al.
26. Wilson, W.J. The truly disadvantaged: The inner city, the underclass, and public policy. Chicago: University of Chicago Press,
1987.
27. Marshall, S. Ethnic socialization of African American children:
Implications for parenting, identity development, and academic
achievement. Journal of Youth and Adolescence (1996)
24:37796.
28. Carlson, C., Uppal, S., and Prosser, E.C. Ethnic differences in
processes contributing to self-esteem of early adolescent girls.
Journal of Early Adolescence (2000) 20:4467.
29. Hernandez, D.J., ed. Committee on the Health and Adjustment
of Immigrant Children and Families, National Research Council.
Children of immigrants: Health, adjustment, and public assistance. Washington, DC: National Academy Press, 1999.
30. Spencer, M.B. Cultural cognition and social cognition as identity
correlates of Black childrens personal-social development. In
Beginnings: The social and affective development of Black children.
M.B. Spencer, G. Kearse-Brookins, and W.R. Allen, eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 1985.
31. Quinton, D., and Rutter, M. Parenting breakdown: The making
and breaking of intergenerational links. Aldershot, England: Avebury, 1988.
32. Van Ijzendoorn, M.H. Intergenerational transmission of parenting: A review of studies in nonclinical populations. Developmental Review (1992) 12:7699.
33. Kelley, M.L., Power, T.G., and Wimbush, D.D. Determinants of
Volume 12, Number 1
disciplinary practices in low-income Black mothers. Child Development (1992) 63:57382.

34. Collins, W.A., Maccoby, E.E., Steinberg, L., et al. Contemporary
research on parenting: The case for nature and nurture. American Psychologist (2000) 55:21832.
35. Atkinson, L., Chisholm, V.C., Scott, B., et al. Maternal sensitivity, child functional level, and attachment in Downs syndrome.
Monographs of the Society for Research in Child Development
(1999) 64(3):4566.
37. Belsky, J. The determinants of parenting: A process model. Child
Development (1984) 55:8396.
38. The prevailing scientific view is that [nature and nurture] are
inextricably interwoven all along the pathway from birth to maturity, [and that science] should concentrate on the ways in which
[nature and nurture] jointly function. See note 3, Maccoby, p.
23. See also Reiss, D., with Neiderhiser, J.M., Hetherington,
E.M., and Plomin, R. The relationship code: Deciphering genetic
and social influences on adolescent development. Cambridge, MA:
Harvard University Press, 2000; OConnor, T.G., and Rutter, M.
Risk mechanisms in development: Some conceptual and methodological considerations. Developmental Psychology (1996)
32:78795; and see note 36, Shonkoff and Phillips, p. 40.
39. Ge, X., Conger, R., Cadoret, R., et al. The developmental interface between nature and nurture: A mutual influence model of
child antisocial behavior and parent behaviors. Developmental Psychology (1996) 32:57489.
40. Magnuson, K.A., and Duncan, G.J. Parents in poverty. In Handbook of parenting. Vol. 4, Social conditions and applied parenting.
2nd ed. M. Bornstein, ed. Mahwah, NJ: Lawrence Erlbaum
Associates, 2002.
41. McLoyd, V.C., Jayaratne, T.E., Ceballo, R., and Borquez, J.
Unemployment and work interruption among African American
single mothers: Effects on parenting and adolescent socioemotional functioning. Child Development (1994) 65:56289.
44. Avenevoli, S., Sessa, F.M., and Steinberg, L. Family structure,
parenting practices, and adolescent adjustment: An ecological
examination. In Coping with divorce, single parenting, and remarriage. E.M. Hetherington, ed. Mahwah, NJ: Lawrence Erlbaum
Associates, 1999, pp. 6592.
Press, 1994.
46. Downey, D.B. When bigger is not better: Family size, parental
resources, and childrens educational performance. American
Sociological Review (1995) 60:74661.
47. According to 1995 survey data, 2% of preschoolers and 9% of
school-age children were being cared for by siblings on a regular
basis. See Smith, K. Whos minding the kids? Child care arrangements: Fall 1995. Current Population Reports, Household Economic Studies, P70-70. Washington, DC: U.S. Census Bureau,
October 2000, pp. 4, 16.
48. McLoyd, V.C. The impact of economic hardship on Black families and children: Psychological distress, parenting, and socioemotional development. Child Development (1990) 61:31146.
49. Downey, G., and Coyne, J.C. Children of depressed parents: An
integrative review. Psychological Bulletin (1990) 108:5076.
50. Steele, R.G., Forehand, R., and Armistead, L. The role of family
processes and coping strategies in the relationship between
chronic illness and childhood internalizing problems. Journal of
Abnormal Child Psychology (1997) 25:8394.
51. Cox, M.J., Paley, B., and Harter, K. Interparental conflict and
parentchild relationships. In Interparental conflict and child
development: theory, research, and applications. J.H. Grych and
F.D. Fincham, eds. Cambridge, England: Cambridge University
Press, May 31, 2001.
52. Davies, P.T., and Cummings, M.E. Marital conflict and child
adjustment: An emotional security hypothesis. Psychological Bulletin (1994) 116:387411.
53. Friedman, R.J., and Chase-Lansdale, P.L. Chronic adversities. In
Child and adolescent psychiatry: Modern approaches. 4th ed. M.
Rutter and E. Taylor, eds. Oxford: Blackwell Science, April 30,
2002.
54. National Research Council. Understanding and preventing violence. Washington, DC: National Academy Press, 1993.
55. Burchinal, M.R., Follmer, A., and Bryant, D.M. The relations of
maternal social support and family structure with maternal
responsiveness and child outcomes among African American families. Developmental Psychology (1996) 32:107383.
56. Taylor, R.D., Casten, R., and Flickinger, S.M. Influence of kinship social support on the parenting experiences and psychosocial
adjustment of African-American adolescents. Developmental Psychology (1993) 29:38288.
57. Schuster, T.L., Kessler, R.C., and Aseltine, R.H. Supportive interactions, negative interactions, and depressed mood. American
Journal of Community Psychology (1991) 18:42338.
58. Burton, L.M., and Stack, C.B. Conscripting kin: Reflections on
family, generation, and culture. In The politics of pregnancy: Adolescent sexuality and public policy. A. Lawson and D.L. Rhode,
eds. New Haven: Yale University Press, 1993.
59. Simons, R.L., Lorenz, F.O., Wu, C., and Conger, R.D. Social
network and marital support as mediators and moderators of the
impact of stress and depression on parental behavior. Developmental Psychology (1993) 29:36881.
60. McLoyd, V.C., and Steinberg, L., eds. Studying minority adolescents: Conceptual, methodological, and theoretical issues. Mahwah,
NJ: Lawrence Erlbaum Associates, 1998.
61. Garcia Coll, C., and Magnuson, K. Cultural influences on the
child: Are we ready for a paradigm shift? In Cultural processes in
child development: The Minnesota symposia on child psychology, Vol.
29. A.S. Masten, ed. Mahwah, NJ: Lawrence Erlbaum Associates,
pp. 124.
181
62. Cowan, P.A., Powell, D.R., and Cowan, C.P. Parenting interventions: A family systems perspective. In Handbook of Child Psychology. Vol. 4, Child Psychology in Practice. 5th ed. W. Damon, I.E.
Siegel, and K.A. Renninger, eds. New York: John Wiley & Sons,
1998, pp. 372.
63. Gomby, D., Culross, P., and Behrman, R. Home visiting: Recent
program evaluationsAnalysis and recommendations. The Future
of Children (Spring/Summer 1999) 9(1):426.
64. Benasich, A.A., Brooks-Gunn, J., and Clewell, B.C. How do
mothers benefit from early intervention programs? Journal of
Applied Developmental Psychology (1992) 13:31162.
65. Halpern, R. Early childhood intervention for low-income children and families. In Handbook of Early Childhood Intervention.
2nd ed. J.P. Shonkoff and S.J. Meisels, eds. New York: Cambridge University Press, 2000, pp. 36186.
66. Brooks-Gunn, J., Berlin, L.J., and Fuligni, A.S. Early childhood
intervention programs: What about the family? In Handbook of
Early Childhood Intervention. 2nd ed. J.P. Shonkoff and S.J.
Meisels, eds. New York: Cambridge University Press, 2000,
pp. 54987.
67. Examples of such studies include Welfare, Children, and Families:
A Three-City Study (see http://www.jhu.edu/~welfare/); Urban
Change (see http://www.mdrc.org/WelfareReform/
UrbanChange.htm); Fragile Families (see http://crcw.princeton.edu/
fragilefamilies/); and Barriers to Employment and Family Functioning: Womens Employment Study (see Kalil, A., Dunifon,
R.E., and Danziger, S.K. Does maternal employment mandated
by welfare reform affect childrens behavior? In For better and for
G.J. Duncan and P.L. Chase-Lansdale, eds. New York: Russell
Sage Foundation, January 2002, pp. 15478.
68. Morris, P.A., Huston, A.C., Duncan, G.J., et al. How welfare and
work policies affect children: A synthesis of research. New York:
Manpower Demonstration Research Corp., 2001. These randomassignment studies provide a very strong basis for assessing the
impact of the reform packages. A key strength of these experimental demonstrations is that participants were randomly
assigned to a program or experimental group that received the
new welfare reform package or to a control group that continued
to live under the old AFDC regulations. In other words, any differences in parenting found between the participants in the
experimental versus the control groups are more likely to be due
to the impact of the welfare reform package than to preexisting
differences in the families.
69. Much of the available research focuses mostly on mothers, and
that is the focus in this article as well. See the article by McLanahan and Carlson in this journal issue for a more detailed discussion of fathers as parents.
70. This has been referred to as the stimulation hypothesis. See
Zaslow, M.J., Dion, M.R., Hair, E.C., et al. Program impacts on
parenting behavior during the first half-year of an educationallyfocused welfare-to-work program. Unpublished manuscript.
2000.
71. Parcel, T.L., and Menaghan, E.G. Parents jobs and childrens
lives. New York: Aldine de Gruyter, 1994.
72. Hoffman, L.W., and Youngblade, L.M. Mothers at work: Effects
on childrens well-being. Cambridge, England: Cambridge University Press, 1999.
73. Chase-Lansdale, P.L., and Duncan, G.J. Lessons learned. In For
better and for worse: Welfare reform and the well-being of children
and families. G.J. Duncan, and P.L. Chase-Lansdale, eds. New
182
York: Russell Sage Foundation, January 2002, pp. 30722.

74. Ricciuti, H.N. Single parenthood and school readiness in White,
Black, and Hispanic 6- and 7-year-olds. Journal of Family Psychology (1999) 13:45065.
75. Vandell, D.L., and Ramanan, J. Effects of early and recent maternal employment on children from low-income families. Child
Development (1992) 63:93849.
76. Three of the demonstration programs (New Chance, Teenage
Parent Demonstration, and the National Evaluation of Welfareto-Work Strategies) required mothers in the experimental groups
to participate in education, training, job readiness, and job search
activities; and four (New Hope, the Minnesota Family Investment Program, the Family Transition Program in Florida, and
the Self-Sufficiency Project in Canada) required mothers in the
experimental groups to work 30 hours per week. The Minnesota
program required mothers to work only 20 hours per week if
they had children age 6 or younger. For further details, see the
article by Zaslow and colleagues in this journal issue.
77. Quint, J.C., Bos, J.M., and Polit, D.F. New Chance: Final report
on a comprehensive program for young mothers in poverty and their
1997.
78. Morris, P., and Michalopoulos, C. The Self-Sufficiency Project at
36 months: Effects on children of a program that increased parental
employment and income. Ottawa, Ontario: Social Research and
Demonstration Corp., 2000.
79. Bos, J.M., Huston, A.C., Granger, R.C., et al. New Hope for people with low incomes: Two-year results of a program to reduce poverty and reform welfare. New York: Manpower Demonstration
Research Corp., 1999. Mothers in the experimental group indicated significantly higher levels of case management support than
those in the control group had.
80. Huston, A.C., Duncan, G.D., Granger, R., et al. Work-based
anti-poverty programs for parents can enhance the school performance and social behavior of children. Child Development
(2001) 72:31836.
81. See Zaslow, M.J., and Eldred, C.A., eds. Parenting behavior in a
sample of young mothers in poverty: Results of the New Chance
Observational Study. New York: Manpower Demonstration
Research Corp., 1998. See also note 77, Quint, et al.
82. See note 79, Bos, et al.
83. For a description of these programs, see Box 1 in the article by
Zaslow and colleagues in this journal issue.
84. These two demonstrations, New Hope and the Self-Sufficiency
Project, were the exceptions. Most of the demonstrations showed
no significant differences in the use of extended-day programs or
extracurricular activities. In one other demonstration (the Minnesota program), fewer children in the experimental program
were enrolled in lessons, clubs, and activities. See note 68, Morris, et al., for more details.
85. Blank, R.M., and Schmidt, L. Work, wages, and welfare. In The
new world of welfare. R.M. Blank and R. Haskins, eds. Washington, DC: Brookings Institution Press, 2001, pp. 70102.
86. Haskins, R. Effects of welfare reform at four years. In For better
and for worse: Welfare reform and the well-being of children and
families. G.J. Duncan and P.L. Chase-Lansdale, eds. New York:
Russell Sage Foundation, January 2002, pp. 26489.
87. Primus, W., and Daugirdas, K. Reducing child poverty by
improving the work-based safety net. In For better and for worse:
Volume 12, Number 1
Welfare reform and the well-being of children and families. G.J.

Duncan and P.L. Chase-Lansdale, eds. New York: Russell Sage
Foundation, January 2002, pp. 24763.
88. See the article by Zaslow and colleagues in this journal issue.
families, see Stanley, S.M., Markman, H.J., St. Peters, M., and
Leber, B.D. Strengthening marriages and preventing divorce:
New directions in prevention research. Family Relations: Journal
of Applied Family and Child Studies (1995) 44:392401.
89. Miller, C., Knox, V., Gennetian, L.A., et al. Final report on the
Minnesota Family Investment Program. Vol. 1, Effects on adults.
New York: Manpower Demonstration Research Corp., 2000.
103. Chase-Lansdale, P.L. Policies for stepfamilies: Crosswalking private and public domains. In Stepfamilies: Who benefits? Who does
not? A. Booth and J. Dunn, eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 1994.
90. Michalopoulos, C., Card, D., Gennetian, L.A., et al. The SelfSufficiency Project at 36 months: Effects of a financial work incentive on employment and income. Ottawa, Ontario: Social Research
and Demonstration Corp., 2000.
104. Ooms, T. Toward more perfect unions: Putting marriage on the

public agenda. Washington, DC: Family Impact Seminar, 1998.
See also note 99, Horn and Sawhill.
91. Miller, C., and Gennetian, L.A. Reforming welfare and work:
Final report on the Minnesota Family Investment Program. Vol. 2,
Effects on children. New York: Manpower Demonstration
105. Horvath-Rose, A., and Peters, H. E. Welfare waivers and nonmarital childbearing. In For better and for worse: Welfare reform
and the well-being of children and families. G.J. Duncan and P.L.
Chase-Lansdale, eds. New York: Russell Sage Foundation, January 2002, pp. 22246.
92. Also, in the New Hope demonstration, mothers who were

already working full time when randomly assigned to the program group showed better parenting (higher warmth and supervision) two years later than control group mothers did, and they
tended to decrease their work hours. See note 79, Bos, et al.
93. The study that found higher levels of problem behavior was
Canadas Self-Sufficiency Project. See note 78, Morris and
Michalopoulos.
94. Bloom, D., Kemple, J.J., Morris, P., et al. The Family Transition
Program: Final report on Floridas initial time-limited welfare
program. New York: Manpower Research Demonstration Corp.,
2000.
95. See note 1, U.S. House of Representatives, pp. 6, 9.
96. Waite, L.J., and Gallagher, M. The case for marriage: Why married people are happier, healthier, and better off financially. New
York: Doubleday, 2000.
97. Moore, M.R., and Chase-Lansdale, P. L. Sexual intercourse and
pregnancy among African-American adolescent girls in highpoverty neighborhoods: The role of family and perceived community environment. Journal of Marriage and the Family
(November 2001) 63(4):114657.
98. See note 91, Miller and Gennetian.
99. Horn, W.F., and Sawhill, I.V. Making room for daddy: Fathers,
marriage, and welfare reform. In The new world of welfare. R.M.
Blank and R. Haskins, eds. Washington, DC: Brookings Institution Press, 2001, pp. 42141. See also the articles by Greenberg
and colleagues and by McLanahan and Carlson in this journal
issue.
100. Markman, H.J., Floyd, F.J., Stanley, S.M., et al. Prevention of
marital distress: A longitudinal investigation. Journal of Consulting and Clinical Psychology (1988) 56:21017.
101. Cowan, C.P., and Cowan, P.A. When partners become parents:
The big change for couples. Mahwah, NJ: Lawrence Erlbaum
Associates, 2000.
106. Danziger, S. Work, income and material hardship after welfare

reform. Journal of Consumer Affairs (2000) 34(1):630.
107. Zedlewski, S.R., and Loprest, P. How well does TANF meet the
needs of the most disadvantaged families? In The new world of
welfare. R.M. Blank and R. Haskins, eds. Washington, DC:
Brookings Institution Press, 2001, pp. 31134.
108. Welfare reform: Moving hard-to-employ recipients into the workforce. GAO-01-368. Washington, DC: U.S. General Accounting
Office, March 2001.
109. The Teen Parent Demonstration. See Aber, J.L., Brooks-Gunn,
J., and Maynard, R. The effects of welfare reform on teenage
parents and their children. The Future of Children (Summer/Fall
1995) 5(2):5371.
110. Cherlin, A., Burton, L., Francis, J., et al. Sanctions and case closing for noncompliance: Who is affected and why. Policy Brief 01-1.
Welfare, Children, and Families: A Three-City Study. Baltimore,
MD: Johns Hopkins University, 2001. Available at
http://www.jhu.edu/~welfare/.
111. Kalil, A., and Seefelt, K. Economic and personal well-being among
sanctioned TANF recipients. Paper presented at the Fall Research
Conference of the Association for Public Policy Analysis and
Management. Seattle, WA, November 24, 2000.
112. Tweedie, J. Sanctions and exits: What states know about families
that leave welfare because of sanctions and time limits. In For
better and for worse: Welfare reform and the well-being of children
and families. G.J. Duncan and P.L. Chase-Lansdale, eds. New
York: Russell Sage Foundation, January 2002, pp. 81102.
113. Two studies that do include observational assessments are the
evaluations of the New Chance and JOBS programs. See note
81, Zaslow and Eldred; and Hair, E.C., McGroder, S.M.,
Zaslow, M.J., et al. How do maternal risk factors affect children
in low-income families? Further evidence of two-generational
implications. Journal of Prevention and Intervention in the Community. In press.
102. For a discussion of one study that does focus on low-income
183
APPENDIX
Innovative Uses of TANF Funds to Provide

Parenting Supports
Program name:
Teen Living Program
State:
Massachusetts

Coverage:
Statewide
Program goals:
Provide shelter and services for young

mothers and their children
Service population: Pregnant or parenting teens who receive TANF

Funding sources:
Description: The 1996 federal welfare reform law requires teen

mothers under age 18 to live with their parents (or in another adultsupervised setting) to receive TANF cash assistance. Teen mothers
who cannot live with their parents in Massachusetts have an alternativea second chance home. Twenty-one homes across the
state offer shelter and intensive support programs to 120 teen families. Each home has its own unique atmosphere, but all provide
two-generational programs for teen mothers and their children.
These programs not only assist with child care, education, and job
training, but also help the young women develop concrete parenting skills and the ability to advocate for their own and their childrens needs. The homes are popular with young mothers: There is
a list of teens waiting to enter the homes, and some teens stay voluntarily beyond age 18.
Results: The program has shown some positive results. According
to a report from the Social Policy Action Network, the young mothers in a similar program in New Mexico had a 2% rate of repeat
pregnancy. In contrast, teen mothers have a repeat pregnancy rate
of 50% within two years in the general population. Also, according
to a report from the Massachusetts Department of Social Services,
teen mothers participating in the program showed significant
improvement in parenting skills, especially in displaying positive,
nurturing interactions with their children, and in managing their
childrens health needs, such as keeping immunizations current.
The mothers also fared better than average in continuing education
and training, even after leaving the homes.
For further information: See the Massachusetts Department of
Social Services Web site at http://www.state.ma.us/dss/Teens/
AS_Assited_Living.htm; and the Social Policy Action Network Web
site at http://www.span-online.org/outcomes.html.
184
Volume 12, Number 1
Program name:
Nurse-Family Partnership
Program name:
Kinship and Family Empowerment Services
State:
New York
State:
Colorado
Coverage:
First begun in Elmira, New York, this program

has been replicated in 23 states
Coverage:
El Paso County
Program goals:
Prevent foster care placements
Program goals:
Decrease the percentage of children in lowincome families that experience physical, psychological, or emotional abuse
Service population: Low-income, first-time mothers

Funding sources:
Initially funded from research grants, this program is beginning to utilize funds from TANF,
Medicaid, and child abuse prevention funds.
TANF funds have been used to support the
program in six states.a
Description: Under the Nurse-Family Partnership program, registered nurses work intensively with first-time mothers during pregnancy and afterward until the child reaches age two. The program
builds confidence and skills in areas ranging from maternal and
child physical health to care giving, building and maintaining support networks, and achieving economic self-sufficiency. Because
the program relies on the establishment of a trusting relationship
between nurse and family, efforts are made to hire nurses who
share the same racial and/or ethnic background as the families they
serve. Over the past 20 years this program has been rigorously tested and refined. Those wishing to replicate the Nurse-Family Partnership program are required to follow a specific protocol that
preserves the original program design, and program implementers
are provided with specialized training and visit-by-visit guidelines.
Results: The Elmira, New York, program has been the subject of a
long-term clinical trial. Results show that children who participated
in the Nurse-Family Partnership program had 56% fewer visits to
the emergency department during their second year of life compared to children in a control group. In the third and fourth years of
life, participants had 40% fewer physician visits than the control
group. Over their first 15 years of life, children in the nurse-visitation program also had 54% fewer verified reports of child abuse
and/or neglect than did children in the control group.
For further information: See the National Center for Children, Families, and Communities Web site at http://www.nccfc.org.
aThe six states where local agencies have used TANF funds to support the program
are: California, Colorado, Kentucky, Ohio, Pennsylvania, and South Dakota.
Service population: Child-only TANF cases

Funding sources:
TANF
Description: In El Paso County, approximately 30% of TANF cases are

child-only cases. These children often live with a relative, generally a grandparent, who does not receive TANF cash assistance. A special unit funded by TANF supports these caregivers so children do not
have to enter the foster care system. This integrated unit, called the
Family Support Team, provides short-term family preservation services to families receiving TANF or other forms of public assistance as
part of an overall effort to better serve families with both child welfare and financial assistance needs. For many caregivers, this support is financial; some receive supplemental payments or one-time
emergency grants. Others receive help with services they may not
have been aware of, such as child care or health care. Still others
seek emotional support. A key element of kinship support is a bimonthly Grandparents raising Grandchildren group. Many of the
grandparent caregivers have lost their social network due to the
arrival of a toddler or a teen in their lives and are parenting in isolation. The support groups not only combat that isolation, but also help
prevent crisis situations by educating grandparents about the developmental and emotional needs of the children in their care. Many of
these grandparents want to parent differently the second time
around. As they grieve the loss of their idyllic grandparent role, they
also learn concrete skills for helping their grandchildren heal from
past parental neglect or abuse. In addition to the support groups, the
program also sponsors biannual community events attended by hundreds of caregivers. These events honor the role caregivers play, and
offer workshops on topics such as age-specific parenting skills, legal
issues for caregivers, and dealing with media violence and safety
issues around the home.
Results: The department is just beginning an evaluation of these
new services, launched in July 1997, but thus far, the program
appears to be achieving its goal. Of the 450 kinship families identified, only two have had a child placed in foster care.
For further information: See the El Paso County Department of
Human Services Web site at http://www.co.el-paso.co.us/humansvc/
keepingfamiliestogether.asp.
185
Five Commentaries:
Looking to the Future
COMMENTARY 1
Wendell E. Primus
ince welfare reform was enacted, caseloads have

fallen dramatically, single mothers earnings
and employment have increased, and the child
poverty rate has fallen. These statistics do not
tell the whole story, however. As the reauthorization
debates unfold in 2002 around child care and the welfare block grant (Temporary Assistance for Needy
Families, or TANF), one of the key criteria for measurWendell E. Primus, Ph.D., is director of income security at the Center on Budget and Policy Priorities, a
nonpartisan research organization and policy institute
conducting research and analysis on a range of government policies affecting low- and moderate-income
people.
COMMENTARIES
To provide an array of perspectives about policy changes to benefit children

that might be considered during reauthorization of the federal welfare reform
law in 2002, we asked five experts across various disciplines and backgrounds
to respond to this question: How can supports and services for low-income
families be improved under welfare reform to help ensure positive outcomes for
children? Their responses follow.
ing the success of the changes implemented under welfare reform should be whether the well-being of children and families has improved.
Welfare reform has coincided with the longest-running
economic expansion in our nations history. Between
1993 and 2000, average annual unemployment fell
from 7% to 4%.1 At the same time, hourly wage rates
for the lowest-paid workers rose, and the incentive to
take a low-wage job increased substantially with expansion of the earned income tax credit (EITC). One
would expect earnings to increase and child poverty to
decline under these conditions. In 1999, according to
a measure of poverty that includes noncash benefits
and taxes, the child poverty rate fell to 13%, the lowest
level since this more comprehensive poverty measure
became available in 1979.2
These positive trends come with a few caveats, however. First, although the number of poor children has fallen markedly in recent years, U.S. Census Bureau data
show that those who remain poor have, on average,
187
grown poorer when the effects of tax and benefit programs are taken into account.2 The poverty gap for all
families with children has improved only slightly,
despite the significant increase in earnings. Second, the
disposable income of the poorest fifth of single mothers living just with their children (and no other adults)
fell 8% between 1995 and 1999, despite increased
earnings, largely because of the loss of cash welfare
assistance.3 Furthermore, approximately 725,000 independent single-mother families were worse off in 1999
than were such families in 1995. In the past, welfare
programs played a pivotal role in reducing the poverty
gap. Now, most researchers agree that some families,
especially those who left welfare without entering the
labor force, are floundering.4
COMMENTARIES
Finally, how well the new welfare programs will perform in a faltering economy remains an open question.
Will states allow their caseloads to increase as more parents find themselves unemployed? Or will states institute policies that make it difficult for families to come
onto the rolls? Will the federal government ensure that
states have adequate resources to meet the increased
need? With the economy now in recession, both the
federal government and the states will need to take
steps to ensure that a safety net is in place when parentsmany of whom have only recently begun to
worklose their jobs.
Thus, though welfare reform clearly has resulted in
positive changes for some current and former recipients, more work remains to shore up the safety net. I
propose the following action items to improve TANF
and related programs during reauthorization in 2002.
Increase funding substantially and ensure that

states do not supplant state funds
There are several reasons for increasing the size of the
TANF block grant. At a minimum, the federal TANF
and state maintenance-of-effort funds should be
adjusted for inflation. Additional funds will also be
necessary to increase parity in the resources available to
states. Currently, wealthier states have significantly
more TANF dollars per poor child than poorer states
have. TANF established supplemental grants to poorer
states to move toward parity; more funds should be
made available to continue to move toward this goal.
188
Primus
Many policymakers point to reduced caseloads as a reason to cut TANF funding. But TANF funds have a
much broader purpose than providing cash assistance for
very poor families. In fact, in fiscal year 2000, less than
40% of all TANF expenditures were for basic assistance.
TANF funds are used to assist parents in preparing for,
finding, and retaining jobs; to fund child care and transportation services for working families and families participating in welfare-to-work programs; to support teen
pregnancy prevention efforts; to operate subsidized job
programs for parents who cannot find regular employment; and much more. The demand for these important
services is not tied to the size of a states cash assistance
caseload. In addition, a broader range of TANF services
should be targeted to two-parent families and noncustodial parents, as discussed further below.
Welfare reform has achieved a number of successes;
states should not be rewarded for these successes by
having their funds cut, in either real or inflation-adjusted terms. At a minimum, the TANF grant should not
be decreased until child poverty is significantly reduced
and low-income parents are working to their maximum
ability. With increased funding, however, states must
be more accountable for how they use their TANF
block grant funds. Steps should be taken to ensure that
TANF funds are used as intended and do not supplant
existing state funds for programs benefiting lowincome families.
Volume 12, Number 1
Change the laws central focus from caseload

reduction to poverty reduction
Maintain focus on TANF adults with multiple and

significant employment barriers
Provide or restore cash and other benefits to

legal immigrants on the same basis as citizens
According to an Urban Institute study, 44% of welfare

recipients reported at least two significant obstacles to
work, such as limited education, no recent work experience, language barriers, mental or physical health problems and disabilities, or lack of transportation or child
care, and 17% revealed three or more obstacles.6 The percentage of recipients reporting no work activity increased
steadily with the number of significant obstacles.
The 1996 welfare law placed far-reaching restrictions

on legal immigrants eligibility for TANF, Medicaid,
food stamps, and Social Security Income (SSI). There
is now strong evidence that these restrictions have had
an adverse impact on many legal immigrants and their
citizen children. Research shows that food insecurity
increased significantly between 1994 and 1998 among
the immigrant-headed households most likely to be
affected by the restrictions.7 Food stamp participation
among citizen children living with immigrant parents
fell by 74% between 1994 and 1998, compared with a
24% decline among other families with children.8
Generally, if appropriate services and accommodations

are available, then many parents with multiple or significant barriers to employment should be able to succeed in the workplace. For parents to accomplish this
goal, however, states must consider modifying their
TANF programs to address the reality that parents with
multiple or significant employment barriers may need
additional services and continued assistance for longer
periods of time.
Revisit the federal time limit on cash assistance

Two aspects of the federal time limit need to be reexamined. First, a disconnect exists between the federal
time limit and policies that encourage welfare recipients to work, including work participation requirements and policies that make work pay (such as
provisions that allow welfare recipients to keep more
of what they earn without changing their eligibility or
benefit levels, referred to as earnings disregards).
But recipients who work while receiving cash assistance risk triggering the time limit and becoming
ineligible for benefits they may need in the future.
The time limit clock should be stopped while families
are working part time but not earning enough to support themselves.
COMMENTARIES
Reducing child and family poverty should be a principal purpose of welfare. Increases in family income are
key to positive child outcomes.5 States should be
required to explain how they will use block grant funds
to address this goal, and their success in doing so
should be measured.
Second, the number of families that can be exempted

from time limits should be increased. The 1996 welfare
law allows states to exempt 20% of their caseloads from
the time limit. When the law was created, this equaled
about 1 million families. Because caseloads have fallen
so dramatically, however, states may now exempt only
half as many people. More generous exemption policies should also be considered during periods of rising
unemployment and for nonparent caretakers.
As taxpayers, immigrants contribute not only to the

funding of education, roads, and national defense, but
also to safety net benefits for low-income families. They
should not be excluded from these programs when
temporary hardship interrupts their employment. Food
stamp and SSI benefits should be restored for legal
immigrants. The restrictions on states flexibility to provide federally funded TANF and Medicaid benefits to
recently arrived immigrants should be lifted.
Provide services and benefits to strengthen

two-parent families and help fragile families
stay together
Recent research indicates that about half of children
born to single parents actually live with both of their
biological parents at birth; however, as time goes on,
these fragile families tend to separate. (See the article by McLanahan and Carlson in this journal issue.)
These families need additional support to help the parents stay together and escape poverty. One concrete
way to do this is to ensure that two-parent families are
189
Primus
COMMENTARIES
accessing the benefits they are eligible to receive. Currently, two-parent families participate in food stamps,
Medicaid, and cash welfare assistance at a much lower
rate than do single-parent families, even when their
incomes are similar. According to data compiled by the
Urban Institute, 40% of single-parent families with
incomes below the federal poverty line receive TANF,
compared with only 10% of two-parent families with
incomes below the poverty line.9 Among families with
incomes below 50% of the poverty line, almost half of
single-parent families receive TANF, compared with
only 13% of two-parent families. States should eliminate any barriers or eligibility restrictions that cause
two-parent families to be served at much lower rates
than are single-parent families. States should also take
steps to reach out and serve a much larger proportion
of eligible two-parent families.
Assist noncustodial parents in providing for their

children both financially and emotionally
Child support can constitute a large part of families
budgets. For poor families that receive child support, it
makes up more than one-quarter of their annual
income, on average.10 Unfortunately, many eligible
families do not receive child support, and most families
do not get the full amount due to them. Noncustodial parents are a heterogeneous group; there are many
reasons why they may not pay child support. Many
low-income noncustodial parents do not pay child support regularly because they are unemployed or underemployed, and have limited incomes.
Many low-income families also are frustrated because
when noncustodial parents pay child support while the
custodial parent is receiving welfare, the government
retains nearly all of the support as reimbursement for
welfare costs. Little, if any, reaches the children for
whom it was intended. (See the article by Greenberg
and colleagues in this journal issue.) These rules need
to change. When a noncustodial parent pays child support, the children should get that support, and the custodial family should benefit financially.
Earnings subsidies (like the EITC) and welfare reform
encouraged more low-income mothers to enter the
workplace so they could better support their children.
The next step should be to help poor fathers become
190
employed, and to address their difficulties with the child

support system (including child support orders that are
high relative to low-income fathers earnings, large
child support debts, and economic disincentives to pay
child support). Policymakers should also consider providing government-subsidized matching payments (an
earnings-like subsidy) when the child support paid by
low-income dads is inadequate. Fathers need these supports so that they, too, can take more financial and
emotional responsibility for their children. At the same
time, the services provided to noncustodial parents
should not be more generous than, or come at the
expense of, programs for low-income custodial parents.
Increase participation in work support programs

among low-income families
In addition to improving the safety net for the poorest
families, policymakers must consider the many working
families that continue to have incomes at or below the
poverty line. Many of these families do not receive the
benefits for which they are eligible, such as food stamps
and Medicaid. Shrinking caseloads in these programs
would not be so troublesome if they were the result of
decreasing numbers of low-income families. Unfortunately, this is not the casethe declines in food stamp
receipt, for example, have significantly exceeded
declines in poverty. (See the article by Zedlewski in this
journal issue.)
Research indicates that the decline in welfare caseloads
may be driving down participation in other benefit
programsthat is, when families are not receiving cash
assistance, they may be unaware of their eligibility for
food stamps, Medicaid, child care assistance, and other
supports. For example, as discussed in the article by
Fuller and colleagues in this journal issue, the application process for child care subsidies must be made easier, and the availability of subsidies more widely
publicized. Additional housing vouchers for lowincome working families should be provided, and
health insurance programs for children should be
expanded to provide coverage for parents, which has
been shown to be key in efforts to increase the number
of low-income children protected.11 Reforms in these
programs should be judged by their ability to increase
participation among the working poor.
Volume 12, Number 1
Improve the EITC phaseout to support low-wage

work more effectively
A higher benefit level for families with three or more
children should also be added to the EITC. Currently
there are two tiers, or maximum benefit levels, in the
EITC: one level for families with one child, and a higher level for families with two children. Families with
three or more children have higher poverty rates than
do smaller families, and they have experienced more
difficulty moving from welfare to work. Adding a third
tier to the EITC would help address these problems.
In addition, the recently enacted tax cut made the child
tax credit partially refundable. Families receive a partial
credit (10% now, but increasing to 15% in 2005) on
their earnings over $10,000. However, for families

with full-time, year-round work at low wages, this provides little assistance. The $10,000 threshold should be
lowered to 0, or at least to $5,000.
In conclusion, TANF reauthorization provides an
opportunity to refocus attention on the well-being of
children and families. Building on the successes already
achieved, we can now focus on efforts to strengthen
supports for families that have left welfare for work and
help them climb the economic ladder, and on efforts to
provide additional resources and opportunities for families whose well-being has not improved under the first
round of welfare reform.
COMMENTARIES
ENDNOTES
1. Statistics from the U.S. Department of Labor Web site at
http://www.dol.gov.
7. Borjas, G.J. Food insecurity and public assistance. Unpublished

paper. Cambridge, MA: Harvard University, May 2001.
2. Center on Budget and Policy Priorities. Poverty and income

trends: 1999. Washington, DC: CBPP, November 2000.
8. Unpublished tabulations of food stamp quality control data, 1994

and 1998, prepared by Center on Budget and Policy Priorities.
3. Center on Budget and Policy Priorities, forthcoming analysis.
9. Beeson, J., and Primus, W. Safety net programs, marriage and

cohabitation. In Just living together: Implications of cohabitation
on families, children, and social policy. A. Booth and A Cronter,
eds. Mahwah, NJ: Lawrence Erlbaum Associates. In press.
4. Blank, R., and Haskins, R. Welfare reform: An agenda for reauthorization. In The new world of welfare. R. Blank and R. Haskins, eds. Washington, DC: Brookings Institution Press, 2001,
pp. 334.
5. Sherman, A. How children fare in welfare experiments appears to
hinge on income. Washington, DC: Childrens Defense Fund,
August 2001.
6. Zedlewski, S. Work activity and obstacles to work among TANF
recipients. Washington, DC: Urban Institute, September 1999.
against poverty. Washington, DC: Urban Institute, March 2000.
11. Ku, L., and Broaddus, M. The importance of family-based insurance expansions: New research findings about state health reforms.
Washington, DC: Center on Budget and Policy Priorities, September 2000.
191
Haskins
COMMENTARY 2
Ron Haskins
n the spring of 2002, congressional attention will

focus on reauthorization of the landmark welfare
reform legislation of 1996. Logically enough,
many scholars, policymakers, and foundation
executives interested in child development hope to
focus attention on children during the reauthorization
debate. More specifically, as some of the articles in this
volume show, these child advocates hope Congress will
fund projects that improve child care and initiate more
and better programs for children.
COMMENTARIES
I question this course of action. Among both scholars

and policy analysts, there is a divide between those who
think the nation can improve child development by
expanding federal programs and spending more on
federal programs for children, and those who want to
leave issues of child development primarily to families
and state and local governments. I am in the latter
group. In this brief commentary, I outline what I
believe will and should be the major issues in the reauthorization debateand none focus specifically on
children. Yet in the long run, satisfactory resolution of
these issues will have positive consequences for the
health and development of the nations children.
As in the welfare reform debate of 199596, work will
be the central issue of reauthorization. Promoting
work has been an ostensible goal of welfare reform
since at least the 1960s. However, a look beyond the
rhetoric to the text of the actual laws passed to fight the
war on poverty reveals that welfare programs, especially Aid to Families with Dependent Children (AFDC),
were based primarily on entitlement to benefits. The
foremost goal of U.S. social policy was not to promote
Ron Haskins, Ph.D., is senior fellow at the Brookings
Institution in Washington, D.C.; senior consultant at
the Annie E. Casey Foundation in Baltimore, Maryland; and former staff director for the Subcommittee
on Human Resources, Committee on Ways and
Means, U.S. House of Representatives.
192
work and self-reliance, but to provide guaranteed benefits. On the eve of the great welfare debate of
199596, the statutes guaranteed eligible individuals
cash, food stamps, child nutrition, school lunch, and
Medicaid, among other benefits.1 The package of entitlement benefits alone was worth more than $13,000
in the typical state, and total benefits worth $16,000 or
$17,000 for a single family were not unusual.2
The heart of the debate that culminated in the 1996
reforms was whether the entitlement to benefits could
be put at risk to enforce the duty to work. Above all
else, nonwork was the problem that Republicans aimed
to fix with the 1996 reforms.3 The fix, in their view,
required ending entitlement to AFDC cash payments,
and mandating states to place a rising percentage of
welfare recipients in jobs or job programs. These
changes were at the core of the new block grant to
states, Temporary Assistance for Needy Families
(TANF), which replaced AFDC.
What has happened since? Welfare programs across the
nation have been transformed. Programs that had been
committed almost exclusively to determining eligibility
and getting the benefit checks right are now almost
equally committed to transitioning families from welfare to work. Now when clients appear in the welfare
office, they are told that they need to work, and that
the welfare program is just the right place to find help
in preparing for and finding a job. The results have
been remarkable. About 2.5 million fewer families are
getting cash welfare, a magnitude of decline that no
welfare program has ever achieved before. The level of
work among single mothers, and especially never-married single mothers, is at an all-time high.4 They now
receive more income from work and less from welfare,5
and their total income has increased substantially since
1993. In addition, child poverty has dropped quickly,
reaching its lowest level in 20 years. In short, welfare
reform has been a rousing success, arguably achieving
its goals to a greater extent than any other federal social
policy of recent decades.
Now on the eve of the reauthorization debate, the
question arises: What should Congress do next? The
answer is succinct: reenact the original legislation with
only a few modest changes. As outlined in the article
by Greenberg and colleagues in this journal issue,
Volume 12, Number 1
headed families continue to apply for welfare, there is the

potential for widespread welfare dependency. Unless
Congress wants to return to the days when public policy permitted, even encouraged, the increase of dependency (and perhaps of nonmarital births), the current
level of TANF funding is needed to ensure that states
can maintain and expand their work programs.
many issues will receive attention during the reauthorization debate, but most of these issues canand I
predict willbe left to states and will not result in
statutory changes at the federal level. Still, four issues
are of such great importance that they should be major
topics of the reauthorization debate: the TANF funding level, innovative work programs, work programs
for poor fathers, and programs to reduce nonmarital
births and promote marriage.
Beyond helping mothers on welfare find jobs, states

also are investing in new types of innovative work programs to help low-income mothers win the battle for
self-support. Many of the mothers who have left welfare are struggling along with annual incomes in the
$12,000 to $15,000 range, based on earnings and
income supplements like the earned income tax credit.
This income level is better than welfare, but doubtless
many, if not most, of these mothers could do better.
COMMENTARIES
Innovative Work Programs
One of the biggest problems is that most mothers at

the bottom of the income distribution may take jobs
but then quickly lose them.6 If they do not work very
much during the year, their earnings rise less than their
welfare benefits fall, and they are worse off. Thus,
states are beginning to develop innovative programs to
help mothers retain jobs longer and/or quickly find
new jobs. If Congress reduces the TANF block grant,
states will be forced to cut back on these programs,
with direct consequences for mothers and children and
an inevitable increase in welfare dependency.
The TANF Funding Level

Unfortunately, cutting the $16.5 billion TANF block
grant may well be the biggest issue in the reauthorization debate. Powerful members of Congress on the
budget and appropriation committees will argue that
since the welfare rolls have declined by half, states need
less money to maintain benefits. This argument is deeply
flawed, however, and betrays a basic misunderstanding
of the nature of welfare reform. If the welfare program
still simply distributed cash benefits, the argument for
funding cuts would be sound. But welfare is no longer
whollyor even primarilya program of cash benefits.
Rather, it is now the most successful employment program the federal government has ever mounted. And as
long as a substantial number of poor, largely female-
Also, many mothers now working for around $7 per

hour are in all likelihood capable of earning much more,
but they need additional education and training. The
education and training programs implemented before
welfare reform focused primarily on high school
dropouts, and did little to move welfare mothers into
jobs.7 Now a new approach is taking hold: First help
these young mothers get jobs to show them the importance of education and training, then provide them with
opportunities to advance their skills, often while they
continue working. States are beginning to develop precisely this type of program, but again, if Congress
reduces TANF funding, states will not be able to both
maintain benefits and simultaneously continue to develop and run these job advancement programs.
193
Haskins
Work Programs for Poor Fathers
COMMENTARIES
In addition to helping mothers, states need to create

job programs for fathers of children on welfare. For
males with a high school education or less, recent
decades have been disastrous. Not only have their average wages declined, but the amount of time they spend
working or searching for work has dropped as well.8
Even as the economy added a net of 20 million jobs
during the spectacular boom of the 1990s, work effort
by minority men declined. By 1999, the employment
rate of black women ages 20 to 24 actually exceeded
the rate for their male counterparts, a first for any age
or ethnic group in the nations history. Although our
knowledge about how to help males improve their
employment is modest, the problem is so great that, at
a minimum, states should be encouraged or required
to launch programs that help poor fathers prepare for,
find, and retain employment.
Programs to Reduce Nonmarital Births and Promote Marriage

The 1996 welfare reform legislation contained more
than a dozen provisions designed to reduce births outside of marriage, a clear indication that Congress and
the President were greatly concerned about nonmarital
births as a major cause of many of the nations social
problems, poverty chief among them. Indeed, the
nation appears to be making modest progress in reducing births outside of marriage. The most hopeful sign is
that the teen birth rate has declined every year since
1991, and is now almost identical to its all-time low
recorded in 1986.9 Data on the overall percentage of
births outside of marriage are less impressive, but hopeful nonetheless. After rising relentlessly for several
decades, the nonmarital birth rate seems to have leveled
off at just under 33%, where it has remained for five
years.10 As discussed in the article by McLanahan and
Carlson in this journal issue, one of the major reasons
for the explosion of nonmarital births in recent decades
is the decline of marriage. Between 1950 and 1998, the
percentage of women over age 14 who are married
dropped from 66% to 55%, and among black women,
the percentage plummeted from 62% to only 36%.11
Although Congress placed great emphasis on reducing
nonmarital births in the 1996 reforms, little informa-
194
tion is available about the impacts of these provisions.

A major evaluation is in place for the abstinence education program, but it will be several years before
results are available.12 One of the most important and
expensive provisions of the 1996 legislation was the
$25 million per year allocated for bonuses to states that
reduced out-of-wedlock births. So far, 10 states have
won these bonuses, but it is unclear whether state
action, or simply demographics, was driving the
declines in nonmarital births.13 For example, the District of Columbia has won the bonus twice, but the
most significant factor contributing to this success
appears to be that the nonmarital birth rate has been
declining faster among blacks than any other group,
and Washington, D.C., has the highest percentage of
blacks nationwide.
The upcoming debate over nonmarital births and marriage will not be enlightened much by results from evaluation studies. Nevertheless, the magnitude of these
problems requires some response. Every state already
has some programs aimed at reducing nonmarital
births, but few have programs that promote marriage.
Programs in both these areas need to be expanded.
The Federal Role

The most important item on the welfare reform reauthorization agenda should be for Congress to keep full
funding for TANF so that states can maintain and
expand the work programs put in place since reform.
Over the past five years, these programs have enabled
our nation to achieve perhaps its greatest policy success
ever for programs targeted to low-income families,
helping about two million poor mothers join the labor
force and, in many cases, bring themselves and their
children above the poverty line.
Beyond the call for full funding, however, I propose
only three modest changes to federal statutory provisions to address the issues outlined above. Specifically,
Congress should encourage or require states to:
(1) Develop and conduct both job retention and job
advancement programs;
(2) Create job programs for fathers of children on welfare; and
(3) Design and fund programs to reduce nonmarital
Volume 12, Number 1
births and promote marriage among families with

children on welfare or at risk of going on welfare.
In addition to the bonuses for reducing out-ofwedlock births, the TANF performance bonuses
should reward states that increase marriage among
low-income families.
The next steps for welfare reform should be to create
programs to help mothers stay employed on a more
regular basis, and to help them get the experience, education, and training they need to secure better jobs. The
major source of action in designing and implementing
such programs is no longer at the federal level, however. Rather, the proper role of the federal government is
to finance programs, set broad goals, insure financial

accountability, and support program evaluation. It is
also a proper and useful role for the federal government
to stimulate the further growth of programs aimed at
helping poor fathers become better parents and
providers, reducing nonmarital births, and increasing
and supporting marriage, especially among poor parents. If advocates on both the left and the right concentrate their fire on full funding for TANF and on
these few issues that everyone agrees are central to the
nations poverty problem, then children, families, and
taxpayers will all be better off in the years ahead.
COMMENTARIES
ENDNOTES
1. Many families also received nonentitlement benefits such as housing, education, job training, and energy assistance. U.S. House of
Representatives, Committee on Ways and Means. 2000 green book:
Background material and data on programs within the jurisdiction
of the Committee on Ways and Means. Washington, DC: U.S. Government Printing Office, 2000, appendix K.
2. Calculations based on information provided by the Congressional
Research Service.
3. Years of experience with state-designed work programs had shown
that most individuals on welfare did not respond voluntarily when
called to work and that states did not aggressively pursue work
policies. See Mead, L. The new politics of poverty: The nonworking
poor in America. New York: Basic Books, 1992.
4. Around 60% of the adults leaving welfare have a job at any given
moment, and over a period of months about 80% of them work at
least part of the time. See note 1, U.S. House of Representatives,
Committee on Ways and Means, Appendix L.
5. Haskins, R. Effects of welfare reform on family income and poverty. In The new world of welfare. R. Blank and R. Haskins, eds.
Washington, DC: Brookings Institution Press, 2001, pp. 10336.
6. Herr, T., and Wagner, S.L. Moving from welfare to work as part of
a group: How pathways makes caseload connections. Chicago: Erikson Institute, 1998.
7. See note 1, U.S. House of Representatives. Committee on Ways
and Means, appendix L.
8. Blank, R. It takes a nation: A new agenda for fighting poverty.
New York: Russell Sage Foundation, 1997.
9. Ventura, S.J., Curtin, S.C., and Mathews, T.J. Variations in

teenage birth rates, 19911998: National and state trends.
National Vital Statistics Reports. Vol. 48, no. 6. Washington, DC:
National Center for Health Statistics, 2000. This decline is primarily the result of declines in teen sexual activity and increases in
use of contraception. See Child Trends. Trends in sexual activity
and contraceptive use among teens. Washington, DC: Child
Trends, March 2000.
10. Curtin, S.C., and Martin, J.A. Births: Preliminary data for 1999.
National Vital Statistics Reports. Vol. 48, no. 14. Washington,
DC: National Center for Health Statistics, 2000; and Martin,
J.A., Smith, B.L., Mathews, T.J., and Ventura, S.J. Births and
deaths: Preliminary data for 1998. National Vital Statistics
Reports. Vol. 4, no. 25. Washington, DC: National Center for
Health Statistics, 1999.
11. Bachrach, C.A. The changing circumstances of marriage and fertility in the United States. In Welfare, the family, and reproductive
behavior. R.A. Moffitt, ed. Washington, DC: National Research
Council, 1998. See also U.S. Census Bureau. Marital status and
living arrangements: March 1998 (Update). Current Population
Reports. Series P-20-514. Washington, DC: U.S. Census Bureau,
March 1998.
12. Maynard, R. Surveys and supporting documentation for the teen
activities and attitudes study. Princeton, NJ: Mathematica Policy
Research, 2000.
13. Haskins, R. Welfare reform and illegitimacy: Republican policies
designed to reduce illegitimacy. In Measuring family dynamics for
welfare reform: Data needs. D.J. Besharov, ed. College Park, MD:
University of Maryland School of Public Affairs. In press.
195
Guerra
Commentary 3
COMMENTARY 3
Fernando A. Guerra
B
COMMENTARIES
efore defining needed supports and services

for low-income families under welfare
reform, it is important to remind ourselves of
the complex circumstances that affect the
daily lives of many poor children and their families.
Many struggle with low income, financial insecurity,
limited job opportunities, marginal living conditions,
lack of transportation, and social inequality. The supports and services provided by programs such as Temporary Assistance for Needy Families (TANF) and
Medicaid can be extremely important to these families,
yet may not be enough to lessen the impact of poverty
and promote positive outcomes for children. The
inability of programs to adapt to families special circumstances or to respond when families are in crisis situations seriously limits their effectiveness. In addition,
the barriers restricting access for poor immigrant families limit the effectiveness of these programs even further. Stories from the lives of some of our nations
poorest families can help to inform the debate concerning reauthorization of the federal welfare reform
law in 2002, and encourage policymakers to look
beyond the numbers of families leaving the welfare
rolls and improve the way supports and services are
provided to our nations poor.
The Precariousness of
Poor Families Lives
For many poor children and their families, there is a
precariousness to their daily lives beyond what may be
apparent on the surface. My observations over the past
few years have convinced me that welfare reform is not
working in the best interests of these families. Moreover, some of the negative impacts appear to be cumulative and far-reaching across groups in a community,
region, or population. Brief descriptions of the circumFernando A. Guerra, M.D., M.P.H., is director of
health at the San Antonio Metropolitan Health District and clinical professor in the department of pediatrics at the University of Texas, San Antonio.
196
stances of several poor families receiving medical services in a primary care clinic where I work call attention
to the difficulties these families encounter as they
struggle to cope with the current welfare system.1
A young mother on welfare, with a high school
equivalency certificate, has recently completed a job
training program and is transitioning to work as a
receiving clerk in a laundry. She has two young children, 6 months and 17 months of age. The child care
center staff tell her not to bring her children back to
the center until she gets clearance from a physician
because the younger of the two had a loose stool
on one occasion during the day. The mother has
already missed work three times this month because
of her childrens illnesses, and her employer has told
her that if she misses work again, she may lose her
job. This young mother is overwhelmed with the
warning from work, the demands of child care, and
the need to miss work for what is, at most, nothing
more than a minor illness.
While still in the hospital, a 17-year-old mother of a
newborn infant pleads desperately with a social worker for assistance with housing. The teen mother and
her baby cannot stay with relatives because the familys home is already overcrowded. The infant and
mother are scheduled for early discharge from the
hospital, which only adds to the mothers difficulties.
An immigrant family from Mexico has a U.S.-born
son with irreversible neurological problems resulting
from trauma at birth. The child requires constant
attention and home-based services. The family is facing loss of Medicaid benefits, as well as food stamps
and utility assistance, because the fathers annual
earnings from his produce business are $2,000 above
the maximum income allowed.
A functionally illiterate single father with a thirdgrade education has two young school-age sons. The
boys are disruptive and unable to settle into a classroom. The family lives in an unsafe, drug-infested
public housing unit, which contributes to the childrens behavioral problems. The father would like to
participate in a job training program to help him earn
more than he receives from TANF, but he cannot
find a suitable program because he lacks the ability to
Volume 12, Number 1
read and other basic skills. Without access to public

transportation, the father transports his children and
himself on a bicycle.
A stressed, worried, depressed grandmother in her
sixties has taken her three adolescent grandchildren
into her home because their mother is mentally ill.
The grandmothers diabetes and high blood pressure
are out of control, making it almost impossible for
her to cope with the daily needs of young, energetic
adolescents, much less help her daughter.
A 22-year-old mother has just given birth to her
fourth child. She had limited prenatal care and a history of tobacco use, and has admitted to using
cocaine regularly during pregnancy. She recently tested positive for hepatitis C. She would like to get into
a treatment program for drug-dependent mothers
because she worries about losing her children to foster care, but no treatment program vacancies are
available.
Since welfare reform was enacted in 1996, many former welfare families have been able to transition to
steady work with the help of a strong economy and
interim supports such as Medicaid, food stamps, subsidized housing, and child care assistance.24 However,
many other families, like those described above, have
not been able to make this transition. Perhaps there
always have been some families experiencing such

problems, especially in large urban centers, but their
numbers and the complexity of their problems seem to
have increased. Targeted supports over longer periods
of time may be required to offset the profound cumulative impacts on these children and their families. For
some, the inability to cope may be so serious that it is
easier to continue providing welfare than to try helping
them overcome their problems.5,6 For others, such as
those in smaller or more rural communities, a lack of
opportunities for jobs, education, literacy, child care,
and access to basic physical and mental health services
presents an even greater challenge.7,8 Finally, poor families of color, families with different cultures or languages, and those with disabilities quite often are
unable to access even the smallest of safety nets.9
The Limitations of the Current

Welfare System
COMMENTARIES
1997, Stephen Shames/Matrix
The family stories shared above illustrate the limitations of the current welfare system. All of these families
are just barely getting by. All are struggling to provide
the basics for their children: food, housing, utilities,
transportation, child care, after-school and summer
programs, periodic medical exams, dental exams, and
immunizations. In addition to dealing with these challenges, all of the families must complete voluminous
forms and meet rigorous eligibility requirements to
access benefits. And some benefitssuch as TANF
come with sanctions or penalties for lack of compliance
with work and other requirements.
Because the procedures to obtain benefits often add to
the daily demands on these families, it should come as
no surprise that, for many, welfare reform has not
worked, and it may even have contributed to adverse
outcomes. When an unanticipated crisis occurs, there is
little opportunity for immediate assistance, and a
downward spiral can sometimes follow with grave consequences, including child abuse, domestic violence,
suicide, or abandonment. There is a compelling need
to assure access to resources and services beyond those
currently being provided.
Throughout the welfare system, workers make decisions based on regulations that cannot be readily adapted to the special circumstances of individual families.
For example, the plight of the young mother about to
197
Guerra
lose her job because her infant was perhaps mildly ill
could be avoided if the child care worker understood
that an occasional loose stool is not associated with
much risk for spreading disease. The family from Mexico with the neurologically impaired child would not
have to face giving up a portion of income to protect
their childs medical benefits if more flexible income
guidelines took into account the variability in income
from agriculture-based businesses. And perhaps the
grandmother would not have had to take in her three
adolescent grandchildren if her daughter had received
more counseling from the mental health system.
COMMENTARIES
In addition, when considering welfare reform and its

consequences, whether positive or negative, it is
important to recognize the influence of contemporary
values, expectations, and political forces in determining
who is considered deserving versus undeserving of
assistance. In particular, as discussed in the article by
Greenberg and colleagues in this journal issue, society
has decided to make it considerably more difficult for
poor immigrant families to qualify for benefits, especially TANF and Medicaid. For the most part, immigrant families can access benefits only if they lived in
the United States prior to 1996 or if they become citizens.9 The benefits available through various immigrant and refugee programs are generally not enough
to help immigrant families overcome the consequences
of poverty and other language and cultural barriers.
Recommended Changes to Improve

Supports for Poor Families
As we prepare for the discussion and debate that will
undoubtedly accompany the legislative process for
reauthorization, it is important to consider the many
challenges faced by poor children and their families.
With their complex and difficult lives in mind, I offer
the following six recommendations:
(1) Ensure that the interpretation of research findings
from studies on the impact of welfare reform take
into account the interrelationships of conditions
that are closely tied to social class and income
inequalities. Many factors other than poverty contribute to poor health, domestic violence, poor
pregnancy outcomes, low levels of educational
attainment, and hopelessness. These factors must
be better understood to keep families that have
198
been negatively affected by welfare reform from

falling further behind.10
(2) Revisit past programs and projects that incorporated components to help low-income families develop coping and problem-solving skills for lessons
about how best to improve these families health
and social well-being. Examples of such programs
might include the longitudinal Ypsilanti Michigan
Study; the California Wellness Project; and community mental health and substance abuse treatment programs.4,6,7
(3) Encourage a greater mix of public and private
resources to help poor families whose daily lives are
so precarious. Such families cannot be the sole
responsibility of government. Responsibility must
be shared with every community, urban or rural,
and with the families themselves. Leaders from the
private sector and the community can offer mentoring, networking, and promotion of social and
interactive skills that can help families secure
greater financial security.
(4) Clarify which immigrants are qualified to receive
which benefits, and provide guidance to states,
especially those with large immigrant populations.
Eligibility workers and Immigration and Naturalization Service representatives must be more sensitive and willing to assist immigrant families
through the maze of bureaucracies in a way that
recognizes their cultural and language needs.
(5) Assure families that when they are no longer eligible for TANF, they may still be eligible for other
benefits as they transition to work, such as food
stamps, Medicaid, and subsidized housing.
(6) Establish a job safety net of publicly funded,
low-wage service jobs that allows for a sense of purpose and future opportunity for individuals who
are willing to work but unable to find employment.
What is most apparent in my work with poor families is
that many barely get by, and they are struggling with
the complex circumstances in their lives. Some families
do just fine under welfare reform, but they are the
exception.11 Many others will continue to need support
from government and society. Helping these families is
Volume 12, Number 1
a shared responsibility. For communities with large

numbers of families in transition, the burden of social
and economic costs can be excessive and must be distributed across public and private sectors. When possible, individuals who benefit should assume personal
responsibility. But as policymakers debate and craft the
reauthorization legislation, they must also recognize
and support the critical dimensions of human need of
those struggling under the system. To do otherwise is
unconscionable.
COMMENTARIES
ENDNOTES
1. These brief case histories reflect the circumstances of several families receiving services in a primary care clinic in San Antonio,
Texas. The descriptions were kept general to protect the families
identities.
2. Daly, M., Duncan, G., Kaplina, G., et al. Macro-to-micro links in
the relation between income inequality and mortality. The Milbank Quarterly (September 1998) 76(3):31555.
3. Danziger, S. Approaching the limit: Early lessons from welfare
reform. JCPR working paper 195. Chicago: Joint Center for
Poverty Research, Northwestern University/University of Chicago, June 2000.
4. Sorensen, E., Mincy, R., and Holgen, A. Redirecting welfare policy toward building strong families. Washington, DC: Urban Institute Press, May 2000.
5. Finding Common Ground. National and state-level welfare
reform developments and the health of women and children. Presentation for staff at the San Antonio Metropolitan Health District. San Antonio, Texas. June 2000. See also Romero, D.,
Chavkin, W., Wise, P.H., et al. State welfare reform policies and
maternal and child health services: A national study. Maternal
and Child Health Journal (2001) 5(3):199206.
6. Chafel, J.A., ed. Child poverty and public policy. Washington, DC:
7. Syme, S.L. Social and economic disparities in health: Thoughts

about intervention. The Milbank Quarterly (1998)
76(3):493503.
8. Evanson, E., Manski, C.F., and Scanlan, T.M., eds. Evaluating
food assistance programs in an era of welfare reform: Summary of a
workshop. Washington, DC: Committee on National Statistics,
National Research Council and Institute of Medicine/National
Academy Press, 1999.
9. Zimmermann, W., and Tumlin, K. Patchwork policies: State assistance for immigrants under welfare reform. Washington, DC:
Urban Institute, 1999. See also note 5, Romero, et al.
10. A good example of such an approach can be found in Healthy
People 2010, with its specific goals and measures for mothers,
children, and adolescents, including risk-taking behaviors and a
variety of other measures that allow for comparison with national
trends. See U.S. Department of Health and Human Services.
Healthy people 2010: Understanding and improving health. 2nd ed.
Washington, DC: U.S. Government Printing Office, November
2000.
11. See the Urban Institutes Assessing the New Federalism Newsletter, January 2000. Available online at
http://newfederalism.urban.org/html/newsletter.html. See also
note 4, Sorensen, et al.; and note 6, Chafel.
199
Anderson
Volume:
COMMENTARY 4
Eloise Anderson
COMMENTARIES
elfare reform has reduced the number

of families receiving assistance, but has
not necessarily addressed the underlying problems that detract from positive
outcomes for children. A greater emphasis on family
issues could be key to promoting positive child outcomes in families receiving welfare. Shifting the focus
of welfare reform from job development to family
development could have the added benefit of strengthening many community support systems that provide
ongoing and renewing supports for families on welfare
as well as other poor families.
In 1996, when Congress created the new welfare program, Temporary Assistance for Needy Families
(TANF), it established four goals for the program and
allowed TANF monies to be used to address any of these
purposes. (See Box 1.) Although states have used their
TANF funds for a wide range of services, including
transportation, child care, housing assistance, substance
abuse treatment, and domestic violence programs,1
these services tend to support only one TANF goal: to
end dependence on government benefits.
Some policy analysts argue that, because of the reduction in the number of families receiving cash aid,
TANF resources should now be redirected to lowincome working families that are not receiving assistance. This idea seems to stem from a belief that the
goals of TANF have been addressed successfully. In
most states, however, a number of families are still
receiving TANF assistancefamilies that are having
difficulty getting and keeping jobs, and that may need
special services to achieve self-sufficiency.
Eloise Anderson is director of the Program for the

American Family at the Claremont Institute, a nonprofit educational organization in Sacramento, California, and former director of the California
Department of Social Services.
200
Therefore, I disagree with those who would redirect

TANF funds away from TANF families to other lowincome families. There are still a few things that need
to be accomplished with the families receiving assistance. Moreover, I believe that the focus should be on
child well-being and efforts to address the other goals
of TANFspecifically, keeping families together,
reducing out-of-wedlock births, and encouraging twoparent families. Although many post-TANF studies
suggest that efforts should focus on moving these families off assistance and into work,2 efforts that focus first
on these other goals may, in the end, prove more useful both for promoting child well-being and for ending
dependence on government benefits.
When restructuring TANF-funded programs and services to focus more on creating positive outcomes for
children, three important factors need to be considered: first, that underlying conditions may contribute
to a familys precarious economic situation; second,
that strengthening families requires attention to fathers
as well as mothers; and third, that a childs neighborhood and community also play an important role in
supporting child rearing.
Underlying Causes of Families Barriers to

Self-Sufficiency
To help families on welfare provide positive outcomes
for their children, it is necessary to understand the
issues they sometimes face. In my many years of working with poor families, I have observed that many
advocates, policy analysts, and legislators seem to
believe that poverty causes these families problems.
But I would argue that the problems some TANF families face are the main cause of their poverty, not the
other way around. Many problems stem from conditions that occurred in the mothers early childhood:
health issues (including sexual and physical abuse),
father absence, parental separation, and divorce. Many
cognitive disabilities are also the result of early experiences.3 Such conditions can place families in fragile
economic situations, which in turn can jeopardize childrens well-being.
As a result, the barriers to employment that many
recipients experience can best be addressed by focusing
more on family issues. In a recent study from the University of Michigan, the authors indicate that mental
Volume 12, Number 1
Box 1
The Four Goals of TANF
(1) To provide assistance to needy families so that children
may be cared for in their own home or in the home of
relatives;
(2) To end the dependence of needy parents on government
benefits by promoting job preparation, work, and marriage;
(3) To prevent and reduce the incidence of out-of-wedlock
pregnancies and establish annual numerical goals for preventing and reducing the incidence of these pregnancies;
and
Source: The Personal Responsibility and Work Opportunity Reconciliation

Act of 1996. Public Law 104-193, title I, section 103, 110 stat. 2113,
August 22, 1996.
and physical health problems are often issues for these

families.4 Moreover, childrens health problems and
severe physical abuse are found in the welfare population in numbers disproportionate to the general population. Finally, many recipients experience multiple
barriers that make work and parenting more difficult.
According to the report, the most difficult barriers to
overcome are perceived discrimination, few work skills,
and drug dependence. By shifting the focus and structure of services to family issues, more effective supports
can be provided to assist these families in overcoming
their barriers.
Consider, for example, children in the child protective
services system (CPS), most of whom are also in
TANF-funded programs.5 I believe that many children
are in CPS because of family problems involving substance abuse, family violence, and mental health
issuesfamily and personal issues that can best be dealt
with through the mental health system. Therefore,
TANF funds should be used to provide these families
In addition, some of the difficulties TANF parents

face in maintaining work may be related to cognitive
disabilities. Recipients with this kind of condition
may take longer to learn a work task or skill than the
employer is willing or able to accept. They may be
unable to cope with the demands of child rearing and
work, or unable to organize and regulate behaviors.
To help families address such problems, TANF funds
should be used to support the developmental disabilities (DD) system. The DD system has a long history
of addressing life skills issues, as well as placing people with cognitive disabilities into jobs, including
people who have more severe disabilities than TANF
recipients are likely to have. The DD work programs,
both sheltered and supportive, may be appropriate for
some TANF recipients who are unable to keep a job.
In addition, the DD system could provide parenting
support for TANF recipients with cognitive difficulties, as well as provide a network of resources to support families long term: supports that will be needed
when the children in these families become adolescents and their cognitive capacity exceeds that of their
parents.
COMMENTARIES
(4) To encourage the formation and maintenance of two-parent families.
with mental health services. Moreover, the work

requirements for these families should be seen as part
of their treatment program, monitored and supervised
by the mental health system, similar to how the mental
health system works with people who have chronic
mental illness.
Strengthening Families with Services for Fathers

and Partners
Most families on welfare are single-parent families, and
it is clear from a variety of studies that children living
in single-parent families are more likely to be poor and
at higher risk for negative outcomes than are children
living in two-parent families.6,7 But when children in
single-parent families receive financial support from the
noncustodial, nonresident parent, and the custodial
parent works, they are less likely to live in poverty.
These findings suggest that focusing on family issues,
especially the TANF goals of reducing out-of-wedlock
births and promoting marriage, could lead to better
outcomes for children.
Two points are important here. First, the design of
many programs seems to assume that single parents
201
Anderson
function in isolation, without relationships, especially

a relationship with an intimate partner, but this flies
in the face of reality. Such relationships generally do
exist. They can be supportive, or they can be a barrier to recipients employment success and their childrens well-being. Either way, relationships are an
important influence in these families lives. Second,
fatherswhether they live with the child or notare
often an extremely important influence on children.
Thus, if childrens well-being is a serious policy concern, then the needs and aspirations of fathers also
must be served.
Many of the fathers of children in families remaining

on welfare need jobs and job skills, and many also need
to overcome personal barriers to employment such as
substance abuse, physical or mental health problems,
or violent behavior. Although TANF fathers have issues
and needs similar to those of TANF mothers, fathers
have been neglected by social service policy and programs, and may distrust those programs. As a result, it
may be difficult for these programs to serve fathers as
well as mothers. For many fathers, a more user-friendly way to deliver services might be the substance abuse
system coupled with a mental health component, as
opposed to the welfare system with its traditional focus
on mothers.
In addition, studies show that children are seriously
affected when there is violence in the family.8 Programs
for dealing with violent behavior and domestic violence
are important because fathers and other men in mothers lives often serve as role models for children and
could cause children harm. TANF funds should be
used to support the mental health system to assist these
men and their families in overcoming violent behavior.
202
Associated Press, THE ADVOCATE
COMMENTARIES
The term family should be redefined to mean the

family of the child, which would automatically include
the noncustodial, nonresident parent. A change in definition would give noncustodial parents access to
TANF employment services and other supports. Furthermore, a focus on child well-being in developing
and providing services to TANF families requires consideration of men in their roles as fathers, husbands,
and boyfriends. Such services should include job programs for fathers, efforts to reduce family violence, and
new strategies for promoting marriage.
Finally, marriage is important for childrens well-being,
but poor couples often do not have access to premarital, marriage, or divorce counseling. Additionally,
because many unmarried parents did not themselves
have parents who married or remained married, they
lack models of successful marriage. TANF funds should
be used to provide families with information and counseling regarding marriage as soon as they apply for aid.
As discussed in the article by McLanahan and Carlson
in this journal issue, studies of unwed fathers indicate
that both parents are hopeful about marriage during
pregnancy and soon after the birth of a child. Thus, if
a childs birth leads a family to seek TANF assistance,
this would be an opportune time to introduce marital
services to both parents, with an emphasis on what is
good for the child. Moreover, TANF funds could be
used to make marriage counseling more available in
community programs and churches. My experience has
convinced me that faith- and community-based organizations are the best resource to provide these services
because they can utilize couples who have been married for many years to mentor younger couples.
Volume 12, Number 1
The Role of Communities and

Neighborhoods
Strong communities and neighborhoods play an
important role in supporting families in their childrearing function. Local residents and neighbors of welfare recipientsthe people who know them and live
around themshould become the providers of services for TANF families. Having programs provided by
the community helps ensure mutual trust and longterm support. Many of these programs already exist in
communities and neighborhoods where TANF recipients live, but they need to be expanded and developed
to accomplish the remaining goals of TANF.
COMMENTARIES
Community-based programs could strengthen and

revitalize not only TANF families, but also whole
neighborhoods and communities. Although the
emphasis would remain on TANF families, a community development approach to programming could
benefit other low-income families as well, since welfare
and low-income families live together in the same
neighborhoods.
In sum, TANF-funded programs should continue to

focus on families receiving welfare and not be redirected to the working poor, as much still needs to be done
to help families remaining on welfare overcome their
barriers to employment. Importantly, however, a
restructuring of programs and services is needed. The
current emphasis on job development and moving families off welfare and into work is designed to meet the
TANF goal of ending dependence on government benefits, but is not well suited to meet the needs of lowincome families. Instead, a greater emphasis on child
well-being and family development, designed to meet
the other goals of TANF, is much more likely to help
these families address their barriers to employment, as
well as promote positive outcomes for children.
ENDNOTES
1. Lazere, E. Welfare balances after three years of TANF block grants:
Unspent TANF funds at the end of federal fiscal year1999. Washington, DC: Center on Budget and Policy Priorities, 2000.
5. Sinanogla, P., and Maluccio, A., eds. The challenge of partnership:

Working with parents of children in foster care. New York: Child
Welfare League of America, 1981, p. 6.

employment of welfare recipients. Ann Arbor: Poverty Research and
Training Center, University of Michigan, 1999, p. 5.
6. McLanahan, S., and Garfinkel, I. Single mothers and their children: A new American dilemma. Washington, DC: Urban Institute Press, 1986.
3. Perry, B. The effects of traumatic events on children. ChildTrauma

Academys interdisciplinary education series. Houston: Baylor
College of Medicine, November 1994. See also Perry, B. Incubated in terror: Neurodevelopmental factors in the cycle of violence. In Children in a violent society. J.D. Osofsky, ed. New
York: Guilford Publications, 1997, pp.12448; and Karr-Morse,
R., and Wiley, M. Ghost from the nursery: Tracing the roots of violence. New York: Atlantic Monthly Press, 1997.
4. See note 2, Danziger, et al., pp. 14, 17.
7. Rickel, A.U., and Becker, E. Keeping children from harms way:

How national policy affects psychological development. Washington,
DC: American Psychological Association, 1997.
8. Hernandez, D.J. Americas children: Resources from family, government, and the economy. New York: Russell Sage Foundation,
1993, p. 58. See also note 7, Rickel and Becker; and a recent
journal issue devoted to the topic of domestic violence and children: The Future of Children (Winter 1999) 9(3).
203
Blum
COMMENTARY 5
Barbara B. Blum
s reauthorization of the Personal Responsibility and Work Opportunity Reconciliation Act approaches, a review of prereform
experience may help to inform the debate.
This commentary provides a historical perspective on
the impact of welfare policy on childrens lives, based
on my personal experiences working with and studying
low-income families for more than 30 years.
COMMENTARIES
My first exposure to the welfare system and foster care

services came in 1969, when I was asked to administer
the child welfare system in New York City. Almost all
of the 28,000 children then in public care came from
families that were eligible for Aid to Families with
Dependent Children (the public assistance program in
place before Temporary Assistance for Needy Families,
or TANF). This child welfare assignment was extraordinarily instructive. Despite the availability of assistance, children frequently were removed from their
homes because their parents lacked sufficient resources
to purchase shelter and sustenance. Clearly, public
assistance allowances were inadequate to pay for housing and food in New York City. Moreover, once parents were deemed fit to have their children returned to
them, the major barrier to reunification was the availability of affordable housing. To respond to this need,
New York Citys child welfare agency began in the
1970s to test whether an advance payment to parents
could support reunification. Families used the onetime payments of about $600 for rent deposits and, a
bit surprisingly, for washing machines. Basic needs
were being addressed.
Following my stint in child welfare services, in the late
1970s and early 1980s I served as commissioner for
Barbara B. Blum is director of the Research Forum

on Children, Families, and the New Federalism at
the school of public health at Columbia University in
New York.
204
New York States Department of Social Services, which

administered welfare, Medicaid, food stamps, and the
entire range of social service programs. In that role, I
learned from county commissioners about the diversity of problems faced by families on our New York caseloads and the complex requirements for accessing
benefits across various programs. Most of all, I remember the incomprehensible funding streams for child
care and the challenges of establishing eligibility levels
and fees for families using day care.
Subsequently, I became president of the Manpower
Demonstration Research Corporation (MDRC), a
nonprofit, nonpartisan social policy research organization. At MDRC, I saw the promise of work experience
begin to emerge as findings from the welfare-to-work
experiments of the 1980s consistently demonstrated
modest improvements in income and employment,
and modest reductions in the receipt of public assistance. Stemming from these findings, my later work at
the Foundation for Child Development examined
whether two-generation effectsthat is, benefits to
both mothers and their childrenmight be achieved if
intervention programs were designed to address the
needs of adults and children simultaneously.
These past experiences lead me to three key observations that deserve emphasis in the current debate over
reauthorization. First, income security is extraordinarily important for all families. Second, nutrition and
health care are also extraordinarily important for childrens healthy development. And third, child care, and
particularly early childhood education, should receive
special attention, as these programs can provide a powerful intervention for young children and their parents.
Income Security
The reauthorization process needs to encourage states
to provide sufficient support to working families, as
well as to vulnerable families, to ensure that stable and
decent housing, proper nourishment, and adequate
clothing are available to children. Without such support, optimum child development is impeded. Strategies states can use to increase family incomes include
financial incentives to help make work pay for families
currently receiving assistance, and improved child support collections from noncustodial parents.
Volume 12, Number 1
Nutrition and Health Care Programs

Several different programs help to ensure that lowincome families have adequate food and health care
coverage, which is so critical to childrens healthy
development. The food stamps program and TANF are
the major sources of support for ensuring that children
receive adequate nourishment. Medicaid and the more
recent program, the State Childrens Health Insurance
Plan (SCHIP), are the major funding streams for childrens health care. All these programs, however, have
problems with eligibility determination, access, financing, and quality.
For example, differences in income, assets, citizenship,
and recertification standards create confusion and
errors in determining eligibility for children and their
parents who seek benefits from TANF, food stamps,
Medicaid, and SCHIP.5 Reauthorization of the federal
welfare reform law provides an opportunity for Con-
COMMENTARIES
Child support is another important source of income.

As described in the article by McLanahan and Carlson
in this journal issue, recent improvements in the Child
Support Enforcement program have increased payments from noncustodial parents.4 We should expand
efforts to assist noncustodial parents by improving
their employment opportunities, and whenever feasible, encouraging beneficial relationships between parents and children.
nonstck/William Koechling
As described in the articles by Zedlewski and by Zaslow

and colleagues in this journal issue, several studies have
demonstrated the value of financial incentives to
increase family income and help make work pay as families transition from welfare to work.1 In addition, by
using state dollars to fund a separate assistance program for working families, Illinois has shown that
financial incentives can encourage families to work
while they are still receiving assistance, without using
time from the federal five-year clock.2 (Five years is the
maximum length of time that families are allowed to
receive federal assistance under the TANF block grant.)
Less has been done to examine how adequate income
can be provided for vulnerable families that are not
working. In fact, research findings are beginning to
highlight how susceptible these families are to sanctions and termination of benefits.3 Much more attention to children in these vulnerable families is needed.
gress to reconsider and reconcile these differences with
a focus on child well-being.
Moreover, eligibility for nutrition and health programs
does not ensure access to the benefits. Frequently, services are provided at times and in locations that make
access difficult for low-income working families.6 In
addition, although childrens health insurance regulations require extensive outreach, some states and localities have been less than vigorous in their efforts to
establish Medicaid eligibility, because federal reimbursement rates for Medicaid are less than for SCHIP.7
Without extensive outreach and advocacy, many eligible families may never receive available benefits. States
and localities, with federal encouragement, have an
opportunity during reauthorization to amend and
improve practices related to access.
In addition, the financing and quality of these programs, especially the health programs, have long raised
concerns. With Medicaid and SCHIP, fees for preventive services in most states have been very low for many
years, impeding the expansion of services for children.8
Furthermore, financing issues are frequently linked to
quality. Again, both federal and state governments
need to reexamine reimbursement practices and the
allocation of existing resources to ensure that reimbursement is adequate to support a reasonable quality
of care for low-income children.
205
Blum
Child Care and Early Childhood Education

Child care, and early childhood education in particular,
should receive special attention during the reauthorization of federal welfare reform. High-quality, stable
child care can be a powerful intervention for young
children and their parents.9 It can stimulate childrens
healthy development and strengthen parents childrearing skills. It is an essential service for parents in the
workforce. But problems related to eligibility, access,
and financing also hinder current child care programs.
COMMENTARIES
As the TANF block grants were implemented, many

states made generous investments in child care. However, too few jurisdictions addressed issues concerning
the link between eligibility for child care and the parents employment status. To ensure continuity of care
for young and often vulnerable children, eligibility
should be connected to childrens needs rather than to
the categorical standards of various funding streams.
Access is also an issue, as the supply of child care, particularly for infants, is not sufficient to ensure that children in families receiving TANF funds and other
low-income working families will receive quality care.
The financing of child care is clearly an issue.10 Perhaps

a commission or other appointed entity is needed to
outline options for financing more adequate and
robust child care services and to stimulate national
debate. The problem is of sufficient scale and of such
significance for Americas children that major changes
in financing are required. Possible options include a
relatively small tax increase devoted to child care that
could be added to the federal payroll tax. Responsibility for providing full-day early childhood education for
children age three and older might be assigned to state
and local education authorities, with resources from
state and local tax revenues.
Altogether, the reauthorization process for welfare
reform provides an opportunity to address the income
security, nutrition, health, and child care needs of poor
children in the United States today. Greatly improved
information systems, administrative practices, and
financing mechanisms will be required to achieve these
goals. Hopefully, national policymakers and the general public are ready to take on these challenges.
ENDNOTES
1. For example, the Minnesota Family Investment Program (MFIP),
Canadas Self-Sufficiency Project, and others have all demonstrated the effectiveness of financial incentives. See Bloom, D., and
Michalopoloulos, C. How welfare and work policies affect employment and income: A synthesis of research. New York: Manpower
Demonstration Research Corp., May 2001.
2. Lazere, E., Fremstad, S., and Goldberg, H. States and counties are
taking steps to help low-income working families. Washington, DC:
Center on Budget and Policy Priorities, May 2001.
3. Cherlin, A., Burton, L., Francis, J. et al. Sanctions and case closings for noncompliance: Who is affected and why. Policy brief. Baltimore: Johns Hopkins University, February, 2001. See also
Danziger, S. Approaching the limit: Early lessons from welfare
206
reform. Joint Center for Poverty Research working paper 195.

Chicago: JCPR, Northwestern University/University of Chicago,
June 2000.
4. See the article by McLanahan and Carlson in this journal issue.
5. Schott, L., Dean, S., and Guyer, J. Coordinating Medicaid and
food stamps: How new food stamp policies can reduce barriers to
health care coverage for low-income working families. Washington,
DC: Center on Budget and Policy Priorities, September 2001.
6. Broaddus, M., and Ku, L. Nearly 95 percent of low-income uninsured children now are eligible for Medicaid or SCHIP: Measures
need to increase enrollment among eligible but uninsured children.
Washington, DC: Center on Budget and Policy Priorities, December 2000. See also ONeill, H., Garcia, K., Amerlynck, V., and
Volume 12, Number 1
Blum, B. Policies affecting New York Citys low-income families.

Research Forum on Children, Families, and the New Federalism,
September 2001.
7. U.S. General Accounting Office. Medicaid and SCHIP: Comparisons of outreach, enrollment practices, and benefits. GAO/HEHS00-86. Washington, DC: GAO, April 2000.
8. American Academy of Pediatrics. Medicaid Policy Statement
(RE9918). Pediatrics (August 1999) 104(2):34447.
9. Collins, A., Layzer, J., Kreader, J.L., et al. National study of child
care for low-income families: State and community substudy interim report. Washington, DC: U.S. Department of Health and
Human Services, Administration for Children and Families,
November 2000. See also NICHD Early Child Care Network.
Child care and motherchild interaction in the first three years of
life. Developmental Psychology (1999) 35(6):13991413.
10. See earlier journal issues on caring for infants and toddlers, The
Future of Children (Spring/Summer 2001) 11(1); and on financing child care, The Future of Children (Summer/Fall 1996) 6(2).
COMMENTARIES
207
List of Acronyms
AFDC
Aid to Families with Dependent Children
CBPP
Center on Budget and Policy Priorities
CCDF
Child Care and Development Fund
CLASP
Center for Law and Social Policy
CSE
DHHS
U.S. Department of Health and Human Services
EITC
earned income tax credit
ESL
English as a second language
FIP
Iowas Family Investment Program
FTP
Floridas Family Transition Program
GUP
Growing Up in Poverty Project
IDA
Individual Development Accounts
INS
Immigration and Naturalization Service
JOBS
Job Opportunities and Basic Skills Training
MDRC
Manpower Demonstration Research Corporation
MFIP
Minnesota Family Investment Program
MOE
maintenance-of-effort (required level of state spending based on previous years

spending levels)
NEWWS
National Evaluation of Welfare-to-Work Strategies
NICHD
National Institute of Child Health and Human Development
PFS
Parents Fair Share
PRWORA Personal Responsibility and Work Opportunity Reconciliation Act of 1996
208
SCHIP
State Childrens Health Insurance Plan
SES
socioeconomic status
SPDP
State Policy Documentation Project
SSA
Social Security Administration
SSI
Supplemental Security Income
SSP
Canadas Self-Sufficiency Project
TANF
Temporary Assistance for Needy Families
TPD
Teenage Parent Demonstration
Volume 12, Number 1
The Future
of Children
Board of Trustees
Susan Packard Orr,
Chairman
Nancy Burnett,
Vice Chairman
Jane Lubchenco
Dean O. Morton
Franklin M. Orr, Jr.
Julie Packard,
Vice Chairman
Lewis E. Platt
William K. Reilly
Allan Rosenfield
Richard T. Schlosberg, III,
President and CEO
Robert Stephens
Colburn S. Wilbur
Honorary Emeriti
Trustees
Robin Chandler Duke
Robert J. Glaser, M.D.
Frank H. Roberts, Esq.
Edwin E. van Bronkhorst
The Future of Children 0
(ISSN 1054-8289) 2002
by The David and Lucile
Packard Foundation, all
rights reserved. Printed
in the United States of
America. Cover photo
Stone/Bruce Ayres
Printed on acid-free,
recycled paper with soy Ink.
(The electronic edition of this
issue can be found at
on the World Wide Web.)
Volume 12 Number 1
Winter/Spring 2002
Published by
The David and Lucile Packard Foundation
Editor-in-Chief
Editorial Advisory Board
Issue Editor
Issue Editorial Advisor
Senior Staff Editors
Courtney Bennett, Ph.D.
Kathleen Reich, M.P.P.
Production Editors
Patricia Fewer
Roselyn Lowe-Webb
Copy Editor
Lee Engfer
Web Editor
Forrest Bryant
Design and Publishing
Barbieri & Green, Inc.
We also appreciate the contributions to this journal issue by the
Foundations research librarians.
Peter P. Budetti, M.D., J.D.

Professor and Director, Institute
for Health Services Research and
Policy Studies
Northwestern University
Felton J. Earls, M.D.
Professor of Child Psychiatry in the
Department of Psychiatry
Consolidated Psychiatry Department
Leon Eisenberg, M.D.
Presley Professor of Social Medicine
in the Department of Social Medicine
Deanna S. Gomby, Ph.D
Special Advisor, The David and
Lucile Packard Foundation
David E. Hayes-Bautista, Ph.D.
Professor of Medicine and Director
Center for the Study of Latino
Health and Culture
University of California, Los Angeles
Philip R. Lee, M.D.
Professor (Emeritus) of Social
Medicine, School of Medicine
Senior Scholar, Institute for Health
Policy Studies
San Francisco
Eleanor E. Maccoby, Ph.D.
Professor (Emerita) of
Developmental Psychology
Stanford University
1. Blank, R.M., and Haskins, R., eds. The new world of welfare.
Washington, DC: Brookings Institution Press, 2001.
President and Senior Scholar
Child Trends
Paul W. Newacheck, Dr.P.H.
Professor of Health Policy
Institute for Health Policy Studies
San Francisco
Judith S. Palfrey, M.D.
Chief, Division of General Pediatrics
The Childrens Hospital
Harvard University
Nigel S. Paneth, M.D., M.P.H.
Chair, Department of Epidemiology
Professor of Pediatrics and Epidemiology
College of Human Medicine
Michigan State University
Lisa Simpson, M.B., M.P.H.
Deputy Director
Agency for Healthcare Research
and Quality
Barbara Starfield, M.D., M.P.H.
University Distinguished Professor
Professor of Health Policy and Pediatrics
Johns Hopkins University
Heather B. Weiss, Ed.D.
Director, Harvard Family
Research Project
Harvard University
Daniel Wikler, Ph.D.
Senior Staff Ethicist
World Health Organization
Rebecca A. Maynard, Ph.D.

University Trustee Professor of
Education and Social Policy and
Education
Graduate School of Education
University of Pennsylvania
The Future of Children is published twice annually by The

David and Lucile Packard Foundation, 300 Second Street,
Suite 200, Los Altos, California 94022. Fourth-class
postage paid at Los Altos, California, and at additional
mailing offices. The Future of Children is a controlled-circulation publication distributed free of charge. Opinions
expressed in The Future of Children by the editors or the
writers are their own and are not to be considered those of
The Packard Foundation. Authorization to photocopy articles for personal use is granted by The Future of Children.
Reprinting is encouraged, with the following attribution:
From The Future of Children, a publication of The David

and Lucile Packard Foundation. Correspondence should be
addressed to Richard E. Behrman, M.D., Editor-in-Chief,
The Future of Children, 300 Second Street, Suite 200, Los
Altos, California 94022. To be added to the mailing list,
write to the Circulation Department at the same address; or
to sign up for our e-newsletter, go to our Web site at
http://www.futureofchildren.org. Photographs that
appear in The Future of Children were acquired independently of articles and have no relationship to material discussed therein.
Policy Priorities. The state policy documentation project.
Washington, DC: CLASP and CBPP, June, 2000. Available online
at http://www.spdp.org.
16. McGroder, S.M., Zaslow, M.J., Moore, K.A., and Le Menestrel, S.

National Evaluation of Welfare-to-Work Strategies: Impacts on
young children and their families two years after enrollment:
Findings from the Child Outcomes Study. Washington, DC: U.S.
Department of Health and Human Services and U.S. Department
of Education, 2000.
1994.
care for low-income families: State and community substudy interim
report. Washington, DC: U.S. Department of Health and Human
Services, Administration for Children and Families, November
2000.

policies affect children: A synthesis of the research. New York:

employment of welfare recipients. In Prosperity for all? The economic boom and African Americans. R. Cherry and W.M. Rodgers III,
eds. New York: Russell Sage Foundation, 2000.
1999.

7. Duncan, G.J., and Chase-Lansdale, P.L., eds. For better and for
New York: Russell Sage Foundation, January 2002.
8. Edin, K., and Lein, L. Making ends meet: How single mothers survive welfare and low-wage work. New York: Russell Sage
Foundation, 1997.
9. Fix, M., and Passel, J.S. Trends in noncitizens and citizens use of
public benefits following welfare reform: 199497. Washington, DC:
Urban Institute, March 1999.
10. Garfinkel, I., McLanahan, S., Meyer, D., and Seltzer, J., eds.
Fathers under fire: The revolution in child support enforcement. New
11. Gennetian, L.A., and Miller, C. Reforming welfare and work: Final
report on the Minnesota Family Investment Program. Vol. 2, Effects
on children. New York: Manpower Demonstration Research Corp.,
2000.
12. Greenberg, M., and Savner, S. The final TANF regulations: A preliminary analysis. Washington, DC: Center for Law and Social
Policy, 1999.
13. Kirby, D. Emerging answers: New research findings on programs
to reduce teen pregnancy. Washington, DC: National Campaign to
Prevent Teen Pregnancy, May 2001.
14. Knitzer, J. Promoting resilience: Helping young children and parents
affected by substance abuse, domestic violence, and depression in the
context of welfare reform. Children and welfare reform issue brief 8.
New York: National Center for Children in Poverty, 2000.

Census Bureau, October 2000.
against poverty. Assessing the New Federalism Policy Brief No. B10. Washington, DC: Urban Institute, 2000.
24. U.S. Department of Health and Human Services. Temporary
Assistance for Needy Families (TANF) program: Third annual
report to Congress. Washington, DC: DHHS, August 2000.
well-being of Americas children and youth: 2000. Washington, DC:
U.S. Government Printing Office, 2000.
26. U.S. General Accounting Office. Education and care: Early childhood programs and services for low-income families. HEHS-00-11.
2000 green book: Background material and data on programs within the jurisdiction of the Committee on Ways and Means.
Washington, DC: U.S. Government Printing Office, 2001.
28. Zaslow, M.J., Brooks, J.L., Moore, K.A., et al. Experimental studies of welfare reform and children. Washington, DC: Child Trends,
2001. Available online at http://www.childtrends.org.
30. Zimmermann, W., and Tumlin, K.C. Patchwork policies: State assistance for immigrants under welfare reform. Washington, DC:
Volume 12, Number 1

Journal and Executive Summary
Journal
Executive
Summary
U.S. Health Care for Children

Winter 1992 Vol. 2, No. 2
NA
Protecting Children from Abuse and Neglect
Spring 1998 Vol. 8, No. 1
Health Care Reform

Summer/Fall 1993 Vol. 3, No. 2
NA
Children and Divorce

Children and Managed Health Care

NA
Sexual Abuse of Children

Home Visiting: Recent Program Evaluations

Spring/Summer 1999 Vol. 9, No. 1
NA
Critical Health Issues for Children and Youth

When School Is Out

Fall 1999 Vol. 9, No. 2
NA
Low Birth Weight

Domestic Violence and Children

NA
Critical Issues for Children and Youths

NA
Unintentional Injuries in Childhood

Children and Computer Technology

Fall/Winter 2000 Vol. 10, No. 2
Back Issues in Print
NA
Special Education for Students with Disabilities

Caring for Infants and Toddlers

Financing Child Care

Current Issue
The Juvenile Court


Winter/Spring 2002 Vol. 12, No. 1
Welfare to Work
Upcoming Issues
Children and Poverty

Financing Schools
All Upcoming Issues (see our Web site at

http://www.futureofchildren.org)
Remove my name from mailing list
Name
Street Address
City
State
)
NA
To order free Journals or Executive

Summaries, please send requests to:
Title /Affiliation
Phone (
Executive.
Summary.
Zip
E-mail
To sign up for e-mail newsletter, check here.
Circulation Department
The David and Lucile Packard
Foundation
Los Altos, CA 94022
E-mail
circulation@futureofchildren.org
Please assist us by indicating which of the following categories

best describes you:
Policymaker/policy analyst
Media/journalist
Service provider
Overall comments:
Advocate
Academic/researcher
General Public
FAX
(650) 941-2273
Please allow 46 weeks for delivery.
Journals, Executive Summaries, and
related Issues and Ideas guides are
also available online at
Note: We regret that, at this time, only
Executive Summaries can be shipped
internationally.
Suggested topics for future issues:
ORDER FORM
Long-Term Outcomes of Early Childhood Programs

Journal
PLEASE FOLD HERE FIRST WITH ADDRESS FACING OUT
NO POSTAGE
NECESSARY
IF MAILED IN
THE UNITED
STATES
BUSINESS REPLY MAIL

FIRST-CLASS MAIL PERMIT NO. 1053 LOS ALTOS, CA
POSTAGE WILL BE PAID BY ADDRESSEE
THE DAVID AND LUCILE PACKARD FOUNDATION

Circulation Department
Los Altos, California 94022
PLEASE FOLD HERE LAST WITH ADDRESS FACING OUT, TAPE AND MAIL

10 1 1 193 6844

Uploaded by

Copyright:

Available Formats

10 1 1 193 6844

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10 1 1 193 6844

Uploaded by

Copyright:

Available Formats

The Future of Children

Children and Welfare Reform

Children and Welfare Reform

Children and Welfare Reform

Editorial Advisory Board

Peter P. Budetti, M.D., J.D.

Rebecca A. Maynard, Ph.D.

The Future of Children is published twice annually by The

From The Future of Children, a publication of The David

16. McGroder, S.M., Zaslow, M.J., Moore, K.A., and Le Menestrel, S.

4. Dalaker, J. Poverty in the United States: 2000. Current population

5. Danziger, S., Corcoran, M., Danziger, S., et al. Barriers to the

6. Duncan, G.J., and Brooks-Gunn, J., eds. Consequences of growing

The Future of Children

22. Smith, K. Whos minding the kids? Child care arrangements.

Volume 12, Number 1

he primary purpose of The Future of Children

headed by single mothers. At the same time, many

Children and Welfare Reform

Children and Welfare Reform: Analysis and Recommendations

Reforms and Child Development

Experimental Studies of Welfare Reform and Children

Welfare Reform and Child Care Options for Low-Income Families

Family Economic Resources in the Post-Reform Era

Welfare Reform, Fertility, and Father Involvement

Welfare Reform and Parenting: Reasonable Expectations

Five Commentaries: Looking to the Future

Inside back cover

Children and Welfare Reform

Children and Welfare Reform:

looks like for children, how low-income children are

Analysis and Recommendations

The New Social Policy Landscape

Provisions of the law that addressed these goals

Volume 12, Number 1

Children and Welfare Reform

How Low-Income Children Are Faring

Economic Well-Being and Its Implications

The Future of Children

work requirements, time limits, and financial incentives

Analysis and Recommendations

higher on assessments of their cognitive skills, but also

Volume 12, Number 1

Children and Welfare Reform

to less frequent behavioral and emotional problems

Childrens Daily Lives

The Future of Children

Children in Families Moving from

Analysis and Recommendations

Total hours of care, stability of care, and type of care all

Volume 12, Number 1

Children and Welfare Reform

with opportunities to explore new skill areas, discover

The Future of Children

released in early 2002, which may help to clarify how

Analysis and Recommendations

domestic violence, child neglect or abuse, and removal

cedures to loss of eligibility and the impact of sanctions

Volume 12, Number 1