Child Welfare and TANF Implementation: Recent Findings

Report for Congress
Child Welfare and TANF Implementation:
Recent Findings
July 8, 2002
Emilie Stoltzfus
Analyst in Social Legislation
Domestic Social Policy Division

Congressional Research Service ˜ The Library of Congress

Child Welfare and TANF Implementation:
Recent Findings
This report examines recent research findings about Temporary Assistance for
Needy Families (TANF) implementation as it has affected the nation’s child welfare
system. The nation’s program of cash aid for needy families with children (TANF)
and its program to protect and care for children who are abused or neglected (child
welfare services) are linked by history and share some of the same clients who have
similar service needs. Assessing the full significance of the 1996 welfare reform law
(P.L. 104-193) to the child welfare system is complicated by the 1997 enactment of
the Adoption and Safe Families Act (P.L. 105-89), which made direct and major
changes to the child welfare system, and by the fact that TANF implementation
generally occurred during a strong economy.
Some child advocates were concerned that 1996 welfare reform initiatives –
including time limits, family cap policies, and work requirements – might harm
children and/or place new strains on the child welfare system. Research on the effect
of TANF implementation has not produced conclusive findings. Changes in the size
of the child welfare caseload have not been attributed to welfare reform, although
certain work requirements may increase the incidence of child neglect.
Welfare reform also changed specific child welfare funding streams and
renewed calls for stronger collaboration between child welfare and cash aid
programs. Here too the evidence of an effect on the child welfare system is
inconclusive. Greater collaboration has occurred in some locations and the flexible
funding of TANF, combined with a strong economy and rapidly declining cash
assistance caseload, meant new resources became available for child welfare services
in some states. Child welfare administrators are concerned, however, that because
this money is not dedicated to their agencies it may not always be available. Further,
the 1996 welfare reform continues to base eligibility for federal foster care and
adoption assistance on whether the family a child is removed from was (or would be)
eligible for cash aid as it existed in prior law (Aid to Families with Dependent
Children, AFDC). This increased the complexity of determining eligible federal
foster care and adoption assistance costs and, as the 1996 income and resource
standards become dated, might diminish federal child welfare funds by establishing
a diminishing pool of children eligible for federally reimbursable assistance.
Finally, some child welfare advocates have also been concerned that children
in families where no adult is receiving a TANF cash benefit (“child-only” cases)
receive the necessary support, services and protection they require. The percentage
of child-only cases has grown since enactment of welfare reform, although the actual
number of these families being served has declined. Some states have used TANF
funds to enact programs specific to the needs of child-only families, and many
require involvement of child welfare agencies to access these new supports.

In troduction ..................................................1
Background ..................................................1
How are TANF and Child Welfare Linked Today?....................2
What are the Concerns and Opportunities?..........................3
Child Welfare Trends and Research Findings........................4
National data on child maltreatment and foster care...............4
Survey findings about the child welfare caseload.................5
Findings about child maltreatment and out-of-home placements.....6
Benefit loss, child welfare involvement and speed of reunification...7
Collaboration and Funding......................................7
Collaborative efforts.......................................8
Funding changes..........................................9
Child-only and Relative Caregiver cases...........................11
Relative caregivers........................................12
Child-only cases with parents...............................13

Child Welfare and TANF Implementation:
Recent Findings
Public cash assistance for needy families with children and public child welfare
systems operate independently but share significant links. The 1996 Personal
Responsibility and Work Opportunity Reconciliation Act (P.L. 104-193) repealed the
prior cash aid for families program (Aid to Families with Dependent Children,
AFDC) and enacted the Temporary Assistance for Needy Families (TANF) block
grant. This change significantly altered how public cash assistance for families is
offered and its implications for the public child welfare system are still being
assessed; no clear picture has emerged. This paper examines survey and other
research findings related to TANF implementation and its relationship to child
maltreatment and use of public child welfare systems. It also discusses the
importance of TANF to collaborative efforts between the family cash assistance and
child welfare systems, changes in the child welfare funding landscape attributable to
TANF, and the growth in the share of TANF families that receive a “child-only”
Assisting children of needy sole-support mothers was the purpose of the
nation’s cash welfare program for poor families when it was established in the 1935
Social Security Act. Both then and now, however, eligibility for aid rests a great deal
on the circumstances of a child’s adult caregiver and federal reform efforts have
focused on these adults. In 1950, Congress made caregivers of needy children
eligible for federal cash aid (in addition to the children themselves) and as early as

1956 it amended the program to include purposes oriented to adult caregivers.

Accordingly, the goal of family self-support joined the program goal of allowing
children to be cared for in their own home (or the home of a relative). Over the next
four decades reform efforts added employment, training and other services to the
program to bolster this family self-support goal. These efforts culminated in the
1996 enactment of the Temporary Assistance for Needy Families (TANF) block
TANF strengthened efforts to reduce dependence on government cash aid by
requiring adult recipients to work and establishing time limits on their federal
support. To the long-standing goals of self-support and care of children in their own
homes, TANF added the goals of reducing the number of births occurring outside of
marriage and encouraging the formation of two-parent families. Through all of these
changes, however, responsibility for support of a child remained a condition of cash
aid and in May 2002, the House passed TANF reauthorization legislation (H.R. 4737)

that would reassert the centrality of child well-being by making it the overarching
aim of all four TANF purposes.
Federal support for state child welfare systems also grew out of the 1935 Social
Security Act. The initial legislation provided funds to state public welfare agencies
to ensure services in rural areas for the “protection and care of homeless, dependent
and neglected children.” Eligibility for the services rested solely with a child’s need
for care and protection. In 1962 (P.L. 87-543) Congress expanded funding
authorization for child welfare services, required states to coordinate these services
with their cash aid program to dependent children and their families, and sought to
ensure child welfare services would be made available in all parts of each state (by
July 1975). Also in 1962, Congress made permanent the provision of federal foster
care assistance to children who were necessarily removed (by court order) from
families receiving AFDC – TANF’s predecessor program.
Federal foster care assistance had first been authorized as a component of the
cash aid program in 1961 and it remained a part of the AFDC program until 1980.
In that year the Adoption Assistance and Child Welfare Act (P.L. 96-272) established
foster care assistance as an independent program with its own set of state plan
requirements. The 1980 law also created federal adoption assistance. In 1993, P.L.
103-66 established a new program and funds for child welfare services that focused
on family preservation and support. The most significant change to the child welfare
system since 1980, however, came with the 1997 enactment of the Adoption and Safe
Families Act (P.L. 105-89). This law emphasizes the paramount importance of safety
for children and seeks to increase permanency for foster care children by establishing
strict time lines for achieving family reunification, adoption, guardianship or another
permanent placement for a child.
How are TANF and Child Welfare Linked Today?
TANF policies are relevant to the well-being of all poor children and, because
they are the most often served, especially those in single-parent families.1 Child
welfare services, by contrast, are designed for children who experience abuse and
neglect and they can come from families of any income level or structure. Although
the vast majority of children in cash aid families do not experience maltreatment,
research shows higher rates of abuse and neglect among poor and single-parent
families and many children in public foster care have been removed from families
that receive cash welfare.2 A 1995-1996 three-state study (California, Illinois, North
Carolina) found that fewer than 3% of children on or recently removed from cash aid

1 For discussion of research concerning outcomes among children in cash welfare
households as a whole, see the CRS Welfare Reform Briefing Book page on Research, Child
Outcomes [].
2 U.S. Department of Health and Human Services. Andrea J. Sedlak and Diane D.
Broadhurst, Third National Incidence Study of Child Abuse and Neglect. Washington, D.C.


were victims of maltreatment but that 60% of all children entering foster care were
in families receiving (or recently receiving) cash welfare.3
Federal child welfare and cash aid policies are also linked by federal financing
requirements. As noted, Congress first funded foster care payments for children as
part of TANF’s predecessor program, AFDC, and this aid was specifically for
children who were removed from families receiving federally supported cash aid
(P.L. 87- 31).4 Eligibility for federal foster care assistance remained linked to AFDC
even after Congress established this assistance as an independent program in 1980.
And, although the 1996 welfare reform law repealed AFDC, it provided that
eligibility for federal foster care assistance remains tied to children who are placed
in foster families after initially being removed from families receiving (or eligible
for) cash aid under AFDC (as it existed on July 16, 1996).5 Nationwide an estimated

55% of the foster care caseload is eligible for federal foster care assistance.6

What are the Concerns and Opportunities?
Some child advocates feared that TANF requirements – mandatory work
participation, benefit loss (sanction) for failure to comply with work or other
requirements, time-limited aid, and broad restrictions on aid to immigrants – would
cause harm to children and would negatively affect the public child welfare system.
At the same time, supporters of the welfare reform law argued that the changes,
especially requiring work, would boost self-esteem among parents on welfare and
improve their ability to parent and serve as role models for their young children.
Aside from the effect of the new law on children and the size of the child
welfare caseload, enactment of TANF renewed calls for better collaboration between
child welfare and cash aid agencies and some new initiatives have followed. The
broad purposes of the TANF block grant, combined with a declining cash aid
caseload, provided a potential source of flexible funds for child welfare services. The
1996 welfare reform law, however, continues to link federal reimbursement for foster
care and adoption assistance to AFDC income and resource standards (as they existed
on July 16, 1996). Beyond the complexity of documenting eligibility for these
children, these increasingly dated standards ensure a diminishing pool of federally
eligible children (and thus declining federal child welfare funds).

3 U.S. Department of Heath and Human Services. Dynamics of Children’s Movement
Among the AFDC, Medicaid and Foster Care Programs Prior to Welfare Reform: 1995-

1996. Washington, D.C., 2000; online at [].

4 In 1961 the program was called Aid to Dependent Children but was changed to Aid to
Families with Dependent Children in 1962 by (P.L. 87-543).
5 P.L. 104-193 originally set June 1, 1995 as the eligibility look-back date. Technical
amendments made by P.L. 105-33 changed this date to July 16, 1996 and made that date
effective as if enacted in P.L. 104-193.
6 U.S. Congress. House. Ways and Means Committee. Green Book 2000, Committee
Print, Washington, D.C. p. 724.

Finally, a post federal welfare reform increase in the percentage of TANF
families that are “child-only cases” (the amount of cash aid offered is for the
child(ren) and no adult portion is granted), has raised new questions about the
adequacy of care and services available to these children and their families.
Child Welfare Trends and Research Findings
Research studying the effect of welfare reform initiatives in child welfare has
not yet produced a clear picture of changes.7 The fact that many states implemented
“TANF-like” policies before the enactment of TANF in 1996 and the post welfare-
reform passage of extensive federal child welfare policy changes (via the Adoption
and Safe Families Act of 1997, P.L. 105-89) complicates the task of assessing the
importance of federal welfare reform to changes in the child welfare system.
Researchers also note that the implementation of TANF took place primarily in an
expanding economy – when many low-skilled parents could find jobs – and that more
time must pass before the implication of federal time limits on cash aid can be fully
understood. Information now available to assess the meaning of TANF
implementation to the child welfare system is drawn from survey data obtained
primarily from child welfare workers, as well as some welfare recipients, and
administrative data regarding child maltreatment victims, foster care caseloads, and
welfare receipt. Some of the research looks exclusively at periods before welfare
reform, but in these cases the researchers are studying specific locations where
federal waivers allowed establishment of “TANF-like” policies. Available research
varies widely in scope and methodology, and only one experimental study (using
random assignment and a control group) has been done.
National data on child maltreatment and foster care. In CY2000 an
estimated 879,000 U.S. children were found to be victims of abuse or neglect for an
incidence rate of 12.2 child victims per every 1000 children. This is above the
number (826,000) and incidence (11.8) of child maltreatment victims in CY1999 but
below the CY1993 high of an estimated 1 million child victims with an incidence
rate of 15.3. For CY2000 states also reported an increase in the percentage of child
victims who experienced neglect, as compared to other kinds of maltreatment. The
proportion of child victims experiencing neglect (including abandonment) rose from
58% in 1999 to 63% in 2000 while the proportion of children who experienced
physical or sexual abuse continued to decline. Researchers do not yet know whether
the 2000 data represent the start of a new upward trend in substantiated child8
maltreatment, particularly neglect cases, or are simply a single-year anomaly.

7 For research surveys see Kristen Shook Slack. Assessing the Influence of Welfare Reform
on Child Welfare Systems. Focus v. 22, no. 1, February 2002. p. 98-105; online at
[] and Rutledge Hutson. Red Flags: Research Raises Concerns
About the Impact of Welfare Reform on Child Maltreatment. Center for Law and Social
Policy, October 2001; online at [].
8 U.S. Department of Health and Human Services. National Child Abuse and Neglect Data
System (NCANDS). Summary of Key Findings from Calendar Year 2000, April 2002;
online at [].

Researchers believe the number of children in foster care has at least stabilized
and perhaps declined in recent years. (Complete data is not yet available, and
changes in the quality of state-reported data may have impacted the reported trend.)
A preliminary HHS estimate puts the number of children in foster care on September
30, 2001 at 565,000. This compares to an estimated 581,000 children in care at the
end of FY1999 and an estimated 556,000 children in care on September 30, 2000.9
The foster care caseload had previously grown in every year for at least two decades,
with the sharpest growth occurring in the last half of the 1980s. From FY1996
through FY1999 the caseload rose as some 1,047,000 children entered foster care
while considerably fewer, an estimated 941,000, left care (i.e., to family
reunification, adoption, legal guardianship, emancipation, or other custody). As of
June 28, 2002, new estimates for the number of entries, as opposed to exits, from the
foster care system were not available. As a result, it is not possible to determine
whether fewer children have entered foster care in recent years or whether the
caseload decline is primarily attributable to a greatly increased number of children
exiting the system.10
Survey findings about the child welfare caseload. A 2000 survey of
46 state child welfare administrators, conducted as part of the National Survey of
Child and Adolescent Well-Being (NSCAW) reports that the majority said they were
unsure of, or that it was too early to tell, the effect of TANF on the number and
characteristics of their state child welfare caseload. Fifteen percent said there
appeared to be an increase in the proportion of TANF families using prevention,
child protective services (CPS) investigation, and ongoing or long-term child welfare
services. (At least one state legislature mandates a CPS investigation of neglect for
families sanctioned because of TANF violations and, in that state, the number of
TANF families referred to CPS by TANF workers increased from 15 sanctioned
families in 1998 to 59 in 1999.)11
The Urban Institute compared the size of 1996 and 1998 child welfare caseloads
in 12 states (AL, CA, CO, FL, MA, MI, MN, NY, NJ, TX, WA, WI) and found them
relatively unchanged in each state (or that change was attributed to state or local
legislation and events). In some locations, however, child welfare workers surveyed
cited federal welfare reform as the cause for growth in particular kinds of child

9 See “Alert” online at []. The
caseload numbers for the last day of FY1999 and FY2000 are “interim” estimates. Interim
caseload estimates are based on more and better data and so are more reliable then
preliminary estimates. Not all data is available yet for FY2001 so the estimated size of the
foster care caseload on the last day of FY2001 is a preliminary number.
10 Child welfare system concerns related to TANF implementation primarily predicted
growth in the number of foster care entries. In the meantime many of the provisions of the
1997 Adoption and Safe Families Act, especially the emphasis on quickly finding
permanency for children in foster care, were intended to ensure quick exits of children from
foster care.
11 U.S. Department of Health and Human Services. National Survey of Child and
Adolescent Well-Being. State Child Welfare Agency Report, June 2001. Online at
[ h t t p : / / www.acf p r o gr ams/core/ongoing_research/afc/wellbeing_i nt r ml ] .
(Hereafter cited as NSCAW, State Survey, 2001)

welfare cases. These included cases related to neglect or poverty (as opposed to
physical abuse), and cases where a parent was a domestic violence victim or was
ineligible for aid due to immigrant status. Overall, workers in six states reported an
increase in abandoned children, which they attributed to stricter welfare
requirements; workers in five states reported more referrals for “inadequate
supervision,” due to parents working and using inadequate child care; and workers
in five states reported an increase in maltreatment among immigrant children.12
Findings about child maltreatment and out-of-home placements.
A comparative analysis of state-level data related to welfare and child welfare
policies from 1990-1998 suggests that certain policies, including benefit size and
work requirements, may affect maltreatment rates and/or the number of out-of-home
placements. This study’s strongest finding was that more generous welfare benefits
are associated with significantly lower levels of neglect and smaller numbers of
children in out-of-home care. (Or conversely, reduced benefits are associated with
higher levels of neglect and larger numbers of out-of-home placements.) These
researchers also report that family cap policies, which typically mean that a family’s
benefit is not increased to meet the needs of an infant born 10 months or more after
the family begins to receive cash aid, are associated with lower numbers of
substantiated cases of maltreatment but increased numbers of out-of-home
placements. Other findings from this research suggest that more single mothers
working leads to increased child maltreatment (especially neglect).13
Other evidence of the effect of welfare reform on child maltreatment is mixed.
An experimental study in Delaware also suggests a connection between strong work
requirements and increased levels of child neglect. The Abt Associates evaluation
of Delaware’s pre-TANF waiver program found higher rates of neglect (after 3 years)
among families that were subject to strong work requirements and time-limited14
benefits (as compared to those subject to the old cash aid program). Researchers
who used administrative data to study the effect of welfare reform on the incidence
of child maltreatment among Illinois children entering AFDC/TANF, however, found
a decline in reported child maltreatment after enactment of federal welfare reform.
For children who received cash benefits in the years 1994-1996 (during which Illinois
instituted a Work Pays program to encourage employment among adult welfare

12 Rob Geen, Lynn Fender, Jacob Leos-Urbel, and Teresa Markowitz. Welfare Reform’s
Effect on Child Welfare Caseloads. The Urban Institute, Washington, D.C., 2001.
(Hereafter cited as Geen et al., Welfare Reform’s Effect on Child Welfare); online at
[] (click successively on Research, Children, Child Welfare and then
scroll down to this report).
13 Christina Paxson and Jane Wadefogel. Welfare Reforms, Family Resources, and Child
Maltreatment. [In] Bruce Meyer and Greg Duncan, eds. The Incentives of Government
Programs and the Well-being of Families. Joint Center for Poverty Research, Northwestern
University/University of Chicago, 2001. (Hereafter cited as Paxson and Wadefogel, Welfare
Reforms); online at [http://].
14 David Fein and Wang S. Lee. The ABC Evaluation: Impacts of Welfare Reform on Child
Maltreatment. Abt Associates. Prepared for Delaware Health and Social Services,
December 2000; online at
[ ht t p: / / www.abt a ssoci at r epor t s / wel f a r e -downl oad/ ma l pap-2.pdf ] .

recipients) the incidence of reported child maltreatment 1 year after first receiving
cash benefits remained stable. But for children who first received cash aid in 1997
– after federal welfare reform was enacted – there was a 10% decrease in
substantiated maltreatment cases; researchers expected this lower rate to hold for the
children who entered the cash aid program in 1998.15
Benefit loss, child welfare involvement and speed of reunification.
A study comparing 187 Utah families who left TANF because of sanctions to 200
randomly selected Utah families who left for other reasons found the sanctioned
families were no more likely than non-sanctioned families to have a child welfare
case opened at some time within 3 years after their cash aid ended. Both groups,
however, had a relatively high involvement with the child welfare system (20.3% and
22.6%, respectively). And among families with child welfare involvement, those that
had been sanctioned were much more likely to have been found without adequate
resources to support their families (67.6% compared to 28.9% of non-sanctioned
families) and to have an open child welfare case at the end of the three years (21%
compared to 2%).16
Separately, a Chicago area study that combined Illinois state administrative data
and survey responses from 173 cash aid recipients (1995-1996) found that the odds
of child welfare involvement were three times greater for respondents who had their
benefits reduced by at least $75 in a month and who did not have some employment
income, as compared to those who experienced this benefit loss while maintaining
some wage income. A study in Cuyahoga County, Ohio, however, found that for
mothers whose children were already in foster care, employment might slow
reunification. The study used administrative data to track the effect of income on the
reunification of children who were placed in foster care before (n=378) and after (n=
525) welfare reform. It found that pre-reform children were reunified at a rate that
was 65% faster than the reunification rate for post-reform children. Although higher
income was a significant factor in speeding reunification and post-reform children
were more likely to have mothers with income from wages, reunification was slowed
when earnings represented the greatest proportion of a mother’s income.17
Collaboration and Funding
Federal child welfare law has for decades required states to coordinate their
child welfare services with assistance and services provided by AFDC/TANF but this
has not, historically, led to extensive collaboration. Enactment of federal welfare
reform spurred calls for greater cooperation. This was true, in part, because the

15 Robert Goerge and Bong Joo Lee. Changes in Social Program Participation in the

1990s: Initial Findings from Illinois. Working Paper, Joint Center for Poverty Research,

Northwestern University/University of Chicago, 2000; online at
[ h t t p : / / www.j c pr .or g/ wpf i l e s/ go er ge r e vi sed.pdf ] .
16 Michelle Derr and Vicky Cooley. Does Welfare Grant Sanctioning Increase Child
Welfare Involvement Among TANF Families? Unpublished, in review at Social Services
17 Kathleen Wells, Case Western University and Shenyang Guo, University of Tennessee,
Interim Report II, September 2001.

flexible funding of TANF made it potentially available for more child welfare uses
and because of concerns that “dual-system”clients (those served by both child welfare
and TANF) might have trouble simultaneously meeting requirements for TANF
assistance, especially work participation, and child welfare case plan requirements
(e.g., court hearings, child visits, or participation in substance abuse treatment,
parenting skills training or other comparable activity).
Collaborative efforts. Three kinds of collaborative efforts have been
identified. These are 1) front-line worker efforts to coordinate case planning of dual-
system clients; 2) agency efforts to create programs that meet shared service needs
regardless of whether a client is identified through a child welfare or TANF office
(e.g., relative caregivers, teen parents, people needing mental health or substance
abuse treatment); and 3) systemic efforts such as joint planning, co-locating workers,
and cross-agency training. Obstacles to collaboration include: no shared information
system between TANF and child welfare agencies (this might be a technical issue but
can also raise confidentiality concerns for child welfare workers); heavy workloads
for TANF and child welfare employees who often have different professional
training, experience and priorities; no shared focus between agencies and their
workers. (The primary attention in the TANF agency is usually adults while children
are the central focus of child welfare agencies.)18 Other issues include concerns that
TANF clients may have about child welfare intervention and restrictions TANF
workers may have on working with families after a child has been removed from the
Urban Institute researchers found recognition of potential for collaboration in
most of the 12 states and multiple county sites it visited in 1999 but many efforts19
were informal or just beginning. Basing their projections on a random sample of
92 county child welfare agencies surveyed in 1999-2000, the National Survey of
Child and Adolescent Well-Being (NSCAW) determined that since the 1996
enactment of federal welfare reform more than one-third (38%) of all counties
nationwide have created multi-program teams that include both TANF and child
protective services (child welfare) workers. Other more intensive kinds of
collaboration are less common. An estimated 22% of counties routinely refer TANF
clients who experience benefit loss (sanctions) to child welfare services while an20
estimated 24% routinely link employment and child welfare services.
El Paso County, Colorado is frequently cited as a leader in collaborative efforts.
Seeking to leverage available TANF funds for child welfare preventive purposes,
county leaders created or merged a range of programs. These include joint family
preservation and family support services (regardless of whether the initial social
service contact was made through child welfare or TANF) and support and services

18 Jennifer Ehrle, Karin Malm, Lynne Feeder and Roseanna Bess. Welfare Reform and
Opportunities for Collaboration between Welfare and Child Welfare Agencies. Urban
Institute, Washington, D.C., August 2001.
19 Ibid.
20 U.S. Department of Health and Human Services. National Survey of Child and
Adolescent Well-Being. Local Child Welfare Agency Report, v. 37, June 2000, online at
[ ht t p: / / www.acf pr ogr ams/ cor e / ongoi ng_r esear ch/ a f c / wel l bei ng_i nt r ml ]

for relative caregivers, teen parents, families experiencing domestic violence, and
others who may have cross-agency concerns. Other county efforts include cross-
systems training on TANF and child welfare as well as substance abuse, mental
health, and domestic violence issues.21
Collaboration initiatives in other parts of the country include a 2-year multi-
county project in North Carolina that drew on many of the principles established in
El Paso County, Colorado. Overall the North Carolina project sought to integrate the
goals of family self-sufficiency and child well-being by creating a holistic focus on
service to families. Among a range of activities it established regular
communications between TANF and child welfare workers.22 And in California, a
4-year initiative to improve TANF and child welfare service links began with a
spring 2001 survey intended to identify current practice. At that time, roughly half
of the state’s 58 counties required child welfare workers to determine if clients had
a “possible need” for TANF services and about the same number had established
steps that require TANF workers to determine whether a client is being served by
child welfare. (However, most counties (37) lacked an information system that
could display a client’s child welfare status for the TANF worker.) Thirty-one of the
counties indicated they had begun to implement coordinated case planning, although
only six said this kind of case planning had been fully implemented.23
Funding changes. The 1996 law that created the TANF block grant (P.L.
104-193) also made significant changes to child welfare funding. The cumulative
effect of these changes on the amount of funding available for child welfare services
and the kind of services offered is still being judged. In some instances it appears
child welfare has received new funding, but because this money is not dedicated to
child welfare purposes there is uncertainty about its availability in future years.
As part of the 1996 welfare reforms, Emergency Assistance funds (often used
for child protective services among cash aid recipients) were folded into the larger
TANF block grant. The 1996 law also placed new restrictions on children’s
eligibility for Supplemental Security Income (SSI). Many states apply for SSI on
behalf of disabled foster care children and the benefit is used to offset the cost of
maintaining custody of these children.24 An Urban Institute survey comparing state

21 David Berns and Barbara Drake. Combining Child Welfare and Welfare Reform at a
Local Level. Policy and Practice of Public Human Services, v. 57, no. 1, March 1999. p.


22 For more information go to the TANF/Child Welfare Collaborative website at
[ ANFindex.htm] .
23 Kate Karpilow, Richard Speiglman, and Linda Orrante. Linkages between TANF and
Child Welfare Services in California Counties: A Report of Baseline Survey
Presented at 51 Annual NAWRS Conference, Baltimore, MD., August 22, 2001. For more
information on the CalWORKS/Child Welfare Partnership Project go to
[ h t t p : / / www.ccr wf .or g/ cal wor ks/ i ml ] .
24 The U.S. Supreme Court has agreed to hear a Washington state case (during its 2003 term)
regarding the legality of this practice. In October 2001 the Washington State Supreme Court
struck down the state’s practice of claiming SSI and Social Security survivor benefits to

fiscal year spending in 1998 and 2000 found a 16% decline in foster care children
who received SSI dollars.25 In addition, the 1996 law reduced the amount of money
available for the Social Services Block Grant (SSBG), another significant, non-
dedicated, source of child welfare funds. At the same time, the federal law allowed
states to transfer up to 10% of surplus TANF funds into the SSBG provided that
money is used to serve children or their families who have income under 200% of the
federal poverty line. Finally, as stated earlier, the welfare reform law continues to
link a child’s eligibility for federal foster care assistance to the prior cash welfare
program, Aid to Families with Dependent Children (AFDC), with that program’s
eligibility standards frozen as they existed on July 16, 1996. As these standards
become dated, the pool of children who are eligible for federal reimbursement of
foster care or adoption assistance is likely to shrink.
In FY2000 states spent close to $640 million in SSBG funds, or about 23% of
their total SSBG expenditures, on four kinds of child welfare services – child foster
care, child protective services, adoption and independent living services. Other child
welfare related services, including case management, prevention/intervention
services, residential treatment and counseling received additional and significant
SSBG funding. Because of reduced SSBG appropriations, the level of SSBG funding
for any purpose is increasingly dependent on large transfers of cash from TANF to
SSBG accounts. In FY2000 approximately 37% of the total SSBG expenditures were
made using dollars transferred from TANF accounts.26 In sum, a number of factors
that have made TANF funds available for SSBG expenditures – including a generally
declining cash assistance caseload are in flux and could change in an economic
downturn and/or as a result of certain legislative proposals. For instance, expanded
work requirements, especially in a less robust economy with fewer entry level jobs
in the private sector, might require states to increase their use of TANF funds in
support of work activities or experience; likewise stricter work requirements could
increase the demand for child day care and thus affect the ability of states to continue
devoting an increasing proportion of SSBG resources for child welfare. At the same
time some pending legislative proposals would increase overall appropriations for
SSBG; these proposals could affect state decisions to transfer TANF dollars to SSBG
The 2000 NSCAW survey of 46 state administrators found that 25% of the
administrators reported increased access to funding or services for child welfare
programs since implementation of TANF. When new TANF funds or resources were
available they were most often used to enhance or expand prevention and
intervention services and out-of-home placements (including kinship care) for

24 (...continued)
reimburse its foster care expenses for eligible children (see Keffeler v. State of Washington).
25 Rob Geen. Shoring Up the Child Welfare–TANF Link. Short Takes on Welfare Policy,
no. 7, Urban Institute, June 2002; available online at [] (click
successively on Research, Children, Child Welfare and then scroll down to the report).
26 The amount of fiscal year state expenditures of SSBG funds is generally greater than the
amount of funds appropriated in that fiscal year because states may transfer funds from other
places, not least TANF. U.S. Department of Health and Human Services, SSBG Annual
Report: 2000, in press.

children at risk of further maltreatment.27 A number of the 12 states surveyed by the
Urban Institute reported an increase in funding for child welfare services between
1996 and the present. TANF funds themselves were a notable part of child welfare
spending in at least six of these states although administrators expressed concern that
because these funds were not dedicated to child welfare they might not always be
available. Texas expected to use $64 million in TANF monies for foster care and
adoption support payments in FY2001 (at least part of this money supplanted state
funding). Minnesota used TANF funds to initiate a concurrent planning program that
requires both family reunification and long-term placement plans to be established
for all foster care children. In California, both Los Angeles and Alameda county
officials reported fewer foster care children receiving SSI. Overall TANF funding
for child welfare in California dropped dramatically in FY1998 to about a half
million dollars (far below the $87 million spent under Emergency Assistance 2 years
earlier) but has since rebounded to $167 million in FY2000.
Child-only and Relative Caregiver cases
Close to 1.4 million children lived in families that received child-only benefits
in FY2000. Some child welfare advocates have expressed concern about whether
these “child-only” TANF families (where no adult receives cash aid) have access to
adequate resources and support services. Along with the rest of the TANF caseload,
the number of families receiving child-only benefits declined from 978,000 in
FY1996 to 782,000 in FY2000. However, because the decline in the percentage of
child-only cases has been slower than the overall TANF caseload decline, child-only
families now represent more than one-third of all cases (34.5% in FY2000). Child-
only cases represented well below one-quarter of all cases before welfare reform
(21.5% in FY1996) and this substantial percentage gain has helped to draw new
attention to particular circumstances and needs of child-only families.28
In child-only cases, the family may be headed by a relative (non-parent)
caregiver – frequently a grandparent – who is income ineligible for aid or who
chooses not to receive TANF cash aid because doing so would require compliance
with work and other TANF requirements. Other kinds of “child-only” families
include those where benefits to a child continue after they are ended for the adult(s)
– due to time limits on aid or because of sanctions for failure to meet work or other
TANF requirements. Child-only cases also include families where the adult head is
ineligible for aid because of citizenship status or previous drug felony conviction.
Finally, a child-only case may exist when the adult caregiver is a recipient of SSI.
Nationwide most child-only cases (62.4%) go to families where the child lives
with a parent who is not receiving aid. But this national number hides a great deal
of variation in the composition of an individual state’s child-only caseload. In close

27 NSCAW, State Survey, 2001.
28 U.S. Department of Health and Human Services. TANF: Fourth Annual Report to
Congress. X-192, Exhibit II, (Data in this report is for FY2000). May 2002 Some part of
the growth in the proportion of child-only cases is due to states’ use of separate state
programs, especially those for 2-parent families. When these cases are included in the
overall cash aid caseload the proportion of child-only families drops to 33.0%.

to half of all states (24) more than 50% of the child-only cases involved children who
lived with a nonparent caregiver, most often a grandparent.29 The kind of sanction
policies a state implements (e.g., does the state end the entire family’s benefit for a
violation, or does aid always continue for children), the availability of other state
social programs to serve nonparent relative caregivers, and the size of a state’s
immigrant population are some of the factors that may affect a state’s child-only
caseload composition.
Relative caregivers. Grandmothers, aunts, and other relative caregivers have
been a part of the cash welfare program for decades. More than half a million TANF
children live with a kinship caregiver, although not all of these are in child-only
cases.30 Most children who live with their grandparents or other relatives do so on
an informal basis but some are placed in relative families by child welfare workers
who are seeking a safe home for the child.31 The 1996 federal welfare reform law
encouraged formal relative placement by amending the law to require that child
welfare agencies “consider giving preference to an adult relative over a non-related
caregiver” when determining placement for a child who can no longer safely live
with his/her parents. Relatives are to be deemed “fit and willing” prior to a formal
placement by a child welfare agency, but many children live informally with relative
Although children living with grandparents and other relatives often have needs
comparable to children placed in non-relative foster homes, a TANF child-only case
typically receives far less financial support than do children in foster care. A study
using California’s 1997 rate for “high cost” counties found that a relative caregiver
receiving a TANF child-only grant to care for one child would receive $279 a month
while the basic rate for one foster care child ranged from $345 to $484, depending
on the child’s age, and many counties supplemented this basic rate with additional
payments for specified needs (e.g., clothing allowance). In addition, foster care
families may be eligible for supportive services and are also monitored to ensure a32
child’s safety.

29 Ibid., Table 10:37.
30 Ibid., Tables 10:36 and 10:37.
31 See Amy Billings, Jennifer Ehrle and Katherine Kortenkamp. Children Cared for by
Relatives: What Do We Know About Their Well-Being? May 2002 and Jennifer Ehrle, Rob
Geen, and Rebecca Clark. Children Cared for by Relatives: Who Are They and How Are
They Faring? February 2001 both from the Urban Institute, Washington, D.C.; available
online at [] (click successively on Research, Children, Child Welfare
and then scroll down to the named reports).
32 Jill Duerr Berrick, Barbara Needall and Meredith Minkler. The Policy Implications of
Welfare Reform for Older Caregivers, Kinship Care and Family Configuration. Children
and Youth Services Review, v. 21, no. 9-10, 1999. p. 843-864. See also Jennifer Ehrle and
Rob Geen. Children Cared for by Relatives: What Services Do They Need? Urban
Institute, Washington, D.C., June 2002; available online at [] (click
successively on Research, Children, Child Welfare and then scroll down to the named

A Mathematica survey of child-only cases in New Jersey found that 60% of the
nonparent caregivers reported that the child(ren) they care for did not live with a
parent due to parents’ substance abuse problems; 40% reported that child welfare
services had required that the child not live with the mother and 58% reported that
the child welfare agency had been “involved” with the parents or the child before the
child came to live with them. The Mathematica survey also found that while, on
average, nonparent caregivers had much higher incomes than TANF families that
include a parent in the cash grant, the great majority (70%) have incomes below
200% of the poverty level. The average nonparent caregiver was 52 years of age but
had considerably poorer physical health then residents of that age in the general U.S.
population. Furthermore, despite relatively stable living arrangements, more children
in these nonparent child-only cases scored less well on school achievement and
behavioral measures than did children in TANF cases that included a parent in the
cash aid grant.33
In response to the growing number of relative caregivers and their specific
needs, some states have used TANF funds to establish subsidized guardianship
programs and/or kinship caregiver support groups and services. The advocacy
coalition Generations United reported in June 2001 that as many as 24 statewide
subsidized guardianship programs existed and that the majority access TANF funds
to provide payments. These programs generally allow a non-parent caregiver
(usually a relative) to assume legal custody of a child and to receive a subsidy equal
to or somewhat less than foster care support but typically higher than a child-only
TANF payment. Most states require that a child be under child welfare custody
before a relative caregiver can have access to the subsidized guardianship program,
however, and this significantly limits the scope of service benefit. (FL, LA and MO
are the notable exceptions.) Some states require that the caregiver first access a
TANF child-only benefit before receiving additional assistance, while as many as
nine states simply used TANF funds to operate a subsidized guardianship program.34
Child-only cases with parents. Child-only cases where a parent received
SSI likely existed prior to the 1996 welfare reform law but changes made by the
reform established several new kinds of child-only cases, including families where
parents are ineligible for benefits due to citizenship status or sanctions, and these
have been of particular concern to child welfare advocates. For FY2000, a little less
than half of 782,000 child-only families (46.3%) included a parent who lived in the
household. Of the parents who lived in these 349,000 families, the largest single
group received SSI benefits (41.5%). Parents whose citizenship status was unknown
represented 26.1% of these child-only cases, 11.5% were under benefit sanctions and

21% are categorized as “other/unknown.”35

33 Robert G. Wood and Debra Strong. The Status of Families on Child-Only TANF Cases.
Work First New Jersey Evaluation. Mathematica Policy Research, May 2002. (Hereafter
cited as Wood and Strong, New Jersey Child-only Study, Mathematica, 2002.)
34 Ana Beltran. Grandparents and Other Relatives Raising Children: Subsidized
Guardianship Programs. Fact Sheet, Generations United, Washington, D.C., June 2001;
available online at [].
35 TANF Report FY2000, Table 10:11. Note: Sanctioned parents could also receive SSI

The 1996 welfare reform law severely limited federal aid to legal immigrants
who arrived in the United States after August 16, 1996 and have lived in this country
for fewer than 5 years; citizen children of these immigrants, however, remain eligible
for cash aid. The New Jersey study found that compared to children living in TANF
families that included a parent in the assistance grant, as well as those living in other
kinds of child-only cases, children of immigrant parents who received a TANF grant
were many times more likely to experience extreme poverty (less than 50% of the
federal poverty level). Research concerning the effect of this change on child welfare
is limited. However, analysis of state level data between 1990 and 1998 did not find
an association between the level of substantiated child maltreatment and availability
of benefits to immigrants. At the same time, Urban Institute researchers note that in
5 of the 12 states they surveyed in 1999, workers reported an increase in
maltreatment among immigrant children.36
The 1996 welfare reform law also required states to sanction adults who failed
to meet work and other requirements. Currently states have the option to reduce the
family benefit by the amount of money that would otherwise be paid to the non-
cooperating adult. In these cases a child-only case occurs because of the sanctions.
Some states however require a full family sanction – meaning the entire grant is
ended – and the House has passed welfare reform legislation (H.R. 4737) that would
require full family sanctions in nearly every state.

35 (...continued)
benefits or be in unknown citizenship/alienage status.
36 Wood and Strong, New Jersey Child-only Study, Mathematica, 2002. Paxson and
Wadefogel, Welfare Reforms, Family Resources, and Child Maltreatment, 2001. Geen et
al., Welfare Reform’s Effect on Child Welfare.