Order Code 98-369 EPW
Updated June 7, 2002
CRS Report for Congress
Received through the CRS Web
Welfare Reform: TANF Trends and Data
Vee Burke
Domestic Social Policy Division
Summary
The size and character of the nation’s family cash welfare rolls and the
composition of welfare spending have changed markedly since August, 1996, when
Congress created the time-limited and work-conditioned block grant program of
Temporary Assistance for Needy Families (TANF). Enrollment has plunged more than
50%. The share of recipients who combine welfare and work has risen sharply (26%
in FY2000). The share of “child-only”cases, which are free of work and time limit
rules, has climbed above one-third nationally and in nine states exceeds one-half of all
TANF cases. The smaller caseload holds a rising proportion of black and Hispanic
families. To promote work, TANF programs use tough work sanctions, liberal work
rewards, “Work First” policies, and diversion payments. States accounting for more
than 40% of the nation’s caseload plan to continue aid, with state-only funds, to families
who reach the 5-year limit on federally paid benefits. Many states offer new services
aimed at TANF goals for a broad non-cash welfare population. Before TANF, 75% of
total family welfare outlays were for cash benefits, but in FY2000 cash aid (including
some new kinds of payments) accounted for only 64% of the total. This report will be
updated for new data.
TANF Policy Choices Made by States
P.L. 104-193, which replaced Aid to Families with Dependent Children (AFDC)
with TANF, sets some ineligibility rules and numerous work rules. States may not use
TANF to assist unwed mothers under 18 unless they live in an adult-supervised setting
and, if a high school dropout, attend school. They cannot give federally funded ongoing
basic aid to a family with a member who has received aid for 60 months, and they must
require work, under state definition, as a condition of continued aid after 24 months of
benefits. Otherwise, states are generally free to design their own programs. To avoid loss
of TANF funds, however, the states must engage 50% of most adult recipients in
specified “work activities” for a general average of 30 hours weekly (20 hours for single
parents of preschool children.
Expansionary Policies. More than 30 states have expanded eligibility by one of
these policies: ending special eligibility restrictions for two-parent families, continuing
benefits for those who go to work, and increasing asset limits. Some states made these
Congressional Research Service ˜ The Library of Congress

CRS-2
changes before TANF, under AFDC waivers. About half the states exempt from work
single parents with an infant, and in FY2000 this policy exempted 4% of the caseload (a
monthly average of 94,240 mothers). All but about 11 states have adopted the Family
Violence Option (FVO), which permits exemption from TANF rules for victims of
domestic violence. To reward work, TANF state programs generally disregard a sizable
share of earnings for at least a year. In 15 states TANF benefits do not end until gross
earnings exceed or come very close to the 2001 poverty guideline ($1,220 for a family
of three). Ten states ignore from 20% to 67% of all earnings in all months, and two,
100%, subject to a gross income cap. (However, eight states–three in all months, five
after 4-6 months of work– use flat dollar disregards, under which extra earnings reduce
benefits.) See TANF Benefits and Earnings Limits in the CRS Welfare Reform Electronic
Briefing Book. New Jersey has launched a supplemental work support program ($200
monthly payment for up to 2 years for ex-TANF families with earnings below 250% of
the poverty level), and Montana now offers a one-time Work Support Payment ($494)
for those who leave TANF with jobs.
Restrictive Policies. State TANF policy choices that tend to restrict the caseload
include: durational time limits shorter than outer federal limits, tough sanctions, welfare
avoidance (diversion) payments, and family caps (reduced or zero benefits for a new baby
in a TANF family). Some states adopted these policies under AFDC waivers. Twenty-
two states have benefit cutoff limits shorter than 60 months, and almost a dozen require
work immediately, rather than after 24 months (some defining this to include job search,
some restricting it to a job or community service). Some 19 states penalize recipients for
their first failure to comply with required work activities by suspending the family’s
whole benefit, sometimes until compliance, sometimes for specified periods. Under some
conditions, eight states end benefits permanently (see CRS Report RS21070). Pre-TANF
law required states to continue aid for the child(ren). Twenty-two jurisdictions impose
family caps. Georgia reports that in FY1999, 3,596 children were subject to these caps,
and in December, 2001, North Dakota reported 239 “benefit cap” children. Two states
pay flat benefits for all family sizes. More than 30 states make lump sum payments that
divert some applicants from enrollment. Diversion data: Oklahoma (FY2001) 2,830
families, and (January 2002) 91 families, equal to 4.6% of cases opened; Virginia
(March), 106 cases, 4.6% of approved applications; New Mexico (April), 29 cases, 1.4%
of approved applications; Washington (January;) 305 families, some with 2 adults. In
FY2000, Iowa diverted 304 families from TANF enrollment. (For details of state TANF
programs, see CRS Report RL30695).
Benefit Levels. According to the fourth annual TANF report, cash benefits
averaged $349 per family in FY2000 (and 23% of families had non-TANF income
averaging $580). Of TANF families, 80% received food stamps averaging $224 monthly,
and 18% received subsidized housing. Between July 1996 and July 2001, most
jurisdictions did not change their maximum benefit levels; four reduced benefits (the
District of Columbia, Hawaii, Idaho, and Oklahoma), and 20 raised them, but increases
exceeded the 11.4% rise in consumer prices only in 5 states (Louisiana, Maryland,
Mississippi, West Virginia, and Wisconsin). (See Cash Welfare Benefit Amounts in the
Welfare Reform electronic briefing book.) According to the National Governors
Association, nine states plan to increase benefits in FY2003: Florida, Illinois, Kentucky,
Louisiana, Missouri, Maryland, Montana, South Dakota, and Texas. Two states have
adopted bonuses (Oregon for cooperation with its work program and West Virginia for
marriage). Wisconsin and Idaho no longer adjust benefits for family size. The Wisconsin

CRS-3
Works (W-2) program pays flat monthly amounts of $673 for community service jobs and
$628 for W-2 transitional activities (e.g., sheltered workshops, vocational rehabilitation,
parenting, life skills, job skills training).
TANF Spending. In FY1996, last full AFDC year, spending on AFDC, AFDC-
related child care, Emergency Assistance (EA), and JOBS totaled $30.4 billion ($16.3
billion in federal funds and $14.1 billion in state funds). In FY1998, first full year of
TANF, comparable spending was $21.5 billion (down 29% from 1996), and the average
monthly caseload was down 29% from its 1996 level. Thereafter the cash caseload
continued to shrink (17% in FY1999, 15% in FY2000, and 6% in FY2001), but spending
rose, reaching $21.7 billion in FY1999 and $22.8 billion in FY2000. Composition of
FY2000 spending: cash aid (including diversion payments, refundable tax credits and
IDA contributions) 64%, TANF services, 16%, TANF child care, 10%, TANF work
activities, 10%. Of FY2000 outlays, $12.5 billion (55%) was from federal funds The
annual federal TANF grant is $16.5 billion, and states must spend at least $10.4 billion
of their own funds each year (maintenance-of-effort [MOE] rule) on behalf of TANF-
eligible families. On September 30, 2000, states had an unobligated TANF balance of
$3.2 billion.
Caseload Decline. As Figure 1 shows, the number of AFDC families began
climbing in FY1990, reached a record peak (5.1 million ) in spring 1994 and then turned
downward. Throughout 2001, monthly numbers were relatively stable, rounded at 2.1
million families, 59% below the 1994 record. As of December 2001, enrollment totaled
2.099 million families. Numbers exceeded year-earlier levels in 30 jurisdictions, but the
national caseload still was smaller than in December 2000 (a decline of 54,000 families
in New York accounted for most of the net drop). Food stamp numbers, which have
been rising steadily since April 2001, reached their highest level since June 1998 in
Figure 1. AFDC/TANF Families: October 1976 - December
2001
6 , 0 0 0 , 0 0 0
Ma r c h 1 9 9 4 :
5 , 0 0 0 , 0 0 0
5 .1 millio n f a milie s
4 , 0 0 0 , 0 0 0
3 , 0 0 0 , 0 0 0
2 , 0 0 0 , 0 0 0
De c e mb e r 2 0 0 1 :
2 .1 millio n
1 , 0 0 0 , 0 0 0
0
Oct-76
Oct-78
Oct-80
Oct-82
Oct-84
Oct-86
Oct-88
Oct-90
Oct-92
Oct-94
Oct-96
Oct-98
Oct-00
March (19.2 million persons).
Many factors have helped to shrink the TANF caseload since 1996, including the
“Work First” culture, the improved economy, tougher sanctions, the benefit limit and

CRS-4
widespread adoption of diversion practices. Under TANF, not only have recipients
departed from welfare at a faster rate, but fewer persons have joined the rolls to replace
them. For instance, an Illinois study found that annual TANF entries in the state declined
45% from FY1997 to FY1999, while exits increased by 8%. Florida case closures in
April, 2002 (9,116) outnumbered case entries (8,459) by 8%. California case closures
in December 2001 (42,254) outnumbered case additions (34,880) by 21%.
Rise in Mothers’ Employment. Among families with children that are
maintained by the mother, the share with an employed mother soared from 61.6% in 1994
to 74.7% in 1999, according to the Bureau of Labor Statistics (BLS). During the first
quarter of 2000, the employment rate of mothers maintaining families with children (not
adjusted for seasonal variation) was 75.5%, but in the comparable quarter of 2001 the rate
slipped to 74.9%. In its 2001 annual report, the CEA said that the EITC has increased the
probability of employment for low-income persons, and it cited an estimate that the EITC
was responsible for 34% of the rise in annual employment among unmarried mothers
between 1992 and 1996. The poverty rate of children in female-headed families dropped
from 52.9% in 1994 to 46.1% in 1998, 41.9% in 1999, and 39.8% in 2000. (For all
persons in female-headed families with children, the 2000 official poverty rate was
35.1%).
Time Limits. The General Accounting Office (GAO) estimates that states
excluded 11% of adult recipients of TANF cash aid from the federal time limit, as of
autumn 2001, by three means: paying certain recipients with state-only funds (and
thereby stopping the federal time clock), continuing pre-TANF waivers that permit time
limit exemptions, and granting hardship exemptions (see GAO-02-501T). TANF’s 5-year
anniversary marks the earliest date that a family could accumulate 60 months of federally
paid benefits, and it varies by state, ranging from October 1, 2001 to July 1, 2002. Eight
states containing more than 40% of the 1999 national caseload have decided to use state
funds to continue benefits beyond 60 months (this spending can be credited toward
required MOE expenditures). Maine, Michigan, Maryland, New York, and Vermont
continue full benefits (generally in noncash form in New York), and Arizona (under a
waiver expiring in September), California, and Rhode Island pay reduced benefits.
Michigan reported in December, 2001, that 4,566 families (5% of the caseload) had
reached the 60-month limit. Families in 20 states are subject to state-imposed limits
shorter than 60 months. See CRS Report RS21069. In these states, families often
receive extensions of time. At the end of March 2002, for example, 2,789 Connecticut
families who had exhausted their original 21 months of aid received 6-month benefit
extensions because they had been unable to obtain work that paid more than TANF
(another 446 families lost eligibility because of the time limit). In April 2002, 1,699
Florida families who had reached the state’s time limit (24 or 36 months) received
extended benefits, some because of unsubsidized employment (under which they earned
a one-month TANF extension for each month of work).

Characteristics of TANF Families
(For details and state data, see the fourth annual TANF report to Congress at
[http://www.acf.dhhs.gov/programs/opre/ar2001/indexar.htm].)
Marital Status, Race/Ethnicity. In FY2000, 59.3% of TANF adult recipients
had never married; 18.8% were married and living together; 12.8% had married, but were

CRS-5
separated, 8.5% were divorced, and 0.8%, widowed. Compared with FY1996, the share
of TANF children who were Hispanic or black rose, and the share who were white
declined in FY2000: white, 26.8% (down from 31.6% in FY1996); black, 40.1%
(compared with 38.4%); Hispanic, 26.8% (22.4%). Asians represented 2.8% of TANF
children; American Natives, 1.6%; other, 0.6%; and unknown race, 1.3%.
Child-only Cases. In child-only cases, which are free of TANF work rules and
time limits, the parent or other relative caretaker is ineligible because of being non-needy,
an illegal immigrant, under sanction, or for some other reason (or, though eligible,
declines TANF). During FY2000, 34.5% of TANF families had no adult recipient. This
compares with 11.6% in FY1990, 21.5% in FY1996, and 29% in FY1999. In nine states,
more than one-half of TANF families in FY2000 had no adult recipient (Alabama,
Florida, Idaho, Mississippi, North Carolina, South Carolina, South Dakota, Wisconsin,
and Wyoming). The composition of child-only cases (by status of the caretaker) varies
among states. Examples: Florida (April 2002) non-needy caretaker relative, 63%; SSI
recipient, 23%; non-citizen, 9%; sanctioned, 1.5%, and other, 3.5%; California (FY1999)
non-needy caretaker relative, 11%; SSI recipient, 16%; non-citizen, 44%; sanctioned,
12%; and other, 17%; and Nevada (January 2002), non-needy caretaker, 75%; SSI
recipient, 9%; non-citizen, 13%, and family preservation plan case, 3%.
Recipients with Jobs. Under TANF there has been a sharp increase in welfare
plus work. In FY2000, 26% of welfare adults were employed (in paid jobs), more than
double the FY1996 rate of 11% (but below the FY1999 rate of 28%). Nine states had
TANF adult employment rates of one-third or higher in FY2000: California, Connecticut,
Delaware, Hawaii, Maine, Michigan, Minnesota, Virginia, and Washington. Available
state data provide more recent employment data. In March 2002, Connecticut says that
17% of TANF families had earnings averaging $7.39 hourly ($727 monthly). This
included 32% of families in Jobs First, which imposes a 21-month benefit cutoff.
Percentages of total TANF families (including no-adult families) with earnings in some
other states: Michigan (January), 32% (57% of those required to work); Florida (April)
4.9% (14.5% of work-required group); Pennsylvania (February), 19% (34% of work-
required group); Oregon (February), 1.7% of 1-parent families and 5.2% of 2-parent
families; Illinois (May), 19% (37% of work-required group); Indiana (January) 3.1%
(8.4% of the work-required group); and Virginia (March), 19% (57% of work-required
group). Some state reports provide work exemption data. During March, Connecticut
exempted 5,101 adults and minor parent household heads from work requirements (out
of a caseload of 23,171), more than half because of caring for a child under age one; in
February, Pennsylvania exempted 27% of adult recipients from work requirements, most
because of disability; in April, Florida exempted 10.2% of adult recipients from work,
more than half because of having a child under 3 months old; and in April, 2002,
Massachusetts (under a waiver lasting until September 30, 2005) exempted 91% of the
caseload from work requirements (and 73% from time limits).
Sanctioned Recipients. The fourth annual TANF report attributed 6.5% of
TANF case closings in FY2000 to sanctions (3.9% were sanctions related to work; 2.6%
were sanctions related to child support, teen parent requirements, or failure to meet an
individual responsibility plan). Sanctions accounted for more than one-fifth of closures
in four jurisdictions: Florida, 26.9%; Idaho, 21.2%; Mississippi, 42% and Oklahoma,
30.9%. However, 10 jurisdictions reported no FY2000 case closures attributed to
sanctions: Colorado, Maine, Minnesota, Pennsylvania, Puerto Rico, Rhode Island, Texas,

CRS-6
Vermont, Virgin Islands, and Washington). More recent available state data on sanctions:
In FY2001, sanctions caused closure of 7,352 cases in Oklahoma (28% of all closures).
In January 2002, the proportion of Oklahoma case closures attributed to TANF work
refusal/failure was 40.5%. In April, Florida suspended parental benefits of 524 families
(333 for work violations and 191 for other reasons); the children’s benefits continued.
Under TANF sanction in January were 1,497 Nevada adults (5% of the total number).
During March sanctions caused 3.8% of closures in Connecticut.
TANF Services for An Expanded Population
States are free to use TANF dollars to offer benefits and services to families
ineligible for ongoing cash aid, provided the services promote a TANF goal. In addition
to traditional welfare goals of helping needy children in their own homes and promoting
self-sufficiency of needy parents, TANF seeks to reduce out-of-wedlock pregnancies and
promote formation and maintenance of two-parent families. In their 2000-2001 plans, 40
states describe special programs for at-risk families (some with income limits up to 250%
of poverty) and 31 indicate they will offer special services for non-custodial parents.
New York in September 2000 issued a list of 28 “State TANF 200% Programs” for
persons and families with incomes up to 200% of the poverty level. The list includes
technology training, home visiting, alternatives to incarceration, support services for
caretaker relatives, foster care worker recruitment/retention, summer youth employment.
(The TANF time limit applies only to basic ongoing aid and support services for the
unemployed, not to services generally. Thus, states can use federal funds to help working
poor families without exposing them to time limits.) The single largest TANF-funded
service is child care (federal/state outlays estimated at $4.3 billion in FY2000). Estimated
federal/state TANF spending in FY2000 on various support, rehabilitative, and preventive
services totaled $3.7 billion. Services provided by these funds included transportation
subsidies, parental skill building services, home energy aid, responsible parenthood
counseling, family planning, and numerous others.
Findings about Welfare “Leavers”
State studies explore the circumstances of families who have left welfare: reasons
for departure, employment and earnings, and returns to the rolls. More than 30 leaver
studies completed as of January 2000 show that most families who left TANF or AFDC
waiver programs between 1995 and 1998 did so because of employment. In the quarter
after exit, administrative data indicate that employment rates ranged from 50% to 64%,
but rose to 66%-86% when based upon employment at any time since exit. Studies using
survey data found generally higher employment rates and a broader range, from 35% to
83%, at the time of the survey. Among welfare leavers who worked, survey data indicate
that average hourly wages ranged from $5.50 to $8.16. Annualized (on the basis of
reported weekly or quarterly work hours) their earnings exceeded cash welfare guarantees
in all 22 states studied. For all these leavers, year-round work paid more than welfare.
However, in all but six of the states, income from wages alone would leave a family of
three below the poverty level. Within 1 year of exit or at the time of the leaver study,
from 13% to 36% of leavers returned to welfare. This count excludes “churners,” persons
who returned to welfare almost immediately, within 1 or 2 months of exit. (For a final
synthesis report of 15 HHS-funded studies about welfare leavers, see
[http://aspe.hhs.gov/search/hsp/leavers99/synthesis02/index.htm].)