Welfare Reform: Early Fiscal Effects of the TANF Block Grant (Letter
Report, 08/18/98, GAO/AIMD-98-137).

Pursuant to a congressional request, GAO reviewed the states' fiscal
decisions for the Temporary Assistance for Needy Families (TANF) block
grant and whether states are taking steps to prepare for the effects of
future economic downturns on their welfare programs, focusing on: (1)
how state budgetary resources, including federal aid, have been
allocated since states have had access to TANF funds; (2) what plans
states are making to ensure programmatic stability in times of fiscal
and economic stress; and (3) the extent to which states have used, or
plan to use, the program's federal Contingency Fund for State Welfare
Programs and the Federal Loans for State Welfare Programs (Loan Fund)
which are available for downturns or other emergencies affecting states.

GAO noted that: (1) more federal and state resources are available for
states' low-income family assistance programs since welfare reform
passed in 1996 than would have been available under the previous system
of financing welfare programs consolidated in the TANF block grant; (2)
GAO's estimates showed that, taking caseload declines into account, 46
states would have more total resources--both state and federal--for
their low-income family assistance programs than they would have had
under the previous welfare programs; (3) states are transforming the
nation's welfare system into a work-focused, temporary assistance
program for needy families and generally chose to spend these resources
to expand programs and benefits by shifting the emphasis from
entitlement to self-sufficiency, enhancing support services, and
increasing work participation rates; (4) states also have achieved
budgetary savings by reducing state funds to the statutory
maintenance-of-effort level of 75 or 80 percent of previous state
spending levels; (5) while states have gained greater resources under
the block grant, they also take greater responsibilities for fiscal
risks should program costs increase in the future; (6) most states,
including 7 of the 10 GAO visited, also have a general fund budget
stabilization or rainy day funds that could be used to augment program
spending during an economic downturn, but welfare programs would have to
compete for these resources with other state funding priorities; (7) the
Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA)
includes features that could provide federal funding to cover future
increases in program needs; (8) states can carry forward unused TANF
funds without fiscal year limitation; (9) as of September 30, 1997,
states had left about $1.2 billion in unspent balances in their accounts
with the U.S Treasury, or about 9 percent of the total grant; (10) it is
unclear whether these balances will remain, shrink, or increase as
states gain experience with the problem; (11) PRWORA also creates two
federal safety-net mechanisms--the Contingency Fund and the Loan
Fund--that were designed to provide states with access to additional
funds during times of economic downturn or fiscal stress; (12) as of
February 1998, neither the Contingency nor the Loan Funds had been used
by any state; (13) officials in states GAO visited said they did not
view the Contingency Fund as a viable source of additional resources;
and (14) officials in many states GAO visited indicated they did not
believe their states would borrow from the Loan Fund during an economic
downturn.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  AIMD-98-137
     TITLE:  Welfare Reform: Early Fiscal Effects of the TANF Block Grant
      DATE:  08/18/98
   SUBJECT:  State-administered programs
             Welfare benefits
             Workfare
             Public assistance programs
             Funds management
             Federal aid to states
             State budgets
             Intergovernmental fiscal relations
             Block grants
IDENTIFIER:  HHS Temporary Assistance for Needy Families Program
             Aid to Families with Dependent Children Program
             AFDC
             Contingency Fund for State Welfare Programs
             Federal Loans for State Welfare Programs
             California
             Connecticut
             Louisiana
             Maryland
             Oregon
             Texas
             Wisconsin
             Colorado
             Michigan
             New York
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to the Chairman, Subcommittee on Human Resources, Committee on
Ways and Means, House of Representatives

August 1998

WELFARE REFORM - EARLY FISCAL
EFFECTS OF THE TANF BLOCK GRANT

GAO/AIMD-98-137

Welfare Reform

(935235)


Abbreviations
=============================================================== ABBREV

  AFDC - Aid to Families With Dependent Children
  AFDC-UP - Aid to Families With Dependent Children Unemployed Parent
     program
  APWA - American Public Welfare Association
  CBO - Congressional Budget Office
  CCDBG - Child Care and Development Block Grant
  CMIA - Cash Management Improvement Act of 1990
  EA - Emergency Assistance
  FMAP - Federal Medical Assistance Percentage
  HHS - Department of Health and Human Services
  IRCA - Immigration Reform and Control Act of 1986
  JOBS - Job Opportunities and Basic Skills
  MOE - Maintenance of Effort
  NASBO - National Association of State Budget Officers
  NCSL - National Conference of State Legislatures
  NGA - National Governors' Association
  OMB - Office of Management and Budget
  PRWORA - Personal Responsibility and Work Opportunities Act of 1996
  SLIAG - State Legalization Impact Assistance Grants program
  SSBG - Social Services Block Grant
  TANF - Temporary Assistance to Needy Families
  TEA-21 - The Transportation Efficiency Act for the 21st Century
  UI - Unemployment Insurance program

Letter
=============================================================== LETTER


B-277800

August 18, 1998

The Honorable E.  Clay Shaw, Jr.
Chairman, Subcommittee on Human Resources
Committee on Ways and Means
House of Representatives

Dear Mr.  Chairman: 

In 1996 the Congress made sweeping changes to national welfare policy
in passing Public Law 104-193, the Personal Responsibility and Work
Opportunity Reconciliation Act (PRWORA).  This act creates the
Temporary Assistance for Needy Families (TANF) block grant,\1 a fixed
federal funding stream that replaces the Aid to Families with
Dependent Children (AFDC) program and a number of welfare programs in
which federal funding matched state spending and increased
automatically with caseloads. 

This report is part of a series of GAO work reviewing the
implementation of federal welfare reform.  In this report, we focus
on the fiscal aspects of welfare reform.  You asked us to review
states' fiscal decisions for the TANF block grant and to address
whether states are taking steps now to prepare for the effects of
future economic downturns on their welfare programs.  Specifically,
you asked us to report on (1) how state budgetary resources,
including federal aid, have been allocated since states have had
access to TANF funds, (2) what plans states are making to assure
programmatic stability in times of fiscal and economic stress, and
(3) the extent to which states have used, or plan to use, the
program's federal Contingency Fund and the Loan Fund which are
available for downturns or other emergencies affecting states. 

In a related report, we responded to a request from you and the
Chairman of the Senate Committee on Finance to monitor states'
efforts to implement programs to meet the stated objectives of Title
I (TANF) of PRWORA.  In that report, we described states' efforts to
require and encourage welfare recipients and potential recipients to
assume greater personal responsibility, examined how states are
providing services to support the objectives of TANF, and reviewed
early reported data to assess states' progress.\2


--------------------
\1 Aid to Families with Dependent Children (AFDC), Job Opportunities
and Basic Skills (JOBS), and Emergency Assistance (EA) programs and
the associated administrative expenses were consolidated in the TANF
block grant whereas the AFDC-related child care program and other
child care programs were combined in the Child Care and Development
Block Grant (CCDBG).  In this report, we focus on those programs
consolidated in the TANF block grant. 

\2 Welfare Reform:  States Are Restructuring Programs to Reduce
Welfare Dependence (GAO/HEHS-98-109, June 18, 1998). 


   BACKGROUND
------------------------------------------------------------ Letter :1

Currently authorized through federal fiscal year 2002, the TANF block
grant represents an entitlement to states of $16.5 billion
annually.\3 Federal funding under the TANF grant is fixed, and states
are required to maintain a significant portion of their own historic
financial commitment to their welfare programs as a condition of
receiving their full TANF grant--referred to as their maintenance of
effort requirement (MOE).\4 These two funding streams--federal TANF
and state general funds for MOE--represent the bulk of the resources
available to states as they design, finance, and implement their new
low-income family assistance programs.  (See figure 1.) Under TANF,
states have the flexibility to design their own programs and
strategies for promoting work over welfare and self-sufficiency over
dependency.  At the same time, states must meet federal requirements
that emphasize the importance of work for those receiving assistance. 
To avoid federal financial penalties, in fiscal year 1997 states must
ensure that 25 percent of their TANF families, rising to 50 percent
in 2002, are engaged in work activities.  In addition, the law
prohibits the use of TANF funds to provide assistance for families
with adults who have received assistance for more than 5 years. 

   Figure 1:  Funding for State
   Low-Income Family Assistance
   Programs

   (See figure in printed
   edition.)

In addition to giving states more responsibility and flexibility in
the design of welfare programs, TANF shifts the fiscal risk to
states, thus highlighting the importance of fiscal planning,
especially contingency budgeting.  In the past, any increased costs
were shared by the federal government and the states.  Under TANF,
however, if costs rise, states face most of the burden of financing
the unexpected costs.\5 States must also handle this responsibility
in the context of any limitations--including legislative
restrictions, constitutional balanced budget mandates, or conditions
imposed by the bond market--on their ability to increase spending,
especially in times of fiscal stress. 

States have various options and resources to help them handle this
new fiscal responsibility.  PRWORA provides states with the ability
to save an unlimited amount of their TANF block grant funds for use
in later years.  These resources must be left in the U.S.  Treasury
until they are needed.  States may also respond to the fiscal risks
implicit in the new block grant environment by increasing the levels
of their state "rainy day funds,"\6 or by establishing dedicated
reserves, consisting of state funds, for their welfare programs. 
PRWORA also creates two safety-net mechanisms for states to access
additional federal resources in the event of a recession or other
emergency--the $2 billion Contingency Fund for State Welfare Programs
(Contingency Fund)\7 and a $1.7 billion Federal Loan Fund for State
Welfare Programs (Loan Fund).\8


--------------------
\3 In this report we refer to the fixed entitlement to states (State
Family Assistance Grant) as the TANF grant. 

\4 States' MOE requirements are based on their own spending in
federal fiscal year 1994 on AFDC, JOBS, Emergency Assistance (EA),
related administrative costs, and AFDC-related child care programs: 
AFDC/JOBS child care program, Transitional Child Care, and At-Risk
Child Care programs.  A state that does not meet PRWORA's work
participation rates must maintain at least 80 percent of its MOE.  A
state that meets its work participation rate must maintain at least
75 percent of its MOE.  Preliminary data show that 34 states spent at
least 80 percent of their MOE in fiscal year 1997, with 21 states
spending above 80 percent.  A recent National Conference of State
Legislatures' report on states' enacted appropriations for state
fiscal year 1998 shows that 37 plan to spend at least 80 percent of
MOE, with 30 states planning to spend more than 80 percent. 

\5 The Congress, however, could approve supplemental appropriations
to help states address unanticipated funding deficiencies.  The
federal government, therefore, has an interest in the extent of
contingency budgeting for recessions and other emergencies. 

\6 Forty-five states report having some type of rainy day fund. 
State budgetary distress associated with economic downturns is often
addressed through the use of these funds.  These reserves are often
used to correct short-term imbalances between revenue and
expenditures.  According to the National Association of State Budget
Officers (NASBO), general fund "ending balances" and rainy day fund
balances were estimated to exceed $24 billion or 6 percent of total
state expenditures in 1998. 

\7 The Adoption and Safe Families Act of 1997 reduces the Contingency
Fund for State Welfare Programs by $40 million over the remaining 4
years the fund is authorized.  (P.L.  105-89, 404, 111 Stat.  2134.)

\8 Loans from this fund are to have a maturity of no more than 3
years at an interest rate comparable to the average market yield on
outstanding marketable federal obligations with comparable periods to
maturity. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :2

To address the objectives for this report, we collected fiscal
information on all 50 states from the U.S.  Department of Health and
Human Services (HHS).  In addition, we interviewed officials from
state budget offices and reviewed state general fund budgets and
low-income family assistance budgets.  We selected the seven states
examined in GAO's parallel report, Welfare Reform:  States are
Restructuring Programs to Reduce Welfare Dependence
(GAO/HEHS-98-109)--California, Connecticut, Louisiana, Maryland,
Oregon, Texas, and Wisconsin.  For this report, we added Colorado,
Michigan, and New York to enrich the discussion of states' efforts to
budget for contingencies.  These 10 states represent 53 percent of
total program dollars and administer about half the nation's
caseload.  We did not independently verify the reported levels of
state spending nor whether reported federal or state spending met the
qualifications set forth in the act.  We conducted our fieldwork from
April 1997 through February 1998 in accordance with generally
accepted government auditing standards.  (For more detail on our
scope and methodology, see appendix I.)


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :3

More federal and state resources are currently available for states'
low-income family assistance programs since welfare reform passed in
1996 than would have been available under the previous system of
financing welfare programs consolidated in the TANF block grant.  Our
estimates showed that, taking caseload declines into account, 46
states would\9 have more total resources--both state and federal--for
their low-income family assistance programs than they would have had
under the previous welfare programs.  These additional resources
result from a combination of significant declines in caseloads, the
new fixed-level financing mechanism created by the TANF block grant,
and the maintenance-of-effort requirement on the state share.  Both
the level of the TANF block grant and the state MOE are fixed and
were set based on time periods when caseloads were high.\10

States are transforming the nation's welfare system into a
work-focused, temporary assistance program for needy families and
generally chose to spend these resources to expand programs and
benefits by shifting the emphasis from entitlement to
self-sufficiency, enhancing support services, and increasing work
participation rates.  States have taken advantage of the flexibility
allowed under PRWORA to use federal and state funds to help
recipients, those who are moving off public assistance, and those at
risk of coming on the rolls to obtain child care, transportation
subsidies, and training to promote job placement.  States also
achieved budgetary savings by reducing state funds to the statutory
MOE level of 75 or 80 percent of previous state spending levels. 
This, however, constituted a higher level of spending per recipient
in 21 states because of declining caseloads.  In addition, four of
the states we visited have set aside a portion of the federal and
state resources in reserve for future contingencies.  States
generally have leveraged the higher federal grant levels and employed
a combination of these options; the result was a greater proportion
of federal resources in their programs. 

While states have gained greater resources under the block grant,
they also take greater responsibilities for fiscal risks should
program costs increase in the future.  In their budget choices,
states struck a balance between investing in today's caseload and
saving for tomorrow's.  At the outset, 5 of the 10 states we visited
planned to use all available federal TANF funds to invest in programs
to promote self-sufficiency with expectations that this would reduce
future caseload growth.  In four of the other states, however, a
portion of their federal and state resources were set aside to cover
potential future caseload increases and costs.\11

Officials in these states were concerned that additional resources
would be difficult to obtain in the future if the economy were to
worsen and if costs were to rise unexpectedly.  Most states,
including 7 of the 10 we visited, also have general fund budget
stabilization or "rainy day" funds that could be used to augment
program spending during an economic downturn, but welfare programs
would have to compete for these resources with other state funding
priorities. 

The act includes features that could provide federal funding to cover
future increases in program needs.  States can carry forward unused
TANF funds without fiscal year limitation.  As of September 30, 1997,
states had left about $1.2 billion in unspent balances in their
accounts with the U.S.  Treasury, or about 9 percent of the total
grant.\12 However, this may be less an indicator of state contingency
planning and more a reflection of the transitional nature of the
states' first year with welfare reform.  It is unclear whether these
balances will remain, shrink, or increase as states gain experience
with the program. 

The act also creates two federal safety-net mechanisms--the
Contingency Fund and the Loan Fund--that were designed to provide
states with access to additional funds during times of economic
downturn or fiscal stress.  Officials in states we visited said they
did not view the Contingency Fund as a viable source of additional
resources.  This is largely because a more narrowly defined range of
state funding is used to permit a state to access the Contingency
Fund than is required under the TANF MOE\13 and because the Fund
provides only limited federal matching funds for every additional
dollar spent by the state.  Similarly, officials in many states we
visited indicated they did not believe their states would borrow from
the Loan Fund during an economic downturn. 


--------------------
\9 Because states made the transition to TANF at different points
during federal fiscal year 1997, our calculations reflect the amount
states would have been eligible to receive for the full fiscal year
rather than the amount they actually received.  (See appendix II for
a more detailed discussion of this topic.)

\10 As stated in footnote 4, states' MOE requirements are based on
their own spending in federal fiscal year 1994 for the programs
replaced by the TANF grant and combined in the CCDBG.  The level of
the TANF grant is set based on the higher of federal spending on the
programs consolidated in TANF for federal fiscal year 1994, fiscal
year 1995, or the average for the years 1992 through 1994 periods
during which caseloads and federal spending were at historically high
levels. 

\11 Louisiana did not establish a specific reserve fund for its
welfare programs; however, the state did not plan to spend $31
million--or about 20 percent--of its TANF block grant in its 1998
state fiscal year (July 1, 1997, to June 30, 1998). 

\12 As of September 30, 1997, states had expended $9.9 billion, or
about 74 percent of the $13.3 billion in TANF funds awarded to them
for federal fiscal year 1997; made obligations but not yet outlayed
about $1.7 billion, or 13 percent, of the remaining funds; and
transferred almost $0.5 billion, or about 4 percent, to the Social
Services Block Grant (SSBG) and CCDBG, as permitted by PRWORA.  The
remainder, $1.2 billion, represents levels of TANF funds that states
had neither drawn down nor obligated.  As part of their financial
reporting to HHS, states report their balances of unobligated TANF
funds remaining at the U.S.  Treasury to HHS; however, these balances
reflect obligations from the states' perspective.  From the federal
perspective when the TANF grants are awarded to the states they
become obligated whether the state has spent these funds or not.  To
avoid confusion we will refer to these balances throughout the report
as "unspent" TANF funds. 

\13 To be eligible for the Contingency Fund, a state must certify
that its current year TANF program expenditures for the full year in
which it draws from the fund are at least 100 percent of what the
state spent in fiscal year 1994 on its low-income family assistance
programs excluding child care-related programs.  In addition, state
spending on certain state-only funded programs and child care-related
programs that count towards meeting the basic TANF MOE does not count
towards meeting the Contingency Fund MOE. 


   TANF SHIFTS FISCAL
   RESPONSIBILITY TO THE STATES
------------------------------------------------------------ Letter :4

The act made sweeping changes to the nation's cash assistance program
for needy families with children and eliminated a family's
entitlement to federal assistance.  These reforms gave states
flexibility to design their own programs and strategies for acheiving
program goals, including how welfare recipients would move into the
workforce.  The act changed the way in which federal funds flow to
states for welfare programs.  Under the old system of financing,
matching grants provided states with resources to implement federal
welfare programs.\14

The federal match was largely open-ended so that if a state
experienced caseload and related cost increases, federal funds would
increase with state funds to cover expenditures for the entire
caseload.  This open-ended federal commitment provided that financing
for every dollar spent on these programs was shared between the
federal government and the states, thereby limiting the states'
exposure to escalating costs.\15

In contrast, under the TANF block grant, the federal government
provides a fixed amount of funds regardless of any changes in state
spending or the number of people the programs serve.  During periods
when caseloads are decreasing, federal funds per recipient will be
higher under TANF than under the old program.  Conversely if
caseloads and costs increase, federal funds per recipient would be
lower.  A state would then be presented with several options,
including using resources previously saved for contingencies,
reallocating budgetary resources to maintain program stability,
reducing program benefits and/or services to ensure that previously
allocated resources go further, or raising additional revenues
through taxes or fees. 

PRWORA allows states more choices concerning the mix of services they
can offer and the people they can serve, and these choices are likely
to be affected by differences in the rules regarding the use of TANF
and MOE funds.  For example, state MOE funds may be used with more
flexibility than TANF funds.  TANF grant funds may be used for cash
assistance, child care assistance, work placement programs,
subsidized work programs and other efforts not specifically
prohibited by PRWORA.  MOE funds can be used not only for these
purposes but also to provide benefits to some recipients excluded
from TANF assistance.\16

States make these budgetary decisions as part of their regular
appropriations process.  Since any unspent TANF funds remain
available to states without fiscal year limitation, a decision to
dedicate a portion of these funds for a future contingency represents
one aspect of a state's program budgeting under welfare reform. 

States have modified their policies to require and encourage welfare
recipients and potential recipients to adopt behaviors that
facilitate becoming more self-sufficient.  For example, in our
recently issued report on state program restructuring,\17 we found
that the proportion of recipients assigned to job placement
activities--as opposed to education or training activities--was
substantially higher in 1997 than in 1994.  Furthermore, as states
seek to expand the number of adults participating in work activities,
they have generally expanded the roles of welfare workers to better
support the work focus of their programs.  These workers' new
responsibilities vary but include such tasks as motivating clients to
seek work, exploring the potential for welfare diversions, and
collecting more information about applicants and recipients to
determine what they need to facilitate self-sufficiency.  States are
also expanding their programs to help families address barriers to
employment.  For example, these states are using a range of
approaches to help recipients obtain reliable transportation, such as
providing funding for rural transportation systems, enlisting
volunteers to provide transportation for recipients, and providing
funds for vehicle repairs. 

Substantial declines in welfare caseloads and increases in the number
of welfare recipients finding jobs provide signs of early progress. 
For example, caseloads have dropped on average about 20 percent since
1996, when PRWORA was enacted, and by one-third since 1994.  However,
questions remain about what will happen over the long term to
families that no longer rely on cash assistance but continue to need
other kinds of assistance to maintain their employment and about how
much these services will cost.  Furthermore, as more families leave
welfare programs, states may face increasing challenges in serving
increasing proportions of long-term recipients with multiple barriers
to employment.  This adds to the uncertainty surrounding the resource
needs of low-income family assistance programs in the future. 

Even if TANF caseloads continue to decrease over time, they may
become more volatile.  Such caseload volatility may put the states at
greater risk for budgetary stress than did previous matching grant
programs.  We noted in our recent report on states' efforts to
restructure their welfare programs that although states have had a
great deal of success implementing welfare reform programs in a
strong economy, little is known about how a poor economy will affect
their programs.  It is possible that caseloads may prove more
volatile under the new system than under the old.  This is because
the greater emphasis on work implies a tighter link to the state of
the job market and hence to the economy.  Although research on
pre-reform caseloads found varying degrees of correlation between the
economy (as measured by unemployment rates) and the AFDC
single-parent family caseload, it generally found strong correlation
between the economy and changes in the smaller AFDC-unemployed parent
(AFDC-UP) family caseload.\18 This difference has been attributed to
the AFDC-UP caseload's stronger connection to the labor market. 
Since the new TANF grant emphasizes work-related activities, TANF
caseloads may act more like the AFDC-UP caseload rather than the
single-parent AFDC caseload and hence be more closely aligned to the
economy.  Alternatively, some analysts suggest that future caseloads
under TANF may not be as susceptible to economic downturns because
states are beginning to place much greater emphasis on strengthening
labor force ties.  For example, many states have substantially
increased the levels of resources allocated to job training, child
care, and transitional medical care.  These efforts may make the
former welfare population less susceptible to being the first to be
laid off in the event of an economic downturn than the AFDC-UP
population was.  Those who remain on the rolls may be a smaller as
well as more stable population.  These differing perspectives
highlight the uncertainties that states will face as they implement
and finance their new welfare programs and thus highlight the
importance of budgeting for contingencies. 

Budgetary stress caused by caseload volatility may be compounded by
the limitations placed on most states by constitutional or statutory
requirements to balance their general fund budgets.\19 For example,
if revenues fall during an economic downturn, a state's enacted
budget can fall into deficit.  State balanced budget requirements
often motivate states both to reallocate resources within their
budgets and cut program spending during recessions.  The need to cut
spending can be alleviated if a state has accumulated surplus
balances in "rainy day" funds.  These surpluses may be used to cover
a given year's deficit.  However, unless there are reserves
specifically earmarked for low-income family assistance programs,
these programs will have to compete for "rainy day" fund resources
with all other programs in a state's general fund in times of
budgetary stress.  These factors together--the likelihood of
increased volatility and the limited budgetary flexibility available
during an economic downturn--point to the importance of state
contingency budgeting. 


--------------------
\14 Under the AFDC program, the federal contribution to program costs
ranged from 50 percent to 79 percent--based on a state's per capita
income--while administrative costs were shared at a uniform rate of
50 percent.  Activities funded under the JOBS program were matched at
higher rates, but the federal contribution was capped and allocated
based on the number of AFDC recipients in each state. 

\15 In some states, including three of those we visited, local
governments also provide funding for low-income family assistance
programs. 

\16 Among other restrictions, TANF funds may not be used to provide
benefits or services to any family with an adult who received TANF
assistance for more than 60 months (cumulative lifetime total) nor to
women under the age of 18 who have not received their high school
diploma or equivalent unless they are attending school.  States can
use their own funds to provide services to these and certain other
types of recipients who do not meet all requirements for TANF
assistance; states may count these expenditures towards their MOE
requirements. 

\17 Welfare Reform:  States Are Restructuring Programs to Reduce
Welfare Dependence (GAO/HEHS-98-109, June 18, 1998). 

\18 Eligibility for AFDC-UP was limited to those families in which
the principal wage earner (a) was unemployed but had a history of
work or (b) worked fewer than 100 hours per month.  AFDC-UP caseloads
comprised 7 percent of the total AFDC caseload in 1995. 

\19 General funds are those state funds into which general revenue
receipts are credited and from which discretionary programs are
funded.  In addition, states operate other funds such as capital
funds, enterprise funds, and trust funds.  Although states may be
required to balance these funds as well, they do not necessarily
follow the federal approach to measuring deficit or surplus (that is,
matching current year receipts to current year expenditures).  For
more information, see Balanced Budget Requirements:  State
Experiences and Implications for the Federal Government
(GAO/AFMD-93-58BR, March 26, 1993). 


   STATES HAVE ADDITIONAL
   BUDGETARY RESOURCES TO FINANCE
   THEIR LOW-INCOME FAMILY
   ASSISTANCE PROGRAMS
------------------------------------------------------------ Letter :5

A combination of the decline in caseload levels, the higher federal
grant levels, and the MOE requirement for states' contributions to
their programs means that most states have more budgetary resources
available for their low-income family assistance programs since
enactment of welfare reform than under prior law.  In many states,
caseloads began to decline even before the enactment of PRWORA. 
Since enactment, this trend has continued in all states except Hawaii
and in many cases the trend has accelerated.  (See appendix II for a
more detailed discussion and caseload data.)

The amount of each state's block grant was based on amounts received
by the state in 1994 and 1995, years when caseloads and spending were
at historic highs.  As a result, we calculated that 45 states were
eligible to receive more in federal fiscal year 1997 for the TANF
block grant than they received in 1996 under the previous welfare
programs.  (See appendix II for a more detailed discussion of the
assumptions used in our estimates and our analytical techniques.) We
estimated that if all states had drawn their entire 1997 TANF grant,
the states would have received about $1.4 billion more under TANF
than they received under previous welfare programs in 1996, when
caseloads were much higher.\20 It is important to note, however, that
there is a great deal of disparity among the states in the levels of
additional resources.  These differences ranged from 70 percent more
federal resources for Indiana to 7 percent less for
Pennsylvania--with the median increase about 9 percent for all 50
states. 

Furthermore, states are required to maintain a significant portion of
their own historic financial commitment to these programs.  Like the
TANF grant, this minimum MOE is fixed and does not depend on the
number of people served or the types of services a state chooses to
provide.  The MOE requirement establishes a minimum, or floor, for
state spending, but there is no federal ceiling on how much a state
can spend.  States face severe fiscal penalties if they do not meet
their MOE requirement.\21 The interaction between the MOE and the
lower caseloads means that in 21 states, although total state
spending went down, spending per recipient increased.  For example,
in Idaho the MOE requirement is about 28 percent lower than what the
state spent in 1996; however, state spending per recipient will more
than double from $870 to $1,849 per year. 

Aside from the nominal changes in funding, another way of viewing
resources available for welfare is to compare total federal and state
resources available under the block grant with what comparable
federal-state spending would have been for 1997 caseloads under prior
law.  Overall, we calculated that under the block grant, 46 states
would\22 have more total resources--state and federal--for their new
welfare programs than they would have had under the old welfare
programs--with a median increase of 22 percent--or about $4.7 billion
more nationwide.  This calculation represents the difference between
states' post-reform total budgetary resources--TANF plus MOE--and
what they would have budgeted for their 1997 caseloads if they were
still using the pre-reform cost structure.  These differences are
largely attributable to the change in financing mechanisms:  total
funding under the previous program was based on caseload, whereas
under TANF, funding is based on federal and state spending levels in
a prior period when caseloads were higher.  (See appendix II for
further discussion of the assumptions and analytical techniques used
in our estimates.) Again, there was great variation among the 46
states with the estimated increase ranging from 1 percent in Alaska
and Connecticut to 102 percent in Wyoming. 


--------------------
\20 We calculated that 49 states would have received higher federal
funding levels per recipient under TANF than they received per
recipient under the previous welfare programs in 1996. 

\21 If a state fails to meet its MOE in a given year, the following
year's TANF grant will be reduced by the shortfall (P.L.  104-193,
409, 110 Stat.  2144).  Furthermore, the state will be required to
bring its spending up in the following year to its required MOE and
replace the shortfall from the previous year.  (110 Stat.  2146). 

\22 Because states made the transition to TANF at different points
during federal fiscal year 1997, our calculations reflect the amount
states would have been eligible to receive for the full fiscal year
rather than the amount they actually received.  (See appendix II for
a more detailed discussion.)


   STATES' USE OF ADDITIONAL
   RESOURCES
------------------------------------------------------------ Letter :6

Additional budgetary resources for state welfare programs present
states with a unique opportunity to invest more in programs that can
help people find and keep their jobs and prevent them from returning
to welfare while still saving some resources for a "rainy day." All
10 states we visited planned to use some of their additional
resources to expand their programs.  Most states recognize that
achieving self-sufficiency and job placement calls for significant
investment in social services and incentives.  These states have
generally not increased cash benefit levels; rather, they plan to
spend additional resources for job placement services, child care,
and other supportive services that can help welfare recipients make
the transition to work.\23 For example, Texas increased the budget
for its job placement and training programs by about $100 million (or
about 200 percent) in order to expand access to job placement
services and enhance its "Invest in Long Term Success" initiative. 
This initiative (1) seeks to match employers with welfare recipients
and provides recipients with targeted training opportunities to meet
the needs of those employers, (2) enhances job retention services to
help former welfare recipients keep their jobs, and (3) creates
"local innovation grants" to support innovative welfare-to-work
programs, such as micro-enterprise development funds.  (See text box
1 for examples of how states are using federal and state funds to
enhance their welfare programs.)

   Text Box 1:  Examples of
   Expansion and Enhancements of
   State Welfare Programs

   (See figure in printed
   edition.)

In addition, as part of an effort to "make work pay," many states
have changed their policies relating to the treatment of earned
income from those previously in effect under AFDC to permit
recipients to keep more of their monthly cash assistance payments or
retain them for longer periods once they begin working.  More than
two-thirds of the states have increased the amount of assets and the
value of a vehicle that recipients can own and still remain eligible
for cash assistance.  The asset and vehicle limits in the prior AFDC
program were widely considered to be too low, creating barriers to
families' efforts to become more self-sufficient.  As these changes
allow more people to remain eligible for program benefits and to
remain eligible for transitional benefits, total state program
budgets have generally increased relative to caseload. 

In addition, the higher level of federal funds and lower caseloads
enabled states to reduce their own funding for the program down to
the required MOE level and still maintain higher total program
budgets.  TANF permits states to achieve some budgetary savings but
the MOE requirement constitutes a higher level of spending per
recipient in many states due to declining caseloads--limiting the
level of savings the state can achieve during a period of declining
caseloads.\24

(See text box 2 for additional examples of how states have achieved
budgetary savings in this manner.) In California, budget officials
said that they were frustrated by the MOE requirement because it
limited their budgetary flexibility.  Given the fixed nature of the
MOE levels, these officials noted that the state will no longer
realize any budgetary savings from a declining caseload because they
must spend the same amount of state funds on their welfare program as
they did in the previous year even if their caseloads are lower. 

   Text Box 2:  Examples of
   States' Use of Federal TANF
   Funds to Achieve Budgetary
   Savings

   (See figure in printed
   edition.)

   Note:  PRWORA allows states to
   transfer up to 30 percent of
   their TANF block grant into the
   CCDBG and the title XX Social
   Services Block Grant (SSBG). 
   Of this percentage, only 10
   percent can be transferred to
   SSBG and must be used only for
   programs and services for
   children and families with
   incomes below 200 percent of
   the poverty line.  The
   Transportation Efficiency Act
   for the 21st Century (TEA-21)
   reduces the amount a state can
   transfer from TANF to SSBG to
   4.5 percent for fiscal year
   2001 and after.  Pub.  L.  No. 
   105-178 8401(b), 112 Stat. 
   107, 499 (1998).

   (See figure in printed
   edition.)

Oregon provides an example of a state that used TANF funds to free up
a portion of state funds for other state priorities.  State officials
told us that during budget deliberations for the 1998-1999
biennium,\25 one of the Governor's proposals was a major overhaul of
the state's school financing system.  Although the state's economy
was sound and state revenues exceeded the forecast, the Governor
informed state agencies responsible for program budgets that some
state resources would have to be reallocated to the school financing
initiative.  According to agency officials responsible for Oregon's
welfare programs, many state programs were affected by this
reallocation.  Since their TANF grant was higher than what they had
received in the previous biennium, their MOE requirement lower than
what had been budgeted in the previous biennium, and their caseloads
had declined by over 50 percent since 1994, they were able to
reallocate nearly $55.2 million in state funds from their welfare
program and still meet their MOE requirement.  Those state general
funds shifted out of their welfare program were reallocated to other
programs within the Human Services department to cover budgetary
needs for planned program expansions in, for example, the Oregon
Health Plan and for other state general fund shortfalls resulting
from the Governor's overall budget priorities.  As a result, the
federal share of the state's TANF program expenses now totals 68
percent, compared to the previous federal share of about 56
percent.\26

In contrast, Maryland took a different approach by permitting the
state's Department of Human Resources to reinvest the state's
budgetary savings that result from caseload reductions.  Ten percent
of the total savings achieved in the state each year may be allocated
to demonstration projects to test innovative approaches to reduce
welfare dependency.  Any remaining savings may be distributed by the
state, with about half returning to the local social service
departments that achieved the caseload reductions as a performance
bonus.  These "reallocated savings" may be used for, among other
things, child care, welfare avoidance grants, drug treatment for
targeted recipients, transportation emergency funds, or any other
direct service to applicants or recipients that are considered
appropriate to accomplish the program's goals. 

There are several differing perspectives for assessing states' fiscal
commitment to these programs.  Although states have been able to
reduce their commitments of state funds below previous levels, they
nevertheless must still maintain spending at higher levels than they
would have spent under the matching grant programs.  In fact, given
the caseload decline, these lower levels of state spending are
providing more per recipient in many states.  Some states have
correspondingly argued that the MOE requirement prevents them from
achieving even greater savings and from reaping the budgetary rewards
traditionally associated with a declining caseload.  On the other
hand, some have raised concerns that reductions in overall state
spending could limit welfare reform's potential to provide the
resources necessary to move people from welfare to work. 


--------------------
\23 Since implementing TANF, Maryland has increased cash assistance
payments.  Under the state's welfare reform legislation, cash benefit
levels are in effect indexed to inflation. 

\24 We did not independently verify the reported levels of state
spending nor whether reported federal or state spending met the
qualifications set forth in the act.  States reported their fiscal
year 1997 federal TANF and state MOE expenditures in November 1997. 
Federal regulations specifying which state expenditures qualify as
MOE were still in draft.  Although no comprehensive verification of
state expenditures has been undertaken at this point, annual audits
performed under the Single Audit Act of 1984 should provide a useful
tool for ensuring that states are promoting financial accountability
for block grant programs such as TANF. 

\25 Oregon's current biennium began on July 1, 1997, and ends on June
30, 1999. 

\26 According to state officials, Oregon has historically invested
more state funds in its JOBS program than were required in order to
obtain the maximum federal matching grant.  For example, during the
1996-1997 biennium the state fiscal commitment to the JOBS program
represented 68 percent of the program budget, with federal matching
funds representing 32 percent.  With the passage of PRWORA, the state
was able to maximize all available resources by using federal TANF
funds where state funds had been previously used.  In addition, the
1998-1999 biennial program budget includes a 37 percent expansion to
the state's welfare program that was made possible by the
availability of additional federal TANF funds. 


   STATE CONTINGENCY BUDGETING FOR
   ECONOMIC DOWNTURNS
------------------------------------------------------------ Letter :7

In most of the states we visited, decisions on how to allocate the
additional budgetary resources available for low-income family
assistance programs were made in a context of strong state economies,
and most forecasts expected these trends to continue in the short
term.  Given the strength of their economies, most states we visited
did not see an immediate need to prepare for a recession for their
welfare programs.  Based on past experience, some state officials
said that if the economy worsened and states' revenues fell, the
budgetary impact would be felt in all state programs--including
welfare.  Nine of the 10 states we visited have established general
fund "rainy day" funds to be used for downturns in state economies
and budget shortfalls, but only four of the nine have significant
balances.\27

Some state officials believe that sound fiscal planning should
include some type of dedicated reserve for contingencies and other
future welfare program needs.  These officials said that a future
downturn could reduce funds available for benefits at a time when
they are most needed.  This could undermine welfare reform by
reducing supportive services crucial to the success of a welfare to
work strategy.  Four of the states we visited had enacted budgets
that established dedicated reserve funds although the amounts saved
were small relative to total program budgets.  Three of the four
states budgeted some federal TANF funds to a special program-specific
reserve account,\28 and one state, Maryland, set aside state general
funds for contingencies.  (See table 1.)



                                Table 1
                
                   Earmarked Welfare Program Reserve
                                Balances

        Earmarked reserves
        for state welfare
             programs
        ------------------
                   Percent
        Millions        of
              of   program
State    dollars    budget      Comments
------  --------  --------  --  --------------------------------------
Colora     8.9\a       3.7      Two reserves were established:
 do                              (1) Department of Human Services was
                                 given a $3 million reserve that can
                                 be used if needed without legislative
                                 approval and (2) The other reserve
                                 consists of a $5.9 million long-term
                                 contingency fund that requires
                                 legislative approval for
                                 disbursement. This second reserve is
                                 intended for use during an economic
                                 downturn.
Maryla      15.7       3.3      Within its budget stabilization fund,
 nd                              Maryland earmarked $15.7 million in
                                 state funds for its Family
                                 Independence Program.
Texas       25.3     2.3\c      The balance of all remaining TANF
                                 funds available to the state was
                                 appropriated to a Contingency
                                 Appropriations and made available for
                                 all expenditures relating to caseload
                                 growth and other program needs.
Wiscon      50\b     4.0\c      Reserves consist of (1) a Wisconsin
 sin                             Works agency contingency fund, (2) a
                                 reserve for additional benefit
                                 payments in Milwaukee, and (3) a
                                 reserve for benefits to children of
                                 Supplemental Security Income
                                 recipients.
----------------------------------------------------------------------
\a Colorado subsequently deposited an additional $16 million in the
reserve for long-term contingencies.  This reflected the balance of
its federal fiscal year 1997 TANF grant that remains unspent at the
U.S.  Treasury. 

\b Wisconsin has subsequently drawn down the entire balance of its
reserve for benefits to children of Supplemental Security Income
recipients and all but $2.8 million of its reserve for additional
benefits payments in Milwaukee, leaving a reserve of $27.8 million or
2.25% of its biennial program budget. 

\c As a percent of 1998-1999 biennial program budget. 

States that established reserves cited the possibility of future
economic downturns and other factors in their decision-making. 
Officials in Maryland expected that at some point in the future--as
has happened in the past--an economic downturn will bring higher
caseloads and higher program costs.  Using state general funds,
Maryland established a $15.7 million reserve fund dedicated to
low-income family assistance programs in part to address concerns
about assuring programmatic stability during such a period of fiscal
stress.  In another example, Colorado allocated $5.9 million of its
federal TANF grant to a long-term contingency reserve in the event it
experiences recession-driven caseload increases in the future. 

In contrast, officials in other states felt that sound fiscal
planning should focus on investing maximum resources now in a
welfare-to-work strategy.  For example, officials in Oregon said that
caseload levels have not fluctuated with the health of Oregon's
economy (as measured by the unemployment rate), and they do not
expect this to change.  They believe caseloads will continue to
decline if sufficient funds are invested now in appropriate services
to achieve recipients' long-term self-sufficiency.  These officials
attributed the state's sizeable caseload declines, from a reported
117,656 recipients in 1993 to 56,299 in 1997, to full investment in
the program and the new emphasis on work.  Similarly, Michigan
officials stated that past experience with caseload changes during
recessions may not be relevant in a post-reform environment.  These
officials expect that while caseload levels have not become
completely independent of the economy, they will eventually stabilize
and then fluctuate with the economy around some new, lower
core-caseload level which they believe will be so far below their
historic high that they do not expect to have difficulties financing
the costs of any future caseload fluctuations. 

The National Association of State Budget Officers (NASBO), the
National Governors' Association (NGA), and some state officials have
suggested that proposed federal rules may actually discourage states
from establishing dedicated state reserves composed of general
funds.\29 For example, in Michigan, state budget officials considered
establishing a reserve with state general funds until the state
learned that reserved funds would not count toward meeting the
states' TANF MOE requirement for the year in which they were
reserved.  Although Maryland did establish a state reserve, state
officials there raised similar concerns and there are currently no
plans to add more state funds to this fund.\30


--------------------
\27 Bond rating agencies recommend that states maintain rainy day
funds equal to a minimum of 3 percent of their general fund budgets. 
The four states we visited that projected a rainy day fund balance
above this threshold for state fiscal year 1998 are Colorado (3.9
percent), Connecticut (3.6 percent), Maryland (7.2 percent), and
Michigan (15.0 percent). 

\28 Although the states appropriated these reserves, the funds are
held in the U.S.  Treasury until needed. 

\29 According to HHS' draft regulations, dedicated state reserves for
welfare programs do not count toward a state's maintenance-of-effort
requirement until they are expended. 

\30 As previously noted, in order for a state to gain access to the
Contingency Fund a state must raise its own spending to a meet the
Contingency Fund's higher and more stringent MOE requirement.  These
funds may be hard to come by if the state is in the midst of a
recession.  A state official said, however, that if a state has saved
its own funds during robust economic times, it could more easily
raise its spending during downturns. 


   THE FEDERAL ROLE IN ENCOURAGING
   STATES' EFFORTS TO PLAN FOR
   FUTURE CONTINGENCIES
------------------------------------------------------------ Letter :8

The 1996 reforms of national welfare policy focused considerable
attention on the federal government's role in welfare.  While the
resulting legislation devolves much programmatic and financial
responsibility to the states, a significant federal role remains in
providing a substantial share of the funding for these programs,
setting national program objectives, establishing reporting and
accountability criteria, and ensuring a safety net.  Fiscal planning
responsibilities were devolved to the states.  The states were
granted the ability to save federal funds without fiscal year
limitation--in other words to plan for future contingencies.  The act
also provides two additional sources of federal funds--the
Contingency Fund and Loan Fund--to be available if economic
conditions affect caseloads and increase the fiscal burden on states. 
The Contingency Fund provides states with a limited amount of
matching funds, much like under AFDC, and requires states to increase
their own spending in order to receive federal matching funds.  The
Loan Fund allows states to borrow a limited amount as well, but they
must repay this loan within 3 years at a rate equal to the yield on a
similar Treasury security.  However, states have registered concerns
about the design of these federal contingency mechanisms. 


      LEVELS OF FUTURE UNSPENT
      TANF BALANCES HELD IN
      RESERVE ARE UNCERTAIN
---------------------------------------------------------- Letter :8.1

According to financial data reported by the states to HHS, many
states had not spent all of their fiscal year 1997 TANF block grants
by the end of the federal fiscal year, and some left considerable
balances at the Treasury.  Thirty-one states carried over a total of
more than $1.2 billion.  While these resources can certainly be used
in the event of an economic downturn, the presence of this apparent
fiscal cushion may reflect the transitional nature of the first year
of the grant rather than explicit state savings decisions. 

While we found that some states left a portion of their TANF grants
in reserve at the Treasury, we generally did not find a clear
relationship between the unspent TANF balances and states'
contingency plans.  During this transitional period, states generally
have been unable to forecast caseload levels with any degree of
accuracy.  In all states but one, caseloads continued to decline,
often at rates far faster than expected.  (See appendix II.) State
program budgets are prepared based on the projected caseload levels. 
Declines that are greater than expected have resulted in large
unspent balances.  Furthermore, the timing of the states' draws on
the TANF funds, and thus the levels of unspent resources, depend on
when the states submitted their state plans, enacted their laws, and
implemented their reforms. 

In addition, a Treasury policy statement issued in June 1997 affected
the timing of states' TANF draws.  It requires that for each
allocation of federal funds a state draws down, it must spend a
proportional share of its own MOE funds.  In this policy statement,
Treasury applies principles set forth in the Cash Management
Improvement Act of 1990 (CMIA) to the TANF block grant.  CMIA settled
a long-standing dispute between the federal government and the states
over the disbursement of funds for federal programs administered by
the states.  CMIA helps to ensure that neither party incurs
unnecessary interest costs in the course of federal grant
disbursements.\31 HHS recognized that this policy might, in some
cases, limit a state's financial flexibility.  It noted that because
PRWORA does not specifically exempt the TANF program, CMIA principles
apply whenever state MOE and federal TANF funds can be used
interchangeably.  HHS indicated that if a state were able to
demonstrate a bona fide need to draw its TANF funds under a different
schedule than it spends its state MOE funds, the Office of Management
and Budget (OMB) would consider granting an exemption to the
proportionate draw down requirement. 

Given these various transitional issues, current levels of unused
TANF may not be a reflection of state decisions to save for the
future and therefore may not be a reliable indicator of future
balances.  Also, a great deal of uncertainty exists surrounding
future welfare costs.  PRWORA requires states to place a growing
percentage of their caseload in work related activities over the next
5 years.\32 As we noted in our related report on states' efforts to
restructure their welfare programs, data from states that have
implemented early reforms and experienced large caseload reductions
indicate that many of the remaining recipients have multiple barriers
to participation in work activities, such as mental health and
substance abuse problems, and domestic violence.  Even if the economy
remains favorable, per recipient costs may need to grow as states
will have to place more of their caseloads in work-related activities
and a greater percentage of their caseloads will need services that
address the barriers to participation. 

The way federal policies are implemented may play a role in
influencing states' plans for future contingencies.  Organizations
representing states and officials in some states we visited suggested
that cash management rules may reduce states' incentives to save
federal TANF funds for the future.  Although states may carry forward
any unspent federal TANF funds without fiscal year limitations, these
unspent TANF reserves must be kept at Treasury--not drawn down and
kept in a state reserve.\33 NASBO, NGA, and the National Conference
on State Legislatures (NCSL) have observed that balances left at the
U.S.  Treasury may suggest to the Congress that grant levels are too
high and these funds remaining at Treasury are not needed by the
states.  Citing past experience with other federal grant programs,
such as the State Legalization Impact Assistance Grants (SLIAG)\34
program, where initial levels were reduced over time and federal
requirements increased, officials at NCSL expressed a concern that
unused TANF funds perceived as "excess" would become vulnerable to
reallocation by the Congress to other areas of national need.  These
same concerns were also expressed in some of the states we visited. 
Consequently, some state officials suggested that these concerns
might prompt states to spend rather than save a greater proportion of
their TANF funds. 

In contrast, the Urban Institute argues that the federal application
of CMIA to TANF is important in ensuring that states have some
reserves for a future contingency.  Since CMIA prohibits states from
spending federal TANF funds until they are needed and requires that
each draw of federal funds be matched by state funds, the Institute
believes that the application of CMIA to TANF has helped to ensure
that some federal funds were held in reserve.  This is especially
important, the Institute notes, because the draft HHS regulations
prohibit state funds held in reserve to be counted as MOE,
effectively creating a disincentive for states to create reserves
with their own funds. 

Under HHS' draft regulations, a state must report how much TANF and
MOE funds it spent on a variety of activities, such as cash
assistance, child care services, and work activities, but no
mechanism currently exists for states to inform the Congress about
their future plans for spending or saving TANF balances left at
Treasury.  Moreover, available data on TANF balances are generally
midyear data from the perspective of states' budgets and
appropriations decisions--not data used in state decision-making. 

In commenting on HHS' draft TANF regulations, the Center on Budget
and Policy Priorities suggested that more information about state
plans for saving TANF could aid congressional oversight of welfare
reform.  The Center suggested that as part of state financial
reporting, HHS could give states the option to record the amount of
TANF funds they plan to set aside for future contingencies, similar
to accounts established in three of the states we visited.  Allowing
for more transparency and information regarding states' contingency
budgets and the nature of the balances left in reserve in the U.S. 
Treasury could provide states with an opportunity to clarify their
longer term fiscal plans for the program.  Consequently, this would
help the Congress gain a better picture of the nature of the unspent
TANF balances. 


--------------------
\31 For additional information see Financial Management: 
Implementation of the Cash Management Improvement Act (GAO/AIMD-96-4,
January 8, 1996). 

\32 Each state must meet two separate work participation rates, that
is, the adult recipient must be engaged in work or job search
activities for a minimum of 20 hours per week.  The overall minimum
participation rate was 25 percent in federal fiscal year 1997, rising
to 50 percent in fiscal year 2002 and thereafter.  The minimum
participation rate for adults in two-parent families is 75 percent in
fiscal years 1997 and 1998 rising to 90 percent in fiscal year 1999
and thereafter.  A state that fails to meet its participation rates
is subject to a reduction in its TANF grant. 

\33 According to HHS, placing TANF funds in a state reserve is
prohibited by CMIA's requirement that states must minimize the time
between federal funds transfer and expenditure. 

\34 The Immigration Reform and Control Act of 1986 (IRCA) (P.L. 
99-603) provided appropriations to help states meet the costs, over a
7-year period, of services provided to aliens granted legal status by
IRCA and to cover federal program and administrative costs.  The act
provided for appropriations of
$1 billion for each of the four fiscal years, 1988-91.  Unused funds
from these years were authorized to be used by states through fiscal
year 1994.  In 1989 and again in 1990, the Congress reduced by about
$1.1 billion the annual appropriations for fiscal years 1990 and
1991, promising to make these funds available in fiscal year 1992. 
These actions were taken because the states were not expected to draw
down all appropriated funds through the end of fiscal year 1991. 
However, in 1992 the President's budget request proposed to rescind
the $1.1 billion.  We reported in May 1991 that the estimates of
future costs in the SLIAG program indicated that all of these
remaining funds might be needed.  Ultimately, these grants to the
states were reduced by $456 million, or 11 percent, of what had been
originally authorized.  For more information, see Health and Human
Services:  Funding for State Legalization Impact Assistance Grants
Program (GAO/HRD-91-109, May 23, 1991). 


      STATES ARE UNLIKELY TO MAKE
      EXTENSIVE USE OF PRWORA'S
      CONTINGENCY FUND OR LOAN
      FUND
---------------------------------------------------------- Letter :8.2

Officials in most of the states we visited indicated that they would
not use the Contingency Fund for State Welfare Programs (Contingency
Fund) nor the Federal Loan Fund for State Welfare Programs (Loan
Fund) even if they became eligible to do so.  These state officials
told us that neither the Contingency Fund nor the Loan Fund presented
states with a viable option for future contingencies. 

Complex federal reconciliation provisions and a more stringent
federal definition of qualified expenditures have led some state
officials to conclude that the costs of gaining access to the
Contingency Fund outweigh any benefits to the states.  To be eligible
to receive federal matching funds through the Contingency Fund, a
state must meet certain conditions.  First, a state must qualify as
"needy" under one of two triggers:  (1) in the most recent 3-month
period, its average unemployment rate (seasonally adjusted) must have
been at least 6.5 percent and must have increased at least 10 percent
from the corresponding rate in at least one of the 2 preceding years
or (2) its average monthly food stamp caseload for the most recent
3-month period must have increased at least 10 percent compared to
what enrollment would have been in the corresponding 3-month period
of fiscal year 1994 or 1995. 

Second, a state must meet a higher and more stringent level of MOE
spending.  None of the states included in our study had budgeted 100
percent for MOE.  Most states we visited planned to meet only the
minimum MOE required by PRWORA (between 75 and 80 percent of their
1994 expenditure levels), thereby requiring a substantial increase in
spending to qualify for the Contingency Fund.  In addition to the
requirement that states raise their spending levels to 100 percent of
historical expenditures to gain access to the Contingency Fund, a
more limited range of a state's spending can be counted toward the
Contingency Fund MOE than for the general MOE.  Although states may
count expenditures on separate state programs that can serve
TANF-ineligible clients as part of their general MOE requirement,
these same expenditures cannot be counted toward the 100 percent
Contingency Fund MOE.  HHS agrees that the operation of the
Contingency Fund would be simplified by allowing states to count the
same expenditures toward both the TANF MOE and Contingency Fund MOE. 
However, both HHS and the Congressional Budget Office (CBO) note that
changes that would ease access to the Contingency Fund would increase
the costs of the Contingency Fund in budgetary scoring terms and
could be subject to challenge under budget rules unless offsets were
found. 

Once a state meets these conditions, it is eligible to draw from the
Contingency Fund.  The state's annual draw is limited to 20 percent
of its annual TANF grant.  However, the state must match all draws
from the Contingency Fund with additional state money as determined
by its matching rate under the Medicaid program, or federal medical
assistance percentage (FMAP).\35 Moreover, there is a year-end
reconciliation process which can reduce state allotments depending on
the number of months during the year the state was eligible.  (See
text box 3 for more detailed discussion of this point.) Lags in data
availability mean states qualifying on the basis of food stamp
caseload increases would not even be aware of their eligibility until
some time after the need arose. 

   Text Box 3:  The Contingency
   Fund's Annual Reconciliation
   Process

   (See figure in printed
   edition.)

Furthermore, the Adoption and Safe Families Act of 1997 (Public Law
105-89), 404 reduces the cap on Contingency Fund spending by $40
million over 4 years.  If a state drew funds in a year affected by
the reduction, the amount it could retain would be reduced by its
share of the annual reduction.  For example, the total reduction in
fiscal year 1999 is $9 million.  If two states drew funds in fiscal
year 1999, at the end of the year, these two states' allocation would
be reduced by $4.5 million each.  If the states had already received
their allocation they would have to remit $4.5 million each. 

Although eight states\36 qualified as needy and could have gained
access to the Contingency Fund in fiscal year 1997, according to HHS,
only New Mexico and North Carolina requested and were awarded
funds.\37 Although Hawaii would have been eligible for resources from
the Fund for all of federal fiscal year 1997, the state determined
that it did not have enough qualifying state expenditures to meet the
Fund's 100 percent MOE requirement.  California was also eligible for
Contingency Funds for the first 4 months of federal fiscal year 1997
(October 1, 1996, through January 31, 1997).  Upon completing the
reconciliation process, the state calculated that it would have to
increase its own spending by almost
$1.9 billion in order to receive $249 million from the Contingency
Fund and declined to do so.  According to HHS officials, North
Carolina and New Mexico had been awarded funds on May 29, 1998.  As
of July 24, 1998, neither state had completed the reconciliation
process and HHS officials expect that both states will be required to
remit a large share of these funds. 

State officials also indicated that they are unlikely to borrow from
the Loan Fund established in PRWORA.\38 Officials in some states
indicated that borrowing specifically for social welfare spending in
times of fiscal stress would not receive popular support.  We have
previously reported on states' reluctance to participate in a similar
loan program in the Unemployment Insurance (UI) Trust Fund.\39 The UI
program originally operated as a forward-funded system with benefit
levels and tax rates set so that the program could "save for a rainy
day" by building reserves during periods of economic expansion and to
be able to pay UI benefits during economic downturns.  This
federal-state partnership is financed through payroll taxes that are
used to pay benefits, finance administrative costs, and maintain a
loan account from which financially troubled states can borrow funds
to pay UI benefits.  By the early 1980s, as a result of severe
back-to-back recessions, many states had depleted their reserves and
began to rely on federal loans to sustain UI benefits.  The Congress
enacted several laws designed to move the system toward healthier
reserve balances.  These changes made it more expensive for states to
borrow from the federal government.  State loan repayments increased,
and states took other actions including cutting program benefits,
limiting the length of time recipients could receive benefits and, in
some cases, increasing payroll taxes--jeopardizing the program's
objective of helping to stabilize the economy during recessions. 
Although the UI program is designed to allow states to build reserves
during good economic times in order to pay benefits during
downturns--and allows states to borrow from the federal
government--these provisions have not always provided sufficient
protection against a need for additional federal resources.  For
example, in 1993 the Congress passed a $4 billion supplemental
appropriations bill to finance emergency unemployment benefits with
federal funds due to shortfalls in state unemployment compensation
trust funds and the unwillingness of states to borrow federal funds
or expend state general funds on UI.\40


--------------------
\35 Under AFDC, state spending was matched at a rate based on each
state's per capita income.  This rate, FMAP, is also used in other
federal-state matching grant programs such as Medicaid.  It ranges
from 50 percent for wealthy states to 79 percent for poorer states. 

\36 Alaska, California, Hawaii, New Mexico, New York, North Carolina,
Washington, and the District of Columbia. 

\37 North Carolina was awarded $15,111,980 and New Mexico was awarded
$ 21,017,193. 

\38 States that have not incurred penalties for improper use of TANF
funds are eligible for loans from the Loan Fund.  Such loans are to
have a maturity of no more than 3 years at an interest rate
comparable to the average market yield on outstanding federal
obligations with comparable periods to maturity.  The cumulative
total of loans made to a state from fiscal year 1997 through fiscal
year 2002 is not to exceed 10 percent of a state's TANF grant. 

\39 Unemployment Insurance:  Program's Ability to Meet Objectives
Jeopardized (GAO/HRD-93-107, September 28, 1993). 

\40 P.L.  103-24, 107 Stat.  67. 


   CONCLUSIONS
------------------------------------------------------------ Letter :9

With the passage of PRWORA, the nature of the partnership between
states and the federal government for designing and financing welfare
programs changed.  Much of the downside fiscal risk has been shifted
to the states by virtue of the fixed nature of the TANF block grant. 
States have gained important new flexibility in making decisions, and
their provisions for financing their programs in the near and long
term will have an important bearing on the future success of welfare
reform. 

States currently have more resources available for these programs
under TANF than would have been available under the old financing
system, but future fiscal demands are uncertain.  The fixed nature of
the TANF block grant, the potential volatility of welfare caseloads
and program spending along with the "pro-cyclical" budgetary
pressures states face under their balanced budget requirements
highlight the importance of both the states' own funding for
contingencies as well as PRWORA's provisions for
contingencies--allowing states to save unused TANF funds for future
years' use and the two safety net mechanisms, the Contingency Fund
and the Loan Fund. 

As these provisions attest, the federal government retains a stake in
states' fiscal decisions affecting the sustainability of the program
during downturns.  As we noted, states see limited incentives to use
the Contingency and Loan Funds, in part because the costs associated
with gaining access outweigh many of the benefits that these
mechanisms may offer.  Although improving access might help states
cope with the effects of economic slowdowns, easing these funds'
requirements could prove costly and may lessen the incentives for
states to fulfill their own responsibilities for fiscal planning and
program financing.  While the Congress may very well revisit the
design of these funds as the implementation of TANF unfolds, for now
the TANF balances left at Treasury constitute the principal source of
federal contingency funds for the states. 

While many states had TANF balances at Treasury at the end of fiscal
year 1997, current reporting requirements do not clearly identify
states' plans for these balances.  We identified a number of
transitional issues that may have affected the levels of balances.  A
number of factors unrelated to states' savings decisions have
influenced the levels of these funds, including cash management
practices, slow-starting programmatic spending, and caseload
declines.  States we visited took different approaches to contingency
budgeting for TANF, and states' practices may change as they gain
experience in implementing their reformed programs under the grant. 
Better information on states' plans for future contingencies,
including on states' unused TANF balances, could play a role in the
continuing dialogue between states and the Congress as welfare reform
continues to unfold. 

In the new block grant environment, the federal government has an
interest in encouraging state savings, but what constitutes
"adequate" saving will remain a state judgment made under conditions
of considerable uncertainty.  Finding the right balance between
saving budgetary resources for future contingencies and investing
them in programs that help people make the transition from welfare to
work will be one of the main challenges for states as they develop
strategies to address the needs of low-income families. 


   RECOMMENDATION
----------------------------------------------------------- Letter :10

We recommend that the Secretary of Health and Human Services consult
with the states and explore various options to enhance information
regarding states' plans for their unused TANF balances.  Such
information might include explicit state plans for setting aside
TANF-funded reserves for the future.  Allowing for more transparency
regarding states' fiscal plans for TANF funds could enhance
congressional oversight over the multi-year time frame of the grant
and provide states with an opportunity to more explicitly consider
their long-term fiscal plans for the program. 


   AGENCY COMMENTS
----------------------------------------------------------- Letter :11

We received comments from HHS, which are reprinted in full in
appendix III.  In addition, portions of the report were reviewed for
technical accuracy by officials in the states we visited, and their
comments were incorporated as appropriate.  We also asked the
National Governor's Association (NGA), the National Conference of
State Legislatures (NCSL), and the American Public Welfare
Association (APWA) to review the report.  We incorporated comments
from these organizations as appropriate. 

HHS, NGA, NCSL, and APWA generally agreed that this report is an
accurate and comprehensive portrayal of the current fiscal issues
facing states as they make progress toward implementing welfare
reform.  HHS, however, expressed concern that our analysis of states'
additional budgetary resources did not take into account that states
must now engage a significantly higher percentage of their caseload
in work activities and that the costs of operating a welfare program
before reforms could not be compared to the costs of operating a
welfare program under TANF.  Indeed, HHS, NGA, and NCSL all
emphasized that under TANF, states are expected to do much more than
under AFDC.  Our analysis was not meant to compare the real costs of
operating the AFDC program to the real costs of operating a welfare
program under TANF, nor was it meant to minimize the additional
responsibilities incumbent on states as they make progress
implementing welfare reforms.  Instead, we sought to illustrate the
levels of resources that are available to finance the dramatic
changes in states' welfare programs.  In our recent report on state
program restructuring, we described how states are moving away from a
welfare system that focused on entitlement to assistance to one that
emphasizes finding employment as quickly as possible and becoming
more self-sufficient.  For example, we found that states were using
some of their additional budgetary resources to enhance support
services, such as transportation and child care, for recipients
participating in work activities and poor families who have found
jobs and left the welfare rolls.  We concluded that the confluence of
a strong national economy that fosters employment opportunities and
the availability of additional budgetary resources has created an
optimal time for states to reform their welfare programs. 

HHS concurred with our recommendation, but NGA and NCSL expressed
concerns that it would lead to an increase in the reporting
requirements already imposed on the states.  It is because we agree
that costs associated with collecting information should not outweigh
its usefulness that we suggested HHS and the states work together on
developing a reporting option.  We recognize that estimates of future
caseloads would affect estimates of future unspent TANF balances and
that developing accurate caseload estimates at this early stage in
TANF implementation poses problems for states.  However, as NCSL and
NGA agreed, information on states' plans for unspent TANF balances
could prove useful as the Congress executes its oversight
responsibilities of the TANF program and the program's funding
levels.  We continue to believe that the Congress would benefit from
more complete information on states' plans for future contingencies,
including unspent TANF balances.  As states, HHS, and other cognizant
parties meet to discuss final reporting requirements under TANF, we
urge them to work together to explore reporting options.  This
discussion can form part of the ongoing dialogue on how best to
restructure governmental roles and responsibilities to achieve the
goals of welfare reform. 

NGA, NCSL, and APWA also underscored states' concerns about applying
CMIA to TANF.  In their view, CMIA limits state flexibility by
restricting TANF funds from being held in state reserve accounts and
by requiring that all draws of federal TANF funds be matched with
state MOE funds.  According to APWA, because caseloads have declined
in many states and because states must still meet their MOE
requirements, many states are not even drawing federal funds until
the fourth quarter of the federal fiscal year.  These comments
reinforce the point made in our report that cash management practices
may have had an impact on the level of TANF resources remaining at
Treasury at the end of the federal fiscal year.\41 On a related
issue, APWA cited the Congress' recent decision to reduce the
proportion of TANF a state can transfer to its SSBG program as
effectively limiting states' flexibility in TANF draw downs, which in
APWA's view, penalized states for leaving TANF funds in the Treasury. 

NGA and NCSL urged that the federal Contingency Fund be redesigned to
be a more attractive option for states.  Specifically, NGA
recommended changing the Contingency Fund's MOE provision to conform
to the TANF MOE requirement and to change the reconciliation
requirement to eliminate the reduction in the state's match rate that
is based on the number of months the state was eligible to access the
Contingency Fund.  We noted in our report that very few state budget
officials perceive that the Contingency Fund will serve to help
states maintain stable program financing if caseloads rise during a
recession.  However, HHS noted that redesigning the Contingency Fund
could result in significant increases in federal costs.  The
Contingency Fund was designed to balance competing objectives.  On
the one hand it could not be so generous as to encourage routine or
casual use which would have led to significantly higher federal
costs.  On the other hand, if the Contingency Fund is overly
restrictive many states would be disinclined to use it and it would
not serve as the fiscal stabilizer it was intended to be.  Balancing
these competing objectives will likely challenge the Congress as well
as the states as they continue to make progress in implementing their
welfare programs under a variety of economic and demographic
conditions. 


--------------------
\41 As noted earlier in our report, HHS has stated that because
PRWORA does not specifically exempt the TANF program, CMIA principles
apply whenever state MOE and federal TANF funds can be used
interchangeably. 


--------------------------------------------------------- Letter :11.1

As agreed with your office, unless you release this report earlier,
we will not distribute it until 30 days from the date of this letter. 
At that time, we will send copies to the Chairmen and Ranking
Minority Members of the Senate Committee on Finance and the Senate
Subcommittee on Social Security and Family Policy, Committee on
Finance, and other interested parties.  We will also make copies
available to others upon request. 

If you have any questions, please call me at (202) 512-9573. 

Sincerely yours,

Paul L.  Posner
Director, Budget Issues


SCOPE AND METHODOLOGY
=========================================================== Appendix I

This review was conducted in conjunction with a review conducted by
GAO's Health, Education and Human Services (HEHS) Division.  The HEHS
review studied welfare reform implementation in seven states: 
California, Connecticut, Louisiana, Maryland, Oregon, Texas, and
Wisconsin.  These states were selected because they represent a
diverse range of socioeconomic characteristics, geographic locations,
and experiences with state welfare initiatives.  According to the
U.S.  Bureau of the Census and HHS estimates, the states ranged in
population from about 3.2 million (Oregon) to about 31.0 million
(California) in 1996; in median income for three-person families,
from about $33,337 (Louisiana) to about $52,170 (Connecticut) in
federal fiscal year 1997; and in overall poverty rates, from 8.5
percent (Wisconsin) to about 19.7 percent (Louisiana) in 1995.  Some
states like Wisconsin and Oregon have had reform initiatives in place
for several years that include elements similar to those in PRWORA,
such as time limits for welfare benefits (Wisconsin) and increased
work participation requirements (Oregon and Wisconsin); others, such
as Louisiana, have been operating more traditional cash assistance
programs with welfare-to-work components and were only beginning more
extensive reforms in fiscal year 1997. 

In addition, in order to capture a broader picture of the fiscal and
budgetary implications of welfare reform, we added three additional
states to our review:  Michigan, New York, and Colorado. 
Historically, Michigan's economy and budget have been highly
sensitive to economic change, and in 1977 Michigan created a Budget
Stabilization Fund to help stabilize the state's fiscal policy.  Like
California, New York and Colorado have county administered welfare
programs that share in the costs of the welfare programs.  We added
these states to obtain the views of local officials on the fiscal
implications of welfare reform in their states.  By including New
York, Michigan, and Colorado, we also increased the geographic
diversity of our study states and included states that when combined
with the other seven states administer the welfare programs of about
half the nation's total caseload. 

To meet our objectives, we interviewed state and local officials in
the local low-income family assistance programs and in program and
state-wide budget offices.  Specifically, we met with officials from
the following organizations during our state visits:  executive
branch budget offices; legislative budget/finance committees; social
service agencies; selected county program and budget offices; and
advocacy groups.  We also reviewed state program and budget
documents, the PRWORA legislation, HHS regulations and policy
guidance, prior GAO reports, and welfare experts' studies. 

We also analyzed fiscal data related to all 50 states' low-income
family assistance programs obtained from HHS to determine the level
of additional budgetary resources states' received as a result of
welfare reform.  (See appendix II for a more detailed explanation of
the methodology used in this calculation.) We did not verify the
accuracy of these data. 

We requested written comments on a draft of this report from HHS,
NGA, NCSL, and the American Public Welfare Association (APWA).  These
comments are discussed in the letter, and HHS' comments are reprinted
in appendix III. 


ESTIMATING ADDITIONAL BUDGETARY
RESOURCES AVAILABLE FOR STATES'
WELFARE PROGRAMS
========================================================== Appendix II

States currently have more budgetary resources available for their
welfare programs than they would have had under prior law.  This is
primarily the result of a combination of three interrelated factors: 
(1) the unprecedented declines in caseloads, (2) the new federal
financing mechanism, or block grant, that provides resources to the
states without regard to the numbers of people states' welfare
programs serve, and (3) the maintenance of effort requirement on
states that establishes a minimum, or floor, funding level for their
state welfare programs.  This appendix describes the influence each
of these factors has on total resources available, and then presents
our estimates of the combined effect they have on total available
resources. 

IMPACT OF DECLINING CASELOADS ON
AVAILABLE RESOURCES

Given the fixed nature of the federal funding stream and states'
minimum MOE contributions, caseload volatility will dramatically
affect the resources available per recipient for state welfare
programs.  As caseloads drop, there will be more resources available
to the states to finance their welfare programs since programs'
finance needs are largely driven by caseload assumptions.  In
contrast, if caseloads rise, there will be fewer federal dollars per
recipient when compared to the previous budget period, and states
will need to raise additional resources on their own or adjust their
programs to make their resources go further.  In many states,
caseloads began to decline even before the enactment of PRWORA and
continued to do so after passage of the law as shown in Table II.1. 
While there remains controversy over some of the reasons for the
caseload declines, research indicates that important factors include
the strong economy and changes in federal and state welfare policies. 



                                    Table II.1
                     
                       Changes in Welfare Caseloads 1994-97

              Total AFDC/TANF recipients by state
            ----------------------------------------
                                                           Percent       Percent
                                                      change 1994-  change 1996-
State       January 1994   August 1996     July 1997            97            97
----------  ------------  ------------  ------------  ------------  ------------
Alabama          135,096       100,510        74,097           -45           -26
Alaska            37,505        35,540        33,663           -10            -5
Arizona          202,350       169,440       137,899           -32           -19
Arkansas          70,563        56,230        51,506           -27            -8
California     2,621,383     2,578,450     2,282,389           -13           -11
Colorado         118,081        95,790        59,171           -50           -38
Connecticu       164,265       159,060       151,321            -8            -5
 t
Delaware          29,286        23,650        21,841           -25            -8
Florida          689,135       533,800       407,598           -41           -24
Georgia          396,736       329,160       243,541           -39           -26
Hawaii            60,975        66,480        74,297            22            12
Idaho             23,342        21,800         7,890           -66           -64
Illinois         709,969       640,870       547,958           -23           -14
Indiana          218,061       141,850       107,355           -51           -24
Iowa             110,639        85,940        73,837           -33           -14
Kansas            87,433        63,780        47,434           -46           -26
Kentucky         208,710       170,890       151,190           -28           -12
Louisiana        252,860       228,120       178,335           -29           -22
Maine             65,006        53,790        44,972           -31           -16
Maryland         219,863       194,130       154,166           -30           -21
Massachuse       311,732       219,580       196,630           -37           -10
 tts
Michigan         672,760       501,440       424,612           -37           -15
Minnesota        189,615       169,740       151,201           -20           -11
Mississipp       161,724       122,750        87,118           -46           -29
 i
Missouri         262,073       222,820       182,022           -31           -18
Montana           35,415        28,240        25,258           -29           -11
Nebraska          46,034        38,510        37,455           -19            -3
Nevada            37,908        33,920        27,896           -26           -18
New               30,386        22,940        17,493           -42           -24
 Hampshire
New Jersey       334,780       275,700       240,338           -28           -13
New Mexico       101,676        99,660        69,605           -32           -30
New York       1,241,639     1,143,960     1,002,936           -19           -12
North            334,451       266,470       231,506           -31           -13
 Carolina
North             16,785        13,130        10,508           -37           -20
 Dakota
Ohio             691,099       549,310       449,123           -35           -18
Oklahoma         133,152        96,010        74,567           -44           -22
Oregon           116,390        78,420        56,299           -52           -28
Pennsylvan       615,581       530,520       432,907           -30           -18
 ia
Rhode             62,737        56,460        54,498           -13            -3
 Island
South            143,883       113,430        79,820           -45           -30
 Carolina
South             19,413        15,840        12,497           -36           -21
 Dakota
Tennessee        302,608       238,890       163,236           -46           -32
Texas            796,348       647,790       479,933           -40           -26
Utah              50,657        39,060        31,975           -37           -18
Vermont           28,095        24,270        22,403           -20            -8
Virginia         194,959       152,680       119,430           -39           -22
Washington       292,608       268,930       238,920           -18           -11
West             115,376        89,039        80,359           -30           -10
 Virginia
Wisconsin        230,621       148,890       100,387           -56           -33
Wyoming           16,740        11,400         4,957           -70           -57
================================================================================
Total         14,008,503    11,969,079     9,956,349           -29           -17
Average          280,170       239,382       199,127           -33           -19
Median           152,804       118,090        83,739           -32           -18
--------------------------------------------------------------------------------
Source:  HHS, Administration for Children and Families.  These data
are reported by the states to HHS and have not been independently
verified by GAO. 

Note:  The "total percent change" presented in this table represents
the percent difference in the nationwide totals.  The "average
percent change" is a simple average of the percentage differences
across states, with each state having equal weight. 

Overall, states' caseloads have declined by about a third since 1994. 
However, this national average masks the differences among the states
in the magnitude and timing of their caseload declines.  For example,
in North Carolina, the caseload dropped by about a third since 1994,
with a decline of 20 percent before federal reforms had been enacted
and an additional 13 percent decline in the last year.  In contrast,
in New Mexico, the overall decline is also near the national average;
however, virtually all of the change occurred in the last year--after
PRWORA passed.  There is also a large disparity among the states in
total caseload change since 1994, ranging from an decrease of 70
percent in Wyoming to an increase of 22 percent in Hawaii. 

IMPACT OF BLOCK GRANT ON AVAILABLE
RESOURCES

Tables II.2 and II.3 present estimates of the change in federal
resources available to implement welfare reform.  Table II.2 presents
our estimates of the differences between available nominal federal
resources for family assistance programs under AFDC and under TANF. 
Our analysis shows that 45 states will receive more federal resources
under TANF than they received in the last year before reform.\1 TANF
provides about $1.4 billion more federal dollars to the states than
they received under the consolidated programs in 1996, when caseloads
were on average much higher.  These differences ranged from 70
percent more for Indiana to 7 percent less for Pennsylvania; the
median increase was about 9 percent. 



                               Table II.2
                
                Additional Federal Resources Under TANF

                               Total federal
                                TANF-related
                            program spending    TANF block     Percent
State                                in 1996         grant  difference
--------------------------  ----------------  ------------  ----------
Alabama                          $78,773,572   $93,315,207          18
Alaska                            60,674,212    63,609,072           5
Arizona                          199,721,493   222,419,988          11
Arkansas                          53,694,984    56,732,858           6
California                     3,527,307,569  3,733,817,78           6
                                                         4
Colorado                         138,940,350   136,056,690          -2
Connecticut                      244,355,429   266,788,107           9
Delaware                          30,174,601    32,290,981           7
Florida                          527,068,864   562,340,120           7
Georgia                          300,803,817   330,741,739          10
Hawaii                            98,386,892    98,904,788           1
Idaho                             31,299,902    31,938,052           2
Illinois                         593,427,626   585,056,960          -1
Indiana                          121,379,747   206,799,109          70
Iowa                             123,547,672   131,524,959           6
Kansas                            86,790,703   101,931,061          17
Kentucky                         170,664,524   181,287,669           6
Louisiana                        122,357,457   163,971,985          34
Maine                             73,231,234    78,120,889           7
Maryland                         209,499,543   229,098,032           9
Massachusetts                    372,009,461   459,371,116          23
Michigan                         581,487,845   775,352,858          33
Minnesota                        238,786,264   267,984,886          12
Mississippi                       69,324,485    86,767,578          25
Missouri                         207,861,860   217,051,740           4
Montana                           39,332,524    45,534,006          16
Nebraska                          56,093,250    58,028,579           3
Nevada                            41,241,248    43,976,750           7
New Hampshire                     36,045,607    38,521,261           7
New Jersey                       362,624,796   404,034,823          11
New Mexico                       129,909,486   126,103,156          -3
New York                       2,331,710,268  2,442,930,60           5
                                                         2
North Carolina                   313,314,815   302,239,599          -4
North Dakota                      24,270,062    26,399,809           9
Ohio                             558,275,390   727,968,260          30
Oklahoma                         122,960,328   148,013,558          20
Oregon                           146,431,168   167,924,513          15
Pennsylvania                     769,802,235   719,499,305          -7
Rhode Island                      82,015,807    95,021,587          16
South Carolina                    98,792,638    99,967,824           1
South Dakota                      19,791,424    21,893,519          11
Tennessee                        182,575,643   191,523,797           5
Texas                            435,279,035   486,256,752          12
Utah                              67,756,472    76,829,219          13
Vermont                           42,215,710    47,353,181          12
Virginia                         134,649,524   158,285,172          18
Washington                       390,759,361   404,331,754           3
West Virginia                     91,328,347   110,176,310          21
Wisconsin                        239,847,605   318,188,410          33
Wyoming                           13,628,240    21,781,446          60
======================================================================
Total                        $14,992,221,089  $16,396,057,           9
                                                       420
Average                         $299,844,422  $327,921,148          13
Median                          $126,728,579  $153,149,365           9
----------------------------------------------------------------------
Source:  GAO analysis based on data reported by states to HHS,
Administration for Children and Families.  These data are reported by
the states to HHS and have not been independently verified by GAO. 

Note:  The "total percent difference" presented in this table
represents the percent difference in the nationwide totals.  The
"average percent difference" is a simple average of the percentage
differences across states, with each state having equal weight. 

Table II.3 presents these additional federal resources on a per
recipient basis to take into account the significant declines in
caseload that have occurred since passage of PRWORA.  These estimates
of states' additional federal resources considered on a per recipient
basis present a different picture not only because the estimates take
post-PRWORA declines in caseloads\2 into account but also because of
differences among the states in their expenditures for emergency
assistance and administration, which TANF also replaced.\3 Adjusting
for smaller caseloads, on average, the new financing mechanisms in
TANF provide states with about $614 more federal dollars per
recipient than the consolidated programs provided in 1996.  This
table shows an increase in federal resources for all states but one,
with a median increase of 47 percent more than before reform.  The
change in federal resources per recipient ranged from an increase of
334 percent (Wyoming) to a decrease of 10 percent (Hawaii).  Our
analysis shows that 38 states received an increase of 25 percent or
more. 



                               Table II.3
                
                    Additional Federal Resources Per
                          Recipient Under TANF

                                     Federal resources per recipient
                                    ----------------------------------
                                                               Percent
State                                     1996        1997  difference
----------------------------------  ----------  ----------  ----------
Alabama                                   $733      $1,259          72
Alaska                                   1,717       1,890          10
Arizona                                  1,169       1,613          38
Arkansas                                   912       1,101          21
California                               1,336       1,636          22
Colorado                                 1,408       2,299          63
Connecticut                              1,522       1,763          16
Delaware                                 1,315       1,478          12
Florida                                    921       1,380          50
Georgia                                    822       1,358          65
Hawaii                                   1,479       1,331         -10
Idaho                                    1,339       4,048         202
Illinois                                   900       1,068          19
Indiana                                    829       1,926         132
Iowa                                     1,356       1,781          31
Kansas                                   1,235       2,149          74
Kentucky                                   971       1,199          24
Louisiana                                  514         919          79
Maine                                    1,304       1,737          33
Maryland                                 1,015       1,486          46
Massachusetts                            1,542       2,336          52
Michigan                                 1,091       1,826          67
Minnesota                                1,396       1,772          27
Mississippi                                524         996          90
Missouri                                   878       1,192          36
Montana                                  1,213       1,803          49
Nebraska                                 1,464       1,549           6
Nevada                                   1,027       1,576          53
New Hampshire                            1,480       2,202          49
New Jersey                               1,243       1,681          35
New Mexico                               1,269       1,812          43
New York                                 1,957       2,436          24
North Carolina                           1,120       1,306          17
North Dakota                             1,793       2,512          40
Ohio                                     1,015       1,621          60
Oklahoma                                 1,120       1,985          77
Oregon                                   1,600       2,983          86
Pennsylvania                             1,404       1,662          18
Rhode Island                             1,357       1,744          28
South Carolina                             816       1,252          53
South Dakota                             1,183       1,752          48
Tennessee                                  693       1,173          69
Texas                                      613       1,013          65
Utah                                     1,657       2,403          45
Vermont                                  1,639       2,114          29
Virginia                                   815       1,325          63
Washington                               1,423       1,692          19
West Virginia                              931       1,371          47
Wisconsin                                1,310       3,170         142
Wyoming                                  1,013       4,394         334
======================================================================
Total                                   $1,193      $1,647          38
Average                                 $1,188      $1,802          55
Median                                  $1,224      $1,687          47
----------------------------------------------------------------------
Source:  GAO analysis based on data reported by states to HHS,
Administration for Children and Families.  These data are reported by
the states to HHS and have not been independently verified by GAO. 

Note:  The "total percent difference" presented in this table
represents the percent difference in the nationwide totals.  The
"average percent difference" is a simple average of the percentage
differences across states, with each state having equal weight. 

IMPACT OF MOE ON STATE SPENDING

Given a declining national caseload, the state MOE requirement
further augments the budgetary resources available on a per recipient
basis to finance states' low-income family assistance programs.\4 The
state MOE is based on spending for a larger set of programs than the
TANF block grant and was pegged to state spending in those programs
during a period of high caseloads and high spending.\5 In the absence
of a MOE requirement, states could draw down all of their federal
TANF grants, and then reduce their own financial commitment to the
program to whatever level would maintain a current service budget
baseline.\6 In all states but one--Indiana--the 80 percent TANF-MOE
requirement is less than what the state spent on those programs in
1996 (see table II.4) and would allow them to reduce their own
financial commitment to the program. 



                               Table II.4
                
                State Resource Levels--Pre-Reform Versus
                              Post-Reform

                Differences in State Resources Under AFDC
                 in 1996 and Under the TANF Block Grants
                ------------------------------------------
                   MOE related
                state spending                  80 percent     Percent
State                  in 1996                         MOE  difference
--------------  --------------  --------------  ----------  ----------
Alabama            $53,483,367                  $41,828,39         -22
                                                         3
Alaska              63,412,721                  52,205,229         -18
Arizona            128,934,235                  101,362,85         -21
                                                         4
Arkansas            28,946,483                  22,228,215         -23
California       3,624,777,252                  2,914,566,         -20
                                                       324
Colorado           138,102,387                  88,395,622         -36
Connecticut        265,717,429                  195,649,12         -26
                                                         7
Delaware            34,293,367                  23,222,474         -32
Florida            481,292,387                  395,646,98         -18
                                                         7
Georgia            237,020,582                  184,926,42         -22
                                                         9
Hawaii             101,127,338                  77,846,912         -23
Idaho               20,255,271                  14,590,646         -28
Illinois           657,265,076                  457,621,89         -30
                                                         0
Indiana             99,927,497                  121,093,31          21
                                                         0
Iowa                87,418,746                  66,094,156         -24
Kansas              74,964,352                  65,866,201         -12
Kentucky            96,465,302                  71,913,050         -25
Louisiana           64,490,198                  59,109,470          -8
Maine               49,856,343                  40,296,038         -19
Maryland           230,500,750                  188,763,14         -18
                                                         0
Massachusetts      414,599,899                  382,877,35          -8
                                                         8
Michigan           512,964,169                  499,752,93          -3
                                                         4
Minnesota          230,083,835                  191,728,27         -17
                                                         8
Mississippi         28,619,923                  23,172,595         -19
Missouri           170,397,434                  128,128,82         -25
                                                         6
Montana             21,968,011                  16,735,379         -24
Nebraska            51,366,287                  30,902,916         -40
Nevada              43,879,854                  27,188,122         -38
New Hampshire       40,747,726                  34,256,105         -16
New Jersey         416,194,085                  324,219,20         -22
                                                         6
New Mexico          58,529,496                  39,947,126         -32
New York         2,377,231,367                  1,824,848,         -23
                                                       309
North Carolina     255,019,427                  164,454,14         -36
                                                         7
North Dakota        16,429,294                   9,673,984         -41
Ohio               440,169,491                  416,587,57          -5
                                                         4
Oklahoma            83,846,543                  65,333,660         -22
Oregon             125,393,961                  98,405,163         -22
Pennsylvania       758,034,488                  434,267,30         -43
                                                         6
Rhode Island        76,887,061                  64,391,515         -16
South Carolina      61,055,562                  38,228,678         -37
South Dakota        13,496,812                   9,359,245         -31
Tennessee          140,545,724                  88,330,537         -37
Texas              348,217,343                  251,439,64         -28
                                                         6
Utah                41,359,316                  26,976,586         -35
Vermont             31,431,632                  27,363,633         -13
Virginia           143,919,823                  136,718,04          -5
                                                         8
Washington         422,477,682                  290,198,32         -31
                                                         0
West Virginia       43,458,106                  34,881,108         -20
Wisconsin          196,055,505                  180,510,64          -8
                                                         7
Wyoming             11,749,234                  11,376,348          -3
======================================================================
Total           $14,114,380,17                  $11,055,47         -22
                             3                       9,766
Average           $282,287,603                  $221,109,5         -22
                                                        95
Median             $98,196,400                  $74,879,98         -22
                                                         1
----------------------------------------------------------------------
Source:  GAO analysis based on data reported by states to HHS,
Administration for Families and Children.  These data are reported by
the states to HHS and have not been independently verified by GAO. 

Note:  The "total percent difference" presented in this table
represents the percent difference in the nationwide totals.  The
"average percent difference" is a simple average of the percentage
differences across states, with each state having equal weight. 

However, the minimum MOE requirements, taken together with the
further decrease in caseloads had the effect of increasing the level
of state resources spent on a per recipient basis for a number of
states.\7 In table II.5, we estimated that 22 states must spend more
per recipient than they spent per recipient under AFDC in 1996,
assuming state spending at 80 percent MOE. 



                               Table II.5
                
                 State Resource Levels Per Recipient--
                     Pre-Reform Versus Post-Reform

                                   State resources per recipient
                              ----------------------------------------
                                                               Percent
State                                 1996          1997    difference
----------------------------  ------------  ------------  ------------
Alabama                               $500          $565            13
Alaska                               1,795         1,551           -14
Arizona                                758           735            -3
Arkansas                               494           432           -13
California                           1,374         1,277            -7
Colorado                             1,401         1,494             7
Connecticut                          1,656         1,293           -22
Delaware                             1,497         1,063           -29
Florida                                843           971            15
Georgia                                650           759            17
Hawaii                               1,521         1,048           -31
Idaho                                  870         1,849           113
Illinois                               999           835           -16
Indiana                                685         1,128            65
Iowa                                   962           895            -7
Kansas                               1,070         1,389            30
Kentucky                               551           476           -14
Louisiana                              272           331            22
Maine                                  890           896             1
Maryland                             1,118         1,224            10
Massachusetts                        1,721         1,947            13
Michigan                               964         1,177            22
Minnesota                            1,348         1,268            -6
Mississippi                            217           266            22
Missouri                               722           704            -2
Montana                                680           663            -3
Nebraska                             1,345           825           -39
Nevada                               1,094           975           -11
New Hampshire                        1,675         1,958            17
New Jersey                           1,430         1,349            -6
New Mexico                             574           574             0
New York                             1,996         1,820            -9
North Carolina                         915           710           -22
North Dakota                         1,219           921           -24
Ohio                                   802           928            16
Oklahoma                               768           876            14
Oregon                               1,375         1,748            27
Pennsylvania                         1,385         1,003           -28
Rhode Island                         1,274         1,182            -7
South Carolina                         507           479            -6
South Dakota                           810           749            -8
Tennessee                              536           541             1
Texas                                  492           524             6
Utah                                 1,018           844           -17
Vermont                              1,223         1,221             0
Virginia                               873         1,145            31
Washington                           1,541         1,215           -21
West Virginia                          445           434            -2
Wisconsin                            1,073         1,798            68
Wyoming                                876         2,295           162
======================================================================
Total                               $1,125        $1,110            -1
Average                             $1,016        $1,047             6
Median                                $963          $973            -2
----------------------------------------------------------------------
Source:  GAO analysis based on data reported by states to HHS,
Administration for Families and Children.  These data are reported by
the states to HHS and have not been independently verified by GAO. 

Note:  The "total percent difference" presented in this table
represents the percent difference in the nationwide totals.  The
"average percent difference" is a simple average of the percentage
differences across states, with each state having equal weight. 

COMBINED IMPACT OF CASELOADS, TANF
GRANTS, AND MOE ON TOTAL BUDGETARY
RESOURCES

Another way to estimate the total resources available for welfare
programs is to compare total federal and state resources available
under the block grant with what comparable federal-state spending
would have been for 1997 caseloads under AFDC.  To estimate changes
in total available budgetary resources, we began by constructing a
current-services baseline for pre-reform spending.  We constructed
our baseline by adding actual state and federal expenditures in
federal fiscal year 1996 for the programs TANF replaced.\8 We
calculated total spending per recipient and then adjusted all
baseline components for inflation except cash assistance.\9 Finally,
to take recent caseload declines into account, we applied these per
recipient costs to 1997 caseloads. 

Using a states' total annual TANF grant, we calculated the federal
contribution to the total resources available.\10 Since the federal
contribution is now a block grant, these funds are available,
irrespective of the needs in a state.  Once again, we assumed that
states would budget at 80 percent MOE.  Since the MOE requirement
establishes a minimum, or floor, on state spending, a state can spend
more than minimally required if it chooses--raising the total levels
of budgetary resources available. 

Table II.6 presents our estimates of the total additional budgetary
resources available to states to design, finance, and implement their
family assistance programs due to TANF.  These estimates represent
the difference between states' post-reform total budgetary resources
(TANF plus MOE) and what they would have budgeted for their 1997
caseloads if they were still using the pre-reform 1996 cost
structure.  That is, table II.6 shows "additional resources" as the
difference between states' new total budgetary resources and our
construction of the current services baseline.  The analysis, which
takes caseload declines into account, suggests an even greater change
in resources than merely looking at nominal changes in federal and
state resources.  Combining the effects of the increased federal
resources and the act's mandated floor on state spending, our
analysis indicates that 46 states will have more total--federal TANF
and state MOE--resources available than they would have had without
reform.  Our estimates of these additional budgetary resources
totaled about $4.7 billion--or, on average, states will have 25
percent more in total budgetary resources available for their welfare
reform programs.  As with the other analyses, there is wide variation
among states--ranging from 102 percent in additional resources for
Wyoming to total fewer resources in Delaware, Hawaii, Nebraska, and
Pennsylvania. 



                               Table II.6
                
                  Total Additional Budgetary Resources

                                                  Additional budgetary
                                                        resources as a
                                      Additional            percent of
                                       budgetary   constructed current
State                                  resources     services baseline
--------------------------------  --------------  --------------------
Alabama                              $49,625,502                    40
Alaska                                   968,234                     1
Arizona                               69,179,824                    22
Arkansas                               7,860,157                    10
California                           548,805,296                     8
Colorado                              64,026,861                    24
Connecticut                            5,531,414                     1
Delaware                                -931,459                    -2
Florida                              271,005,350                    28
Georgia                              178,444,087                    35
Hawaii                               -40,945,164                   -21
Idaho                                 29,518,102                    58
Illinois                              68,465,351                     6
Indiana                              179,169,795                    88
Iowa                                  29,756,018                    14
Kansas                                64,178,595                    42
Kentucky                              29,900,372                    11
Louisiana                             87,381,046                    48
Maine                                 21,349,748                    18
Maryland                             106,872,002                    25
Massachusetts                        243,326,902                    33
Michigan                             421,059,851                    39
Minnesota                             65,107,184                    15
Mississippi                           46,934,247                    49
Missouri                              68,048,494                    19
Montana                               15,712,420                    26
Nebraska                             -10,305,381                   -10
Nevada                                14,185,455                    17
New Hampshire                         20,780,275                    29
New Jersey                           127,134,572                    17
New Mexico                            40,342,705                    22
New York                             364,994,597                     8
North Carolina                        25,539,979                     5
North Dakota                           5,111,369                    13
Ohio                                 365,562,536                    38
Oklahoma                              80,850,010                    41
Oregon                               109,551,976                    43
Pennsylvania                         -13,898,411                    -1
Rhode Island                          21,404,178                    14
South Carolina                        36,150,049                    23
South Dakota                           7,012,126                    21
Tennessee                             94,180,795                    31
Texas                                239,221,315                    32
Utah                                  23,115,769                    22
Vermont                               12,842,716                    18
Virginia                             106,414,336                    41
Washington                            24,360,705                     3
West Virginia                         37,688,797                    29
Wisconsin                            270,604,114                    65
Wyoming                               24,380,113                   102

======================================================================
Total                             $4,657,574,920                    16
Average                              $93,151,498                    25
Median                               $43,638,476                    22
----------------------------------------------------------------------
Source:  GAO analysis based on data reported by states to HHS,
Administration for Children and Families.  These data are reported by
the states to HHS and have not been independently verified by GAO. 

Notes:
1.  Differences between our estimates of additional budgetary
resources and those presented in state budget documents are the
result of a variety of factors including:  (1) differences between
the state fiscal year and the federal fiscal year, (2) the difference
between a state's total TANF grant and the amount it was eligible to
receive in federal fiscal year 1997, and (3) assumptions made by
state budget analysts about the effects of program reforms in the
state's baseline that might not have been included in the expenditure
data and assumptions used in our estimates.

2.  The "total" presented in this table represents the percent
difference in the nationwide totals.  The "average" is a simple
average of the percentage differences across states, with each state
having equal weight. 



(See figure in printed edition.)Appendix III

--------------------
\1 This analysis compared actual 1996 federal expenditures for the
programs TANF replaced with states' full year TANF allotments.  These
allotments may not represent resources states actually received in
1997 because states became eligible to receive TANF at different
points during the year depending on when they filed their state
programs plans with HHS.  Sixteen states were eligible for their full
annual TANF grant award for all of federal fiscal year 1997 and 28
states for at least 9 months. 

\2 Under AFDC, caseloads were defined as the number of families
receiving monthly cash assistance.  Although many states had
implemented TANF provisions before or by July 1997, the caseload data
for that month most likely is still based on this pre-PRWORA
definition of cash assistance.  Under HHS-proposed regulations for
TANF, issued in November 1997, the definition of assistance has been
broadened to include ongoing financial support, including work
subsidies and assistance with child care, in addition to ongoing cash
assistance.  As a result, future caseload data will include those
receiving these types of assistance.  While the proposed regulations
note that states may also assist individuals in other ways, such as
with counseling, case management, and employment services and
one-time, short-term financial aid, such as automobile repair, states
will not be required to report on the numbers receiving these forms
of assistance.  As such, these recipients will not count as part of
the TANF caseload even though they are receiving benefits funded with
TANF or state MOE resources. 

\3 A recent study from the Urban Institute found that the
relationship between the caseload decline and the change in federal
funding is not perfectly correlated.  They noted that many of the
states that received less funding actually experienced declines in
AFDC caseloads from their TANF grant base years to 1996 but still
received less total welfare funding in 1997 (relative to 1996)
because they spent more on emergency assistance and/or administrative
costs in 1996 than in their TANF grant base years.  See Gordon Mermin
and C.  Eugene Steuerle, "The Impact of TANF on State Budgets," Urban
Institute Number A-18 in the Series, New Federalism:  Issues and
Options for States (November 1997). 

\4 If a state meets certain work participation rates, its MOE
requirement drops to 75 percent.  However, we assumed that states
would budget conservatively at the higher rate since they could not
know at the start of a fiscal year whether they would meet the work
participation rates.  This assumption produces a conservative
estimate of the additional state resources since states can spend
more than minimally required.  In federal fiscal year 1997, 21 states
reported spending more than 80 percent, 13 reported spending 80
percent, and 16 states met their work participation rates and were
able to take advantage of the lower MOE requirement. 

\5 As noted in the letter, the MOE is based on spending in those
programs consolidated in the TANF block grant as well as spending on
programs that were combined in the Child Care and Development Block
Grant (CCDBG)--AFDC-related Child Care, Transitional Child Care, and
At-Risk Child Care. 

\6 A current-service baseline assumes the continuation of current
policies and reflects anticipated costs of ongoing programs and
activities without policy changes.  It generally includes allowances
for inflation and changes in caseload.  Proposed policy changes that
would affect the costs of programs are compared to the
current-services baseline to estimate the budgetary impact. 

\7 In this analysis, we used AFDC cash-benefits caseload data.  Since
the programs combined in the CCDBG included populations in addition
to those receiving AFDC cash benefits, the data for actual spending
per recipient will appear inflated.  However, the comparison between
spending required under PRWORA and spending under the consolidated
and combined programs is more meaningful if the same caseload data
are used in the denominator.  Furthermore, state TANF-MOE spending on
child care also can be counted toward receiving CCDBG matching funds. 
In effect, these state funds are double counted; once to qualify for
the federal TANF grant and once to obtain federal CCDBG matching
funds.  For more information, see Welfare Reform:  States' Efforts to
Expand Child Care Programs (GAO/HEHS-98-27, January 13, 1998). 

\8 We used federal fiscal year 1996 expenditure data because it
captured spending in the last full federal fiscal year of the
categorical matching grant and was comparable across states.  Our
construction of states' baseline does not include state or federal
spending on the three child care programs combined in the CCDBG.  In
this analysis, we sought to compare the pre-PRWORA AFDC program with
the post-PRWORA TANF program and to minimize the effect of
interactions between TANF and CCDBG. 

\9 We did not adjust cash assistance for inflation because nominal
average cash benefits have been declining in real terms since 1970. 
We used an inflation adjustment of 2.1 percent, based on the increase
in the GDP price index (chain weights) from federal fiscal years 1996
to 1997. 

\10 States' TANF allotments will remain unchanged for the 6 years of
the grant. 


COMMENTS FROM THE DEPARTMENT OF
HEALTH AND HUMAN SERVICES
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)


The following are GAO's comments on the Department of Health and
Human Services' letter dated July 23, 1998. 

GAO COMMENTS

1.  See "Agency Comments" section of the report. 

2.  Text (now on page 3) amended. 

3.  Text of Table 1 of page 18 changed to reflect that figures
represent state reserves and are distinct from the Federal
Contingency Fund for State Welfare Programs. 

4.  HHS refers to 409 (a)(7)(B)(i) to suggest that state MOE funds
may only be used on 4 designated activities; (aa) cash assistance,
(bb) child care assistance, (cc) educational activities designed to
increase self-sufficiency, job training, and work .  .  .  , (dd) and
administrative costs in connection with the matters described in
items (aa), (bb), (cc), and (ee) .  .  ." HHS omits (ee) from its
list of qualified state expenditures.  This part allows states to
spend their own funds in any manner that is reasonably calculated to
accomplish the purpose of TANF. 

In its proposed rule (see 273.2), HHS interprets 409(a)(7)(B)(i) to
mean that a state may count as MOE its expenditures under all state
programs, i.e., the state's TANF program as well as any separate
state program that assists "eligible families" and provides
appropriate services or benefits.  Thus, while MOE funds must be used
on eligible families (as defined by the state) and on activities that
can reasonably be calculated to accomplish the goals of TANF, they
can be used to provide support to certain categories of clients that
are prohibited from receiving federal TANF assistance.  If states
choose to operate separate state programs, they have more flexibility
in the use of state funds than they have in the use of federal funds. 
We continue to believe that these differences will have an impact on
the choices states make with regard to their programs, specifically
the mix of services they can offer and the people they can serve. 

5.  Text (now on page 24) changed to reflect that HHS concurs with
CBO that allowing states to count the same expenditures toward both
the TANF MOE and the Contingency MOE would increase the costs of the
Contingency Fund in budget scoring terms and could be subject to a
challenge under the budget rules unless offsets were found. 

6.  Text (now on page 25) amended to reflect new information. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix IV

ACCOUNTING AND INFORMATION
MANAGEMENT DIVISION, WASHINGTON,
D.C. 

Thomas M.  James, Assistant Director
Bill J.  Keller, Evaluator-In-Charge
Linda F.  Baker, Senior Evaluator

SAN FRANCISCO FIELD OFFICE

Raymond G.  Hendren, Senior Auditor


*** End of document. ***