< Back to Search

TANF Final Rule – Pages 17769 - 17818

Published: July 10, 2012
Tags:
TANF Final Rule

[Federal Register: April 12, 1999 (Volume 64, Number 69)]
[Rules and Regulations]
[Page 17769-17818]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr12ap99-24]

[[pp. 17769-17818]] Temporary Assistance for Needy Families Program (TANF)


[[Continued from page 17768]]
[[Page 17769]]

option) with respect to any period during the month in which the
individual refused, subject to good cause and other exceptions
determined by the State. These exceptions include the statutory
exception for single custodial parents of children under the age of six
who cannot obtain needed child care, which is included in the
regulations at Sec. 261.15. The State also has the option to terminate
the case.
    In addition to the child care exception, each State may establish
its own criteria for determining when not to impose a penalty on an
individual, that is, when an individual has ``good cause'' for not
engaging in work. States may also establish other rules governing
penalties as needed.
    Under the Family Violence Option, a State may waive work
requirements in cases where compliance would make it difficult for an
individual to escape domestic violence or would unfairly penalize
individuals who are or have been victimized by such violence or
individuals who are at risk of abuse. The State must determine that the
individual receiving the program waiver has good cause for failing to
engage in work.
    The final regulations include a cross-reference to the State
penalty for failure to impose sanctions in accordance with section
407(e) of the Act (at Sec. 261.54). We added this reference for the
convenience of the reader; it does not represent an additional
requirement.
    Comment: We received many comments urging us to change the language
of the regulations concerning the pro rata reduction of a recipient's
assistance. The commenters thought that the way in which we paraphrased
the statute altered its meaning and excluded certain types of pro rata
reductions. Most urged us to clarify that a State can make a pro rata
reduction based on any reasonable method; some asked us to indicate
that a pro rata reduction is based on the head-of-household's share of
assistance or on the share of those refusing to work. A few commenters
also noted that States should have the flexibility to define the
timeframe for applying a pro rata reduction. Several commenters
suggested that the NPRM inappropriately restricted a State's ability to
impose a greater penalty.
    Response: We recognize that the language we used in the NPRM may
have caused confusion concerning the meaning of a pro rata reduction,
and we have modified the regulations to reflect the statutory language
more closely. It was not our intention to prescribe one method of
proration or to proscribe other legitimate methods; a State may make a
pro rata reduction based on any reasonable method. With respect to
imposing a greater penalty, we think that the NPRM's regulatory text
and preamble were very clear that a State could impose a penalty
greater than a pro rata reduction, up to and including terminating the
case, and thus have not substantially altered the regulations in that
regard.
    Comment: One commenter, concerned about the burden on caseworkers
of tracking an individual's participation, urged us to establish
specific, fixed penalties on an individual for certain periods of time
for refusal to work. The commenter gave an example of reducing the
grant by the individual's share for the first month of refusal and
gradually increasing it.
    Response: As we indicated above, a State may establish any method
of pro rata reduction that it chooses that comports with section 407(e)
of the Act. Since we do not intend to dictate one proration method over
another, it would not be appropriate to adopt the penalty scheme that
the commenter suggests.
    Comment: Several commenters expressed the same concern in this
section that they did in Sec. 261.13 regarding the applicability of
employment protections to welfare recipients. They urged us to ensure
that good cause exceptions in this section protect recipients from
penalty where the individual refused to work due to a violation of
employment laws, such as sexual harassment or other forms of job
discrimination. Others urged us to provide guidance about appropriate
good cause exceptions.
    Response: States have the flexibility to define ``good cause'' as
they deem appropriate. Because of the States' extensive experience in
this area, we think it is not necessary to provide specific guidance
regarding what good cause exceptions a State should acknowledge.
However, we have included a new regulatory section at Sec. 260.35 to
reference employment protections under other laws that apply to working
welfare recipients. We certainly agree that welfare recipients should
not have to choose between unsafe or discriminatory working conditions
and losing benefits, especially where there are protections under
Federal law.
    Comment: A commenter urged us to exempt from the work requirements
any foster parents with birth children in the home.
    Response: The statute does not provide for an exemption from the
work requirements for such individuals; however, States may define
``good cause'' as they find appropriate. Since the statute specifically
gives States the authority to establish good cause and other
exceptions, we do not intend to dictate specific good cause criteria,
other than the child care exception provided for at section 407(e)(2).

Section 261.15--Can a Family Be Penalized if a Parent Refuses To Work
Because He or She Cannot Find Child Care? (Sec. 271.15 of the NPRM)

    A State may not reduce or terminate assistance to a single
custodial parent caring for a child under age six for refusing to
engage in required work, if the parent demonstrates an inability (as
determined by the State) to obtain needed child care. This exception
applies to penalties the State imposes for refusal to engage in work in
accordance with either section 407 or section 402(a)(1)(A)(ii) of the
Act. The parent's demonstrated inability must be for one of the
following reasons:
    <bullet> Appropriate child care within a reasonable distance from
the individual's home or work site is unavailable;
    <bullet> Informal child care by a relative or under other
arrangements is unavailable or unsuitable; or
    <bullet>  Appropriate and affordable formal child care arrangements
are unavailable.
    This penalty exception underscores the pivotal role of child care
in supporting work and also recognizes that the lack of appropriate,
affordable child care can create unacceptable hardships for children
and families.
    We have substantially modified this section of the regulations, in
part by moving much of what constituted Sec. 271.15 under the NPRM to a
new section, Sec. 261.56. This new section specifies the State's
responsibilities in carrying out the penalty exception, while
Sec. 261.15 describes the impact of the provision on the individual. We
have also moved the State penalty provision associated with this child
care exception (formerly Sec. 274.20) to a newly created Sec. 261.57.
Our intent in making these changes is to preserve the informational and
contextual nature of subpart A of part 261 and to make the State's
responsibilities and the possible penalty associated with them easier
to follow. In this section of the rule, we have added cross-references
to these two new sections for clarity.
    Readers can find all comments associated with this exemption in the
preamble discussion for Sec. 261.56.

Section 261.16--Does the Imposition of a Penalty Affect an Individual's
Work Requirement? (Sec. 271.16 of the NPRM)

    Section 408(c) of the Act, as amended by section 5001(h) of Pub. L.
105-33,

[[Page 17770]]

clarifies that penalties against recipients under TANF ``shall not be
construed to be a reduction in any wage paid to the individual.'' In
the NPRM, we indicated that imposing such a penalty does not require
the State to reduce the number of hours of work required, as it would
otherwise do if the individual's wages decreased, due to the provisions
of the Fair Labor Standards Act.
    In the final rule, we have modified this section of the regulations
to reflect the statutory language more precisely. This change does not
signify any shift in our interpretation of the provision: we continue
to believe that Congress intended to permit a State to sanction an
individual who is subject to the Fair Labor Standards Act (FLSA)
without also being forced to reduce the individual's required hours of
work. FLSA requirements, including the Federal minimum wage, apply to
any welfare recipients that meet the broad definition of ``employees''
under that law, which includes participants in many work activities. By
indicating that a penalty does not reduce the individual's wages, the
State does not need to recalculate hours of work subject to FLSA. A
State is, of course, free to decide to reduce the work hours of a
sanctioned individual or to reassign the individual to activities that
are not subject to FLSA.
    In addition to the comments described below, we received several
comments expressing support for the inclusion of this provision in the
regulations. Others indicated that some readers were confused by the
intent of this section; we hope the explanation above and the change in
the regulatory text have reduced this confusion.
    Comment: Some commenters urged us to delete the last clause in
Sec. 271.16 of the NPRM, which indicated that a penalty would not
result in a reduction in the number of hours of required work. Others
asked us to substitute the word ``participation'' for the word ``work''
in that clause.
    Response: We have removed the last clause from the regulation
because we did not want to preclude a State from reducing an
individual's hours of work.
    Comment: Some commenters thought this provision would act as an
incentive for States to penalize recipients to avoid the minimum wage
requirements and urged us to monitor sanctions under this provision by
collecting data on State sanctions. Another commenter inquired whether
this provision applied where the penalty is disqualification of the
individual from the program, such as for an intentional program
violation.
    Response: The commenters seem to be suggesting that a State would
have an incentive to penalize a recipient because this provision
prevents the State from considering the penalty to be a reduction in
wages and therefore it could engage the recipient in hours of work for
which he or she is not compensated. We do not agree. An individual's
hours of work are established in accordance with the FLSA based on the
benefits the family receives, long before and independent of the
sanctioning process. The State may only impose a work sanction for
failure to engage in required work. If an individual thinks that the
State has penalized him or her inappropriately, he or she has recourse
to appeal the sanction decision; section 402(a)(1)(B)(iii) of the Act
requires the State to provide opportunities for recipients who have
been adversely affected to be heard in a State administrative or appeal
process.
    With respect to monitoring sanctions and application of this
provision, readers should understand that no individual is sanctioned
``under this provision''; rather, this provision applies to any
recipient who is sanctioned. Thus, no sanctioned recipient is
considered to have had a reduction in wages as a result of the penalty.
    Readers should also note that we have improved the information we
are collecting about sanctions and should refer to Appendix A for
further discussion of these data requirements.
    Comment: Some commenters urged us to clarify that a penalty against
a family is not a reduction in assistance or other payments. They
thought the phrase ``reduction in any wage paid to the individual''
raised doubt about this point. One commenter specified that States
should be relieved of FLSA liability regardless of whether the
individual is in a wage or nonwage work assignment.
    Response: We think the language of the provision is clear and does
not need further interpretation. As we indicated above, the FLSA
requirements apply to any welfare beneficiaries that meet the broad
definition of ``employees'' under that law; thus, the term ``wage'' is
the appropriate one to use.
    Comment: One commenter thought we should specifically state that
the FLSA does not apply where the State has sanctioned an individual,
so as to protect a State from reducing an individual's hours out of
fear of violating the FLSA to the point where he or she would no longer
count toward the participation rate. As an alternative, the commenter
suggested that we deem an individual's hours of work, as determined by
the FLSA, as automatically meeting the work requirement or give States
broader authority to include the value of other benefits when
calculating an individual's work obligation.
    Response: Because the FLSA includes other provisions not affected
by this provision, it would not be accurate to state that the FLSA does
not apply. We think the regulatory language explains the interaction of
the FLSA and this provision adequately. Regarding the commenter's
suggested alternative, the statute is very clear about the number of
hours an individual must be engaged in work to count toward the
participation rate (see subpart B). Regarding the comment suggesting
broad authority for a State to include other benefits in calculating an
individual's work obligation, this matter is governed by the FLSA and
thus is outside the scope of these regulations.
    Comment: One commenter urged us to clarify that, although a
sanction would not result in a reduction in the number of required
hours of work, it might result in a reduction in certain activities, in
order to comply with Federal, State and local labor laws.
    Response: As we have indicated, this provision is intended to avoid
forcing a State to reduce the hours an individual must work because his
or her benefits decreased as a result of a penalty imposed under TANF,
as it would otherwise have to do in accordance with Federal labor law.
If the State chooses to reduce the individual's hours of work, or to
shift the individual to other appropriate activities, it has the
flexibility to do so. If there are State or local labor laws that
restrict the State's actions in this area, it is the State's
responsibility to adhere to applicable laws.

Subpart B--What Are the Provisions Addressing State Accountability?

Section 261.20--How Will We Hold a State Accountable for Achieving the
Work Objectives of TANF? (Sec. 271.20 of the NPRM)

    Work is the cornerstone of welfare reform. Research has
demonstrated that early connection to the labor force helps welfare
recipients make important steps toward self-sufficiency. The rigorous
work participation requirements embodied in the legislation provide
strong incentives to States to concentrate their resources in this
crucial area.
    This summary section makes the legislation's focus on work and the
requirements for work clear, while other sections address each of these
areas in more detail.

[[Page 17771]]

    This section describes what a State must do to meet the overall and
two-parent work participation rates. It explains that a State must
submit data to allow us to measure each State's success with the work
participation rates. It notes that a State meeting the minimum rates
will have a reduced MOE requirement, while a State failing to meet them
risks a financial penalty.
    We received only one comment relating to this section alone.
    Comment: Regarding the reference to data that a State must submit
for us to calculate the participation rates, the commenter contended
that the process for calculating the participation rates is too
complicated. As an alternative the commenter suggested that a State
should calculate its own participation rate, which we should then
review.
    Response: Section 411(a) of the Act requires States to report to us
various data necessary to calculate the participation rates. Therefore,
we think that it is clear that Congress intended us to make the
calculations of the participation rates and gives us the authority to
specify the data elements we need. As we have done prior to the
publication of final regulations, we will continue to work in
partnership with States to ensure that data are accurate and correctly
portray their participation rates.

Section 261.21--What Overall Work Rate Must a State Meet? (Sec. 271.21
of the NPRM)

    Section 407(a) of the Act establishes two minimum participation
rates that a State must meet beginning with FY 1997.
    The first, the overall work rate, is the percentage of all families
receiving assistance who must participate in work activities by fiscal
year. This section lists the statutory overall participation rate that
applies to each fiscal year.
    The second is the work rate for two-parent families, which we
address at Secs. 261.23 and 261.24.
    We received no comments concerning this section.

Section 261.22--How Will We Determine a State's Overall Work Rate?
(Sec. 271.22 of the NPRM)

    This section of the regulation restates in clear terms the
participation rate calculation specified in the statute. In particular,
without changing its meaning, we have phrased the denominator in a way
that we think is easier to understand than the statutory language.
    We received many requests for guidance concerning how, for purposes
of the participation rates, we treat a family that the State exempts
from work requirements.
    A State has the flexibility to establish any exemptions it chooses;
however, with two exceptions (discussed below), the legislation offers
no room to remove categories of recipients from the denominator, as
prior law did. PRWORA embodies the views that: (1) Work is the best way
to achieve independence; and (2) each individual should participate to
his or her greatest ability. As waiver projects have demonstrated,
innovative State programs can often find meaningful ways for nearly
every recipient to participate in work-related activities. Therefore,
the statute and the regulation require nearly all families to be
included in the calculation of the participation rates.
    The two exceptions to this requirement are certain families that
are subject to a penalty and, at State option, families in which a
single custodial parent is caring for a child under 12 months of age.
When directed by the State's reported data to do so, we will disregard
from the calculation for a month--that is, not include in either the
numerator or the denominator--families: (1) Receiving assistance that
are subject to a penalty for refusing to engage in work required in
accordance with section 407 of the Act, but that have not been subject
to a penalty for more than three of the last 12 months; and (2) in
which a single custodial parent is caring for a child under one year of
age. The latter exception is limited by statute to a maximum of 12
months for any parent. Although the first exception is not a State
option under the statute, a State may choose to include a sanctioned
family in the rate even though it has been subject to a penalty for
three or fewer months in the last 12 because the family is nevertheless
working enough hours to count toward the participation rate. In such a
situation, we would include the family in both the numerator and
denominator of the calculation.
    The policy described above with respect to families subject to a
penalty is slightly different from that of the NPRM. We are removing
``excepted'' families from the entire calculation, rather than just the
denominator. We have made this change after reexamining Congressional
intent. We think it unlikely that very many individuals would have been
subject to a sanction while still working sufficient hours to count in
the numerator, but we believe it would not be consistent with
Congressional intent to permit inclusion in the numerator but not in
the denominator. By creating the exception to inclusion in the
denominator, Congress intended to avoid penalizing a State when it
tries to get a nonparticipating individual to participate. However,
Congress did not intend to create an advantage for such a State by
allowing ``excepted'' individuals to be included in the numerator when
they were not in the denominator. Therefore, if a State wishes to count
a family in the numerator, that family must also appear in the
denominator.
    The regulation makes clear that a State may count as a month of
participation any partial months of assistance, if, in each full week
of assistance in that month, an adult in the family is engaged in work
for the minimum weekly average number of hours. These families are
already included in the denominator since they are recipients of
assistance in that month.
    This provision ensures that a State receives credit for its efforts
in the first and last months that a family receives assistance. Without
it, a State would have an inadvertent incentive to start and end
assistance as close as possible to the beginning of the month, rather
than as families need it. We think that measuring work in full weeks of
assistance during a partial month is consistent with the spirit of
PRWORA. We have established the same policy for partial months of
assistance under the two-parent rate at Sec. 261.24.
    In the preamble to the proposed regulation for this section, we
included a significant discussion about the relationship among waivers
granted under the Family Violence Option (FVO), work participation
rates, and a State's access to penalty forgiveness under ``reasonable
cause.'' We recognized that there were circumstances under which a
State should and would temporarily waive work requirements for domestic
violence victims. Two questions we considered were: (1) How such
waivers would affect the calculation of the participation rates; and
(2) how they would affect a State's penalty liability.
    As we discussed earlier in the preamble, instead of changing the
basic calculation of the work participation rates, we chose to address
this situation through our penalty liability determinations. We chose
this targeted approach so as not to provide blanket exemptions for
those who have ever suffered domestic violence, but instead to provide
appropriate protections and supports for TANF recipients who need them.
    Because of the nature of the comments we received on the domestic

[[Page 17772]]

violence provisions in the proposed rule, we decided to consolidate the
discussion of those comments in the preamble and to consolidate the
regulatory provisions in a new subpart B of part 260. You can find the
consolidated preamble discussion in the earlier section entitled
``Treatment of Domestic Violence Victims.''
    As the result of the comments and the changes we made to part 260
of the rule, we have also revised the language that was proposed at
Sec. 271.52(b)(1). Under the revised language, we no longer define the
criteria for ``reasonable cause'' related to federally recognized
domestic violence waivers in this section, but cross-reference the
regulatory provisions in part 260. Also, we have added language to
Sec. 261.52 indicating we would take waivers of work requirements
granted under subpart B of part 260 into account in deciding if a State
is eligible for a penalty reduction based on the degree of its
noncompliance. Please see Sec. 261.52 for further discussion of these
issues.
    We received many comments concerning our proposal to redefine
``family'' to include in the participation rate any families the State
has excluded (based on defining a family as ``child-only'') for the
purpose of avoiding a penalty. We have removed this provision from this
section, as well as from Sec. 261.24 describing the two-parent
participation rate. Please refer to the earlier preamble discussion in
the section entitled ``Child-Only Cases'' for further discussion of
this decision and the comments that relate to it.
    Comment: One commenter thought that the overall participation rate
as we described it in this section could be interpreted as either
having a State average the 12 monthly rates or calculate a weighted
average, taking caseload size into account.
    Response: The statute does not provide for a weighted average in
calculating the participation rates; rather, it specifically states
that the annual rate is the average of the State's monthly rates for
the fiscal year. Moreover, readers should understand that States are
not responsible for calculating the participation rates. We calculate
the rates based on the data that States report to us. For further
discussion of the required data and reporting provisions, please refer
to part 265 of this chapter.
    Comment: Several commenters suggested that we exclude certain
groups of individuals from the participation rate, in addition to those
specified in the regulations. In particular, various commenters urged
us to remove from the rate calculation: women in the third trimester of
a pregnancy; cases that include a child and a grandparent who is over
60 years of age; families not receiving cash assistance; individuals
working for employers that engage in discriminatory conduct; cases
engaged in federally mandated administrative reviews prior to a
sanction; and individuals who have received assistance for fewer than
60 days and therefore are not required to participate. Another
commenter agreed with our statement that States should establish
whatever exemptions they choose, but thought those State-exempted
individuals should be removed from the rate calculation.
    Response: As we indicated in the NPRM and the above discussion, we
believe the statute is very clear regarding the calculation of the
participation rates and does not give us the flexibility to exclude
additional categories of individuals from the calculations. The
participation rates are written in terms of ``families receiving
assistance'' that include an adult, thus we could not limit the rates
to those receiving cash assistance. (For further discussion of the
definition of assistance, please refer to Sec. 260.30 of this chapter.)
    Concerning individuals in work activities where the employer
engages in discriminatory conduct, again, we do not think we have the
latitude to remove such families from the denominator; however, we
fully expect States to conduct programs that are lawful and uphold
employment laws that apply to working welfare recipients. Please refer
to the section entitled ``Recipient and Worker Protections'' for a more
detailed discussion of this issue.
    It is not entirely clear to us what the commenter means by
``federally mandated administrative review process prior to being
placed in sanction.'' There is no longer a federally mandated
conciliation process, as there was under the JOBS program. It is
possible that the commenter is referring to the provision at section
402(a)(1)(B)(iii) of the Act, requiring an explanation in the State
plan of how the State will provide opportunities for recipients who
have been adversely affected to be heard in a State administrative or
appeal process. If so, recipients appealing an adverse action may
already be under a sanction and therefore would not be included in the
rate, if they have not been subject to one for more than three months
in the last 12. Further, there is nothing in the statute to suggest
that State administrative or appeal process should be lengthy; on the
contrary, we hope States will establish expedited processes, in the
interests of both the families and the State. We think there is neither
the need nor the authority to remove such families from the
participation rates.
    Regarding the commenter's concern that individuals are not required
to participate in work activities until they have received 60 days of
assistance, the commenter is confusing the requirement on individuals
to work with the requirement on States to achieve certain participation
rates. Although the activities may be the same, they are separate
requirements under the law. Please refer to the discussion at
Sec. 261.10 for further explanation of this distinction.
    The statute is clear in giving a State the flexibility to establish
``such good cause and other exceptions'' as it chooses, but does not
remove those with good cause exceptions from the rate calculations. We
encourage States to adopt fair and practical good cause exceptions.
While we understand the commenter's concern that a State has no
incentive to create good cause exceptions if the excepted families
remain in the denominator, it is worth noting that the overall
participation rate leaves room to grant good cause exceptions under a
variety of different circumstances.
    Comment: A commenter suggested that there should be follow-up on
individuals for three months following employment and that such
individuals should be included in the participation rate as an
incentive to States to find employment for recipients. The commenter
stated that currently individuals are not included in the rate once
they become employed.
    Response: Neither the statute nor the regulations excludes employed
recipients from the participation rate, as long as they are still
actually receiving TANF assistance. In fact, unsubsidized employment is
the first work activity that permits TANF recipients to be considered
``engaged in work'' and other forms of employment immediately follow
it. Moreover, recognizing that the participation rate calculations did
not give States credit for those who became employed and left the
welfare rolls, Congress created a ``caseload reduction credit'' for
that purpose. (See subpart D for discussion of the Caseload Reduction
Credit.)
    We do require States to collect data on families no longer
receiving assistance (please refer to Sec. 265.3), but we believe it is
burdensome and impractical to require all States to follow such
families for any period of time. We do agree that this is important
information in understanding the effect of the TANF program and
encourage States to

[[Page 17773]]

conduct follow-up studies where possible. Also, we have designed the
initial high performance bonus system to give us follow-up information
on the employment of recipients without imposing a substantial new
burden on State TANF agencies.
    Comment: One commenter stated that the denominator of the
participation rate changes daily and that we need a standardized
formula to allow programs to meet their goals. Another asked whether
the rate is calculated based on a sample or the universe of cases,
suggesting that the universe was preferable where feasible.
    Response: While the denominator of the participation rate can
change from month to month, States will have ongoing access to
information about their caseloads, which should enable them to adjust
for shifts in the number and types of cases. The participation rates
are based on monthly data of families receiving assistance that include
an adult. Therefore, a family that receives assistance for even one day
in a month contributes to the total number of families receiving
assistance in that month. We think the participation rate calculations
are quite clear. However, we have incorporated some opportunities in
the penalty relief provisions to consider a State's special
circumstances. For example, in reducing the work participation penalty,
the final rule adds a new adjustment factor that could help States that
substantially increase the number of participants, but fail the
participation rate because they are experiencing significant caseload
increases.
    Regarding whether a State should report the universe of caseload
data, Sec. 265.5 permits a State to report participation and other data
for the universe or a sample of cases and outlines acceptable sampling
methods. States should weigh the advantages and disadvantages of
sampling and make their own decisions about whether to report on a
universe or sample basis.
    Comment: One commenter urged modifying the regulations to ensure
that, if one parent in a two-parent family is subject to a penalty but
the other parent continues to work the minimum hours required for the
overall participation rate, the family should count toward the overall
rate. If the second parent subsequently is subject to a penalty, the
commenter thought we should measure the months of sanction in the last
12 months separately for each parent, thus maximizing the time a family
would be excluded from one or both participation rates.
    Response: First, we think it is clear in both the statute and these
regulations that families, and not individuals, are subject to
penalties. The State has the flexibility to determine the amount of the
penalty, up to and including terminating the case, but must impose a
penalty that is at least a pro rata reduction of the family's
assistance (see Sec. 261.14 for further discussion of pro rata
reductions). Thus, we would look at whether the family, not the
individual, is a sanction case.
    If the family continues to receive assistance and meets the
standard for being ``engaged in work'' under the overall rate while
being sanctioned, as it would in the commenter's example, then the
State may choose to count that family in the numerator and denominator
of the calculation. However, since it is a family and not an individual
that is subject to a penalty, should the other parent subsequently
refuse to work and the State take action, it would simply be a second
sanction for the family and does not call for separate tracking for
purposes of calculating the denominator.
    Comment: One commenter objected to the fact that two-parent
families are counted twice, once in the two-parent participation rate,
and once as part of the overall rate. The commenter thought that two-
parent families should be counted only in their own rate.
    Response: The composition of the overall participation rate is
statutory. The two-parent rate measures State success with that sub-
population, while the overall rate measures success with the entire
caseload of families that include an adult.
    Comment: Several commenters expressed support of the provision
excluding a single custodial parent caring for a child under 12 months
of age from the participation rate calculation. However, some
commenters thought that we should not tie the exclusion from the rate
to whether the State has adopted the option not to require the parent
to engage in work. In essence, they argue that there are two separate
decisions: whether to require the parent to work and whether to exclude
the parent from the rate. Others questioned whether this provision
allows for a one-time exclusion of up to 12 months or whether the
parent could be excluded again should he or she be caring for another
child under one year old.
    Response: Based on the comments and after reexamining the statutory
provision, we agree that we need not link the State's option not to
require a single custodial parent of a child under 1 to work to the
exclusion of such parents from the rate calculations. The State can
make separate decisions about exempting and excluding a family from its
rate. The statute describes a certain individual, that is, ``a single
custodial parent caring for a child who has not attained 12 months of
age'' and then separately indicates that ``such an individual'' may be
disregarded in calculating the participation rates. We have re-written
the regulation to allow disregard of a family with such an individual,
since the rates actually measure families and not individuals.
    Regarding whether this is a one-time provision or is renewable, the
law plainly states that a parent may be disregarded from the rate for
not more than 12 months. We interpret this language to mean a
cumulative, lifetime limit of 12 months for any single custodial
parent, but not necessarily a one-time disregard. Thus, if a parent
were disregarded from the rate for four months while caring for one
child under a year old, he or she could be disregarded for as much as 8
months with a subsequent baby.
    Comment: We received many comments in support of the provision to
give a State credit for a month of participation if the individual is
engaged in work for the minimum average number of hours in each full
week the family receives assistance in a partial month; however, some
commenters found the provision too narrow to accommodate States that
assign an individual to an activity weeks after the beginning of a
benefit period. Some urged us to count an individual's time in
assessment toward the participation rate. Another suggestion was that
we should only consider a month of assistance (partial or full) to
begin from the time the individual is assigned to a countable activity.
One commenter thought we should only count families in the denominator
from the first full month of assistance. One commenter asserted that we
should include only recipients, and no applicants, in the participation
rate; thus, this provision would affect only partial months following
approval of assistance.
    Response: The law and these regulations permit participation in
only 12 specific work activities to count toward the participation
rates. (Please refer to subpart C.) While we appreciate the time it
takes a State to assess an individual and assign him or her to an
activity, we do not have the flexibility to add assessment to the list
of allowable activities. By the same token, we cannot simply decide
that some period of time for which an individual receives assistance--
such as time prior to assignment in a work activity or a partial month
of assistance--should not

[[Page 17774]]

be considered a period of assistance and therefore exclude the
individual's family from the participation rate for that month. On the
contrary, if a family receives assistance for any portion of a month,
then we must include the family in the denominator of the participation
rate for that month, subject to the caveat in the paragraph below.
    With respect to the assertion that we should not include applicants
in the participation rate, we agree that States should not be forced to
count individuals in the participation rates while their applications
are pending. At the same time, we do not want to deny States that are
successful in moving applicants into work activities credit for their
efforts. It is for this very reason that we wanted to give States
credit in the participation rates for a partial month of assistance
where an adult works at a level equivalent to the standard for a full
month. Further, under these final rules, we will give States some
discretion to decide when a family begins to receive assistance, for
the purposes of the participation rates. If a State pays benefits
retroactively, i.e., for the period between application and approval,
the State would have the option to consider the family to be receiving
assistance either during the retroactive period or only during the
month of payment.
    This comment included an example in which the State ``prorated
[benefits] from the date of application,'' even though it did not
approve the application until about four weeks later. Each State has
some flexibility to decide when benefits begin; in this example, the
State chose the date of application. The statute is unclear whether
receipt of assistance for a prior period is assistance in that prior
month or only during the month of payment. Thus, when a State chooses
to pay retroactively back to the date of application, it has the option
to choose whether the recipient is receiving assistance during the
month or part of the month covered by the retroactive payment. Because
many States require applicants to engage in some form of work, such as
job search, this partial month provision should prove to be an
advantage for States that pay benefits retroactively for the
application period.

Section 261.23--What Two-Parent Work Rate Must a State Meet?
(Sec. 271.23 of the NPRM)

    As in Sec. 261.21, this section restates the minimum work
participation rates for two-parent families established in the statute.
    As States are aware, the two-parent participation rate increases
sharply. Congress has high expectations that States will help the vast
majority of adults in two-parent families find jobs or participate in
other work activities. We note that most States had difficulty meeting
the less ambitious JOBS participation rates for unemployed parent
families (UPs), the primary two-parent cases under AFDC, and about half
the States subject to the rates in FY 1997 failed the two-parent TANF
participation rate. For several reasons, the new rates under TANF are
much more demanding than they were under JOBS. First, the TANF rate is
a ``two-parent'' rate, not a rate just for UPs. Secondly, the
denominator includes much more of the caseload; it recognizes many
fewer exemptions. Finally, PRWORA lifted the restrictions on providing
assistance to two-parent families. Thus, in some States, many more two-
parent families could be eligible for assistance and subject to the
work requirements than under prior law.
    We strongly encourage each State to consider carefully what it must
do to get two-parent families working. In some cases, States may need
to make substantial changes to their program designs. In the first few
years of operating TANF, the participation rates are at their lowest
and caseload reduction credits may significantly reduce the minimum
required rates. We think it is important for States to capitalize on
this initial period to invest in program designs that will allow them
to achieve the higher participation rates in effect in later years. We
intend to assist States in this endeavor through technical assistance
and by sharing promising models as they emerge.
    We received only one comment relating to this section.
    Comment: A commenter urged us to eliminate the two-parent
participation rate once the two-parent caseload represents less than
five percent of a State's overall caseload.
    Response: We do not have the authority to eliminate the
participation requirement related to the two-parent caseload. The
statue is very clear about the required minimum rates that States must
achieve and the penalty associated with failing to meet participation
rates. We have tried to give States some relief with respect to the
demanding two-parent participation rate through both the structure of
the caseload reduction credit and the penalty reduction provisions.
Please refer to subparts D and E for further discussion of these areas.

Section 261.24--How Will We Determine a State's Two-Parent Work Rate?
(Sec. 271.24 of the NPRM)

    The regulations express the two-parent work participation rate in
terms very similar to those we used for the overall rate. Any family
that includes a disabled parent is not considered a two-parent family
for purposes of the participation rate. Thus, we do not include such a
family in the numerator or denominator of the two-parent rate.
    It is important to note that, in accordance with the statute, we
calculate both participation rates in terms of families, not
individuals. Whether we include the family in the numerator depends on
the actions of individuals, but an entire family either counts toward
the rate or does not. In the case of a two-parent family, whether a
family counts may depend on the actions of both parents.
    In response to issues raised by the comments, and questions raised
by States dealing with interim participation rate calculations, we have
added language to the final regulations clarifying what constitutes a
two-parent family in the two-parent participation rate calculation. We
have found that States had divergent readings of which parents to
consider in determining whether a family was a two-parent family.
Therefore, we included this provision to ensure greater consistency
across States in measuring participation among two-parent families.
    The final regulations state that, for the purposes of this
calculation, a two-parent family includes, at a minimum, all families
with two natural or adoptive parents (of the same minor child)
receiving assistance and living in the home, unless both are minors and
neither is a head-of-household.
    The preamble to the NPRM indicated that providing a noncustodial
parent with TANF services need not cause a State to consider the family
a two-parent family for the participation rate. This policy has not
changed in the final regulations and is consistent with the new
definition of a two-parent family. A State may, but is not required to,
treat a family in which a noncustodial parent receives TANF assistance
as a two-parent family.
    As in Sec. 261.22, where States direct us to, we exclude from the
participation rate calculation for a month the families that are
subject to a penalty for refusing to engage in work required in
accordance with section 407 of the Act, but have not been subject to a
penalty for more than three of the last 12 months. This is a change
from the NPRM, which only excluded them from

[[Page 17775]]

the numerator of the calculation. Please refer to the discussion at
Sec. 261.22 for an explanation of this change.
    Section 408(a)(7) of the Act limits the receipt of Federal TANF
assistance to 60 months for any family, unless the family qualifies for
a hardship exception or disregard of a month of assistance. (In our
discussion of Sec. 264.1, we explain that months of receipt are
disregarded when the assistance was received either: (1) by a minor
child who was not the head of a household or married to the head of a
household; or (2) while an adult lived in Indian country or in an
Alaska Native Village with 50 percent or greater unemployment.) We have
received inquiries concerning the effect of a time-limit exception or
disregard on the participation rates. In fact, the time limit does not
have a bearing on the calculation of the participation rate. All
families must be included in the participation rate, unless they have
been removed from the rate for one of the two work-related exemptions
(i.e., the family is subject to a penalty, but has not been sanctioned
for more than three of the last 12 months; or the parent is a single
custodial parent of a child under one year of age and the State has
opted to remove the family from the rate).
    We received many of the same comments about the calculation of the
two-parent participation rate that we received in connection with the
calculation of the overall participation rate. In particular, please
refer to the preamble for Sec. 261.22 for discussion of the comments
and our responses about excluding groups of recipients from the
participation rate and counting partial months of participation.
    As we indicated in Sec. 261.22, we have not kept in the final rules
our proposal to redefine families to include in the participation rate
any families that the State has excluded (based on its defining a
family as a child-only family) for the purpose of avoiding a penalty.
Please refer to the earlier preamble section entitled ``Child-Only
Cases'' for further discussion of this decision and the comments that
relate to it.
    Comment: One commenter asked what the definition of a two-parent
family is and whether it includes a household in which both parents are
not available for work. Another commenter stated a family's status as
two-parent or not often changes in the course of a month and that,
therefore, a family should not be considered a two-parent family in a
month in which its status change.
    Response: We believe that Congress did not intend to exclude from
the definition of a two-parent family a family with two parents
receiving assistance, neither of whom is disabled, even if they are
``not available for work'' or the family's status changed during the
month. We interpret the statute to mean that, if a State grants
assistance for both parents in a family (and neither is disabled), then
it must be considered and reported as a two-parent family. If one
parent is coming and going from the family in the month and the State
does not provide assistance for that parent, then it seems reasonable
not to consider it a two-parent family.
    Comment: Several commenters noted that the NPRM did not define the
term ``disabled parent,'' thus making it unclear which families should
be excluded from the two-parent participation rate. Some urged us to
leave the definition to States or to define it broadly to accommodate
State policy. Others specifically urged defining it to include people
who are temporarily disabled or incapacitated.
    Response: We have not defined the term ``disabled parent'' in the
final regulations so that each State may define the term as it deems
appropriate.
    Comment: Commenters urged removing from the denominator all persons
exempt from work requirements based on valid State welfare reform
waivers in effect prior to enactment of PRWORA.
    Response: Please refer to subpart C of part 260 for discussion of
how we will treat welfare reform waivers under the participation rates.
    Comment: One commenter thought that we should remove families that
are subject to a penalty from the calculation for the entire duration
of a penalty rather than only if they have been in penalty status for
less than three of the last 12 months. Alternatively, the commenter
thought we should remove such a family if it has been subject to a
penalty for less than three months in a fiscal year instead of the
preceding 12 months.
    Response: We do not have the authority to make either of the
changes the commenter suggested because the statute is very precise
about this provision. It specifies that sanctioned families are removed
from the rate, but not if the family has been subject to the penalty
for more than three months within the preceding 12-month period.

Section 261.25--Does a State Include Tribal Families in Calculating
These Rates? (Sec. 271.25 of the NPRM)

    States have the option of including in the participation rates
families in the State that are receiving assistance under an approved
Tribal family assistance plan or under a Tribal work program. If the
State opts to include such families, they must be included in the
denominator as well as the numerator.
    Comment: A commenter urged that any rewards or bonuses a State
receives due to including Tribal participants in the calculations
should be shared with the Tribes in question.
    Response: Nothing in these regulations precludes a State from
sharing rewards or bonuses with Tribes; however, we do not have the
authority to require a State to do so.
    Comment: One commenter was confused by our discussion in the
preamble to the NPRM. We said that where the State opts to include
families receiving assistance under a Tribal TANF or Tribal NEW
program, the families must be in the denominator as well as the
numerator ``where appropriate.'' The commenter asked us to clarify
whether a State is free to include or exclude such families from the
numerator and denominator. The commenter also asked us to clarify the
standards of participation and activities that applied for a State to
count such a Tribal family.
    Response: A State may, at its option, include or exclude families
receiving assistance under a Tribal TANF or Tribal NEW program from the
denominator of the State TANF participation rates. To be included in a
State participation rate numerator for a month, a family must meet the
standards for counting a family in that rate, both with respect to
hours of participation and allowable activities. These standards apply
whether the family receives assistance under a State TANF program, a
Tribal TANF program, or a Tribal NEW program. We realize that many
Tribal programs will have different standards of work and different
activities, but to count toward a State rate, the family must meet the
standards associated with that rate.
    We wanted to be clear that, if a State did plan to count a family
receiving assistance in a Tribal program, that family had to be
included in just the same way that a State TANF family would be
included, that is, in the denominator of the rate as well as the
numerator. But since inclusion in the numerator is not automatic
(because the family must meet the hours of participation in allowable
activities), we added the phrase ``where appropriate.'' Since this was
confusing, we modified the preamble discussion in the final rules.

[[Page 17776]]

Subpart C--What Are the Work Activities and How Do They Count?

Section 261.30--What Are the Work Activities? (Sec. 271.30 of the NPRM)

    Section 407(d) of the Act specifies the 12 work, training, and
education activities in which individuals may participate in order to
be ``engaged in work'' for the purpose of counting toward the work
participation rate requirements. Congress did not define these
activities further. While some have commonly understood meanings from
their use over time or from prior employment and training programs,
several of the activities, such as ``vocational educational training''
and ``job readiness assistance,'' are subject to interpretation.
    In considering whether to provide greater definition of the
activities as part of the NPRM, we examined legislative intent and
sought the views of a variety of groups on the matter. Most groups
urged us to leave further definition to the States. Some urged us to
define work activities in ways that fostered education while promoting
work, emphasizing the importance of education and training in
empowering many recipients to find meaningful employment, let alone to
advance. Ultimately, we chose not to define the individual work
activities in the NPRM in favor of giving States greater flexibility;
we have not changed that position in the final regulations.
    Because this flexibility could also be used in ways that do not
further Congressional intent, under the data collection requirements at
Sec. 265.9, we are requiring each State to provide us with its
definitions of work activities for its TANF program and with a
description of work activities for any separate State program that
requires them. We are concerned that different TANF definitions could
affect the vulnerability of States to penalties for failure to meet the
participation rate. This data collection will help us determine whether
this is in fact a serious problem; to the extent possible, we want to
ensure an equitable and level playing field for the States. Over the
next several years, we will carefully assess the types of programs and
activities States develop and will share the results of our findings.
If necessary at some time in the future, we will initiate further
regulatory action.
    We would also like to remind States about some key research
findings from prior welfare-to-work programs. According to the Manpower
Demonstration Research Corporation's publication, Work First, the most
successful work first programs have shared some characteristics: a
mixed strategy including job search, education and training, and other
activities and services; an emphasis on employment in all activities; a
strong, consistent message; a commitment of adequate resources to serve
the full mandatory population; enforcement of participation
requirements; and a cost-conscious management style.
    While the most successful programs consistently and strongly
emphasize work, the actual program designs recognize and address the
critical role education plays in preparing adults for work. As more and
more recipients engage in work, State caseloads may reflect higher
proportions of the educationally disadvantaged. In combination with
other work activities, education may become more important in improving
basic communication, analytical, and work-readiness skills of
recipients. Thus, States may need to integrate adult basic skills,
secondary education, and language training with high-quality,
vocational education programs. Such program designs encourage
recipients to continue acquiring educational skills necessary for
higher-skill, higher-wage jobs.
    We encourage States to adopt program designs that take advantage of
existing educational opportunities. States may use the statutory
flexibility to design programs that promote educational principles by:
    <bullet> Actively encouraging adults and children to finish high
school or its equivalent;
    <bullet> Expecting family members to attain basic levels of
literacy and to supplement their education in order to enhance
employment opportunities;
    <bullet> Encouraging family literacy; and
    <bullet> Promoting community-based work-related vocational
education classes, created in collaboration with employers.
    States could also make it easier for individuals to combine school
and work. For example, they could develop on-campus community work
experience program positions, where child care is also available. They
could also encourage schools to use work-study funds for students on
welfare and then count the hours worked in those programs toward work
requirements.
    While we have not regulated the definition of work activities, we
want to ensure that recipients and children both experience positive
outcomes. This is a particularly significant issue when child care is
the work activity. For this to happen, child care arrangements should
be well developed, implemented and supported.
    Research has found that quality child care is critical to the
healthy development of children and that providers who choose to care
for children create more nurturing environments than those who feel
they have no choice and are providing care only out of necessity. Thus,
States should assess whether recipients have an interest in providing
child care before assigning them to this activity.
    In addition, States should provide training, supervision and other
supports to enhance caregiving skills if they wish recipients to attain
self-sufficiency. Such supports, including training in health and
safety (e.g., first aid and CPR), nutrition, and child development,
would assist the development of both the caregivers and the children in
care.
    Finally, the stability of child care arrangements affects outcomes
for both parents and the children in care. When parents feel
comfortable with their child care arrangements, their own participation
in the work force becomes more stable. This stability, in turn, fosters
emotional security for children. Thus, States should take stability
into account when assigning participants to child care as a work
activity.
    The majority of those who commented on this section of the proposed
regulations supported our decision not to define work activities beyond
the statutory list. We discuss other comments below.
    Comment: Several commenters made suggestions about the content of
work activities. Some urged us generally to ensure the quality of work
activities by establishing minimum standards for the activity and for
the provider. Others made suggestions about specific activities, urging
us to give guidance or make requirements concerning particular elements
of an activity. For example, one commenter thought that vocational
educational training should conform to the definition of vocational
education in the Carl D. Perkins Vocational Education Act; another
suggested discontinuing on-the-job contracts with employers that do not
provide long-term employment. In essence, these commenters wanted us to
define certain of the activities or ensure that certain activities
would be counted as work.
    Response: We appreciate and share the commenters' concerns that
work activities be designed to meet the needs of recipients and be
effective in helping them become self-sufficient. However, we think
that the goals and objectives of the legislation will be better served
by having each State define the work activities. We believe States will
use the flexibility of the statute to formulate a variety of reasonable
interpretations leading to greater innovation,

[[Page 17777]]

experimentation, and success in helping families become self-sufficient
quickly. It is true that States could conceivably include a range of
activities that may not enhance work skills or might not be considered
valid work experience by potential employers. However, in light of the
five-year time limit and the criteria for the high performance bonus,
we expect that States will work to establish programs that promote a
family's long-term success in the workplace.
    Comment: One commenter pointed out that we omitted the statute's
limitation of the activity of ``work experience'' to instances where
sufficient private-sector employment is not available.
    Response: We have amended the regulations to reflect the statute's
limitation.
    Comment: A couple of commenters urged us to require that work
activities comport with Federal employment laws in order to count for
participation. One suggested that we require employers to post
appropriate nondiscrimination notices. Another stated that Congress
failed to include any provision limiting work activities to work at the
minimum wage.
    Response: We agree fully with the commenters that all TANF work
activities should be lawful and should not subject participants to
discrimination, unsafe working conditions or other circumstances
prohibited by employment law. Nevertheless, we do not think the
appropriate way to address the issue is to exclude certain work
activities from the participation rate calculation in some States.
Adjusting the rates would be administratively cumbersome and not
necessarily equitable. As we have discussed earlier in the section of
the preamble entitled ``Recipient and Workplace Protections,'' there
are other entities, such as the EEOC, DOL and our Office of Civil
Rights, that enforce compliance with civil rights and employment laws.
Their mechanisms for monitoring and enforcing compliance are not linked
to the timing of the participation rate calculation. In other words, a
finding of noncompliance that they might issue would not necessarily be
available within our timeframes for calculating participation rates.
Moreover, even if we did receive timely information about noncompliance
with employment requirements, because the participation rates may be
based on sample data, it might be very difficult to determine the
appropriate adjustments to make to the rates based on such findings.
    Given these complications, we have not modified the regulation as
the commenters suggest. We think it makes better sense to support the
enforcing entities in carrying out their responsibilities. If, over
time, we find significant problems that could warrant adjustments
within the TANF program, we will consult with States, labor interests,
Congress, and other interested parties about the appropriate steps to
address these problems.
    We have also included a new regulatory section at Sec. 260.35 that
addresses employment protections available to TANF recipients. Please
refer to the preamble section entitled ``Recipient and Workplace
Protections'' for a discussion of additional comments related to this
issue.
    We have addressed the Fair Labor Standards Act (FLSA) at
Sec. 260.35 of the final rule, in the preamble to Sec. 261.16, and in
the preamble discussion entitled ``Recipient and Workplace
Protections.'' Please refer to those preamble sections for further
discussion of the application of FLSA, including the minimum wage
requirement, to TANF work activities.
    Comment: We received support from a number of commenters for the
discussion of the importance of education to TANF recipients that we
included in the preamble to the NPRM. In response, we have repeated
much of that discussion in this preamble to the final rule. We also
received support for our guidance concerning the provision of child
care as a work activity.
    A couple of commenters urged us to incorporate that discussion,
particularly our program design suggestions, into the text of the
regulation itself. They argued that States would not create broader
activities that combine work and education unless we specifically
regulated in this area. Similarly, one commenter urged us to specify
that we would not penalize a State for including a range of educational
activities, from literacy programs through post-secondary education, in
its definitions of work activities.
    One commenter thought the proposed rule did not truly support the
integration of work and education because it allowed each State to
define the work activities. Instead, the commenter urged us to provide
definitions that guide States in integrating work and education.
    Response: We have not included our discussion of program designs
within the text of the regulation. That discussion is intended to spark
creative thinking about State choices in implementing TANF, rather than
to prescribe a particular design for all States. It is also intended to
underscore the important role we think education can play in TANF.
    We have not included the kind of blanket statement the commenter
suggests absolving States of any potential penalty liability for
including a wide array of educational components in its work
definitions. There are statutory limits on counting educational
activities in the participation rates; a State that exceeded those
limits could be subject to a penalty. Readers should refer to
Sec. 261.33 for a discussion of the limits on counting participation in
educational activities in the participation rates.
    As we indicated above, we have opted not to define the work
activities to a greater degree than the statute does. We think that the
preamble discussion gives States ample suggestions of ways to integrate
education with work activities. We are also available to work with any
State and its education community to help them design programs that
will meet their particular needs.
    Comment: One commenter stated that there is no provision to count
participants in a GED program, adult basic education or English as a
Second Language (ESL) in the participation rates and urged making them
countable work activities. Another commenter urged adding a basic
skills ``refresher'' course for those already holding a GED. A third
commenter encouraged us to include student internships as a work
activity.
    Response: While we have no authority to add to the list of 12 work
activities, a State could provide the education programs described in
the first two comments under the existing activities. In particular, we
point out that GED is explicitly part of the eleventh activity:
``satisfactory attendance at secondary school or in a course of study
leading to a certificate of general equivalence for a recipient who has
not received a high school diploma or a certificate of high school
equivalency.'' Similarly, student internships, depending on their
content, may well meet a State's definition of one or another of its
work activities. We would be glad to provide technical assistance to a
State that has questions about incorporating activities such as these
into its program design.
    Comment: One commenter urged us to require States to include
vocational educational training among their work activities.
    Response: We understand the commenter's interest in seeing that
recipients have the opportunity to enroll in training programs that
will give them the skills to qualify for and keep higher paying jobs;
however, we do not have

[[Page 17778]]

the authority to require a State to provide any specific work
component.
    Comment: One commenter expressed support for the development of
micro-enterprises and other forms of self-employment, particularly in
rural settings. The commenter urged increasing flexibility in this area
by counting the period necessary to develop a business as
participation.
    Response: Again, the State has the flexibility to design and define
work activities that meet the needs of its caseload, including creating
a micro-enterprise development program. It is unclear from the comment
precisely what activities the commenter believes should be considered
work that are excluded either by statute or by a State's policies. We
agree that any legitimate hours of work in the development of a
business could contribute to the participation rate; however, for
example, if the recipient is waiting for a loan approval, but not
otherwise participating, it hardly seems reasonable to count that time
as participation. The fact that something has value or is integral to a
countable activity does not necessarily mean it can count as
participation. We would be happy to work with States that would like
technical assistance in this area.
    Comment: One commenter expressed support for the requirement that a
State provide us with its work definitions, citing its value for
research into effective employment-related services. Another commenter
objected to this requirement, maintaining that States must already
submit this information as part of the State TANF plans.
    Response: We think it is important to know how States are defining
the activities because of the implications they have for penalties. We
want to ensure that we enforce the requirements of TANF in a way that
is as equitable to States as possible. We also agree that the
definitions will help with research into effective program designs. We
think it is reasonable to collect these definitions as an annual
addendum to other data collection. Unfortunately, the TANF State plans
do not necessarily include a State's work activity definitions.
However, we have revised the reporting requirements at Sec. 265.9(d) to
allow a State that included such information in its plan to reference
the plan or attach the appropriate plan pages.
    Comment: One commenter objected to ``such a restricted list of
countable work activities'' protesting that low-grant States will not
be able to make use of several of the components because recipients
will have too much income to continue receiving assistance. The
commenter also stated that low-grant States will be adversely affected
by the minimum wage requirements of the Fair Labor Standards Act
(FLSA).
    Response: States must weigh carefully their decisions about grant
amounts and earnings disregards as they formulate State policy. The
commenter is correct that a State's benefit rules may have implications
for its participation rates, as well as for a family's time limit and
for State budgets. However, these are largely matters of State
discretion; the regulations reflect the statutory work activities.
Readers should refer to the preamble at Sec. 261.16 for a more detailed
discussion of the FLSA and its effect on TANF work activities.

Section 261.31--How Many Hours Must an Individual Participate To Count
in the Numerator of the Overall Rate? (Sec. 271.31 of the NPRM)

    Section 407(c) of the Act specifies the minimum hours an individual
must participate to count in the State's overall participation rate
calculation. There are two related requirements. First, there is a
minimum average number of hours per week for which a recipient must be
engaged in work activities. The average weekly hours are reflected in
the following table:

------------------------------------------------------------------------
                                                               Then the
                                                               minimum
                   If the fiscal year is:                      average
                                                              hours per
                                                               week is:
------------------------------------------------------------------------
1997.......................................................           20
1998.......................................................           20
1999.......................................................           25
2000 or thereafter.........................................           30
------------------------------------------------------------------------

    Second, the law requires that at least an average of 20 hours per
week of the minimum average must be attributable to certain specific
activities. These activities are:
    <bullet> Unsubsidized employment;
    <bullet> Subsidized private-sector employment;
    <bullet> Subsidized public-sector employment;
    <bullet> Work experience;
    <bullet> On-the-job training;
    <bullet> Job search and job readiness assistance for no more than
four consecutive weeks and up to six weeks total in a year;
    <bullet> Community service programs;
    <bullet> Vocational educational training not to exceed 12 months;
    <bullet> Provision of child care services to an individual who is
participating in a community service program.

(Note: the limitation that at least 20 hours come from certain
activities does not apply to teen heads of households; however,
there are other limitations related to teen heads of households.
Please refer to Sec. 261.33 below.)

    After an individual meets the basic level of participation, the
following activities may count toward the total work requirement hours
of work:
    <bullet> Job skills training directly related to employment;
    <bullet> Education directly related to employment for those without
a high school diploma or equivalent;
    <bullet> Satisfactory attendance at a secondary school or GED
course for those without a high school diploma or equivalent.
    In our consultations prior to drafting the NPRM, several people
asked whether a State may average the hours of participation of
different recipients to reach the minimum average hours required by the
work participation rate, as they could in the JOBS program. PRWORA does
not permit combining and averaging the hours of work of different
individuals. However, the regulation and the statute permit averaging
an individual's weekly work hours over the month to reach the minimum
average number of hours per week required for that individual to be
engaged in work.
    We have reorganized the regulatory text slightly from the way it
appeared in the NPRM for the sake of clarity, but this section still
paraphrases the statute in simple, understandable terms.
    The final regulations do not contain the chart we included in the
NPRM depicting which work activities count in the first 20 hours and
which count thereafter. We decided the chart no longer added to
readers' understanding of the provision since legislative changes
simplified the rules and its inclusion disrupted the regulatory text,
making the policy more difficult to follow.
    Comment: We received several comments expressing support for our
clarification in the NPRM that a State may average an individual's
weekly hours of work over a month. One commenter supported averaging,
but without reference to an individual's hours in a month.
    Response: For clarity, we would like to reiterate that the statute
does not permit combining and averaging of hours of work of different
individuals in the overall participation rate. Rather, it is an
individual's hours of work from different weeks within a month that may
be averaged.
    Comment: Several commenters expressed concern about the effects of
the FLSA in restricting the number of hours a State may require an
individual to participate in certain work activities,

[[Page 17779]]

particularly work experience and community service. They emphasized the
importance of these activities to individuals not ready for
unsubsidized employment.
    Concerned that the FLSA will impede a State's ability to meet the
participation rate requirements, some commenters urged us to exempt
these activities from the wage and hour requirements of the FLSA.
    Response: We have no authority to exempt an activity from the
requirements of the FLSA. We have tried to explain the basic effect of
the requirements on TANF work activities in the preamble to
Sec. 261.16, but we urge interested parties to consult the Department
of Labor's guidance entitled ``How Workplace Laws Apply to Welfare
Recipients (May 1997)'' for more information. We would also like to
point out that States have the option of increasing the amount of a
family's grant and thus permitting an individual to engage in more
hours of work in accordance with the FLSA. States should weigh policy
decisions in this area very carefully; the interrelated effects on
participation rates, a family's remaining months under the Federal time
limit, and State spending on the TANF program are crucial aspects of
TANF program design.
    Comment: A couple of commenters expressed opposition to separating
the work activities into those that count for the first 20 hours and
those that count thereafter, questioning the regulation's support for
educational attainment despite the preamble's discussion of its
importance.
    Response: The requirement that the three education-based activities
can only count for participation after the first 20 hours is a
statutory one; thus, we have no authority to alter it. That fact does
not change our commitment to education for recipients who need it. We
have suggested several possible models for combining education with
other activities and stand ready to help States that would like
technical assistance in this area.
    Comment: A commenter urged us to make GED preparation and English
as a Second Language (ESL) ``stand alone, countable'' activities
because substantial portions of some State caseloads need basic
education and language skills before they can hold even entry-level
jobs.
    Response: Clearly, both GED preparation and ESL fit within the list
of 12 work activities enumerated in the statute.
    We presume that the commenter's real concern is that the State
cannot receive full participation credit for such educational
activities because of the requirement that the first 20 hours of
participation be attributable to the noneducational activities. This is
a statutory requirement that we have no authority to change in the
regulations. We urge States and localities to consider combining work
and educational activities where it is appropriate in order to maximize
participation credit. Although some individuals will not be able to
engage in multiple activities, this could be a viable solution for many
recipients.
    Comment: One commenter, in stressing the importance of education to
permanent self-sufficiency, urged us to include time spent on homework
and fieldwork when calculating an individual's hours of participation.
    Response: As we have indicated, it is each State's responsibility
to define its work activities in a reasonable manner; thus a State
could choose to include homework time as part of an activity. However,
we encourage States to consider carefully how Congress intended to
treat homework in determining ``engaging in work'' to ensure that its
interpretation is reasonable.
    It is unclear to us exactly what the commenter means by
``fieldwork''; if this refers to practical, career-based experience
within the context of an educational activity, it might meet a State
work activity definition. We have spoken to this issue in response to a
comment about student internships above in the comments to Sec. 261.30.
    Comment: A commenter urged us to give partial credit for placing
individuals in countable activities for fewer than the minimum average
number of hours. For example, if the required hours are 20 and the
individual participates for 10 hours per week, the commenter would have
us count the case as 0.5 in the participation rate for that month.
Another urged us to develop a means of giving a State credit for an
individual's participation over a longer period of time than one month.
    Response: The statute does not provide for counting a portion of a
case in the participation rate and measures participation on a monthly
basis; either the adult is engaged in work and the family counts in the
rate or it does not and is not in the rate.
    Comment: One commenter urged us to modify this section to include
the provisions of Sec. 261.35 or to include a cross-reference to it.
That section indicates that we will count a single custodial parent
caring for a child under the age of six as engaged in work if the
parent participates in work activities for an average of at least 20
hours per week.
    Response: The language of this section is consistent with the
statute and does not need to incorporate the provisions of Sec. 261.35
to take into account the full range of ways in which a family may meet
the participation rate. We have tried to make these regulations easy to
read. This means, in part, keeping sections reasonably short and
separating different ideas into new sections. In this subpart in
particular, we have tried to group all the provisions that relate to
counting hours of work; it would be simply impractical to include all
these provisions in one section.
    We have decided not to reference Sec. 261.35 to avoid multiple
references to the other sections in subpart C, which we think readers
will readily notice due to their proximity.
    Comment: A couple of commenters urged us to give a State credit for
an individual's ``excused'' absences from work, such as holidays or
jury duty, as opposed to counting only actual hours of work. They
thought that an absence beyond the individual's control should count as
participation. Another commenter suggested that we count at least a
portion of an individual's commute time when he or she must travel an
extended distance to reach the job.
    Response: The statute specifies the standard by which we must
measure whether an individual is engaged in work. That standard is that
a recipient ``is participating in work activities for at least the
minimum average number of hours per week'' specified in the table in
this section. Although the JOBS program gave us the discretion to
establish a participation standard that considered scheduled hours and
actual hours worked, TANF does not provide that flexibility.
    However, consistent with ordinary practice for counting work time,
a State could base the hours of work it reports on an employer's record
of hours for which an employee is paid, thus accounting for paid
holidays and jury duty days. Similarly, consistent with the ordinary
practice for counting work time, we do not believe that commuting time
can reasonably be considered ``engaging in work'' for any activity and
therefore will not count it toward the participation rates.

Section 261.32--How Many Hours Must an Individual Participate To Count
in the Numerator of the Two-Parent Rate? (Sec. 271.32 of the NPRM)

    For two-parent families, section 407(c) of the Act specifies that
the parents must be participating in work activities for a total of at
least 35 hours per week and that a specified number of

[[Page 17780]]

hours be attributable to specific work activities. A State may have one
parent participate for all 35 hours, or both parents may share in the
work activities. If the family receives federally-funded child care
assistance and an adult in the family is not disabled or caring for a
severely disabled child, then the parents must be participating for a
total of at least 55 hours per week. As before, a specified number of
hours must be attributable to certain activities (listed below).
    In the first situation (where the weekly total must be at least 35
hours), at least 30 hours must be attributable to the same narrow group
of activities that applies to the 20-hour standard in the overall rate.
In the second situation (where the weekly total must be at least 55
hours), 50 hours must be attributable to this narrow group of
activities. Again, these are:
    <bullet> Unsubsidized employment;
    <bullet> Subsidized private sector employment;
    <bullet> Subsidized public sector employment;
    <bullet> Work experience;
    <bullet> On-the-job training;
    <bullet> Job search and job readiness assistance for no more than
four consecutive weeks and up to six weeks total in a year;
    <bullet> Community service programs;
    <bullet> Vocational educational training (for not more than 12
months);
    <bullet> Provision of child care services to an individual who is
participating in a community service program.
    Therefore, no more than five of the relevant minimum hours may be
attributable to education related to employment, high school (or
equivalent), or job skills training activities.
    During our consultations prior to developing the NPRM, many thought
it was unclear whether the 35-hour requirement was a minimum for each
week or a minimum weekly average, as is the case in the overall rate.
For example, if a parent participated 40 hours one week and 30 hours
the next, the question arose whether he or she would meet the minimum
requirement for both weeks. To provide maximum flexibility for States
to meet the program goals, we clarified in the proposed rule and have
maintained in the final regulations that, as long as the parents'
average total hours equal at least 35 hours per week, the individual
meets the participation requirement.
    Other than this clarification, we have mirrored the statute. As in
Sec. 261.31, we have reorganized the regulatory text slightly from the
way it appeared in the NPRM to make it clearer, but this section still
paraphrases the statute in simple, understandable terms.
    The majority of the comments on this section expressed support for
our interpretation that the weekly hours requirement was a weekly
average within a month and not a fixed number of hours for each week.
Commenters emphasized that this will help States work flexibly with
families and respond to emergencies or other family needs that affect
hours of work in a particular week.
    We also received many of the same comments in this section that we
received in connection with the hours of work required for the overall
participation rate. In particular, please refer to the preamble for
Sec. 261.31 for discussion of the comments and our responses about: the
requirements of the FLSA; counting ``excused'' absences from work
toward the participation rate; giving partial participation credit for
participating below the hours of work standard; and reporting
requirements for a week that spans two months.
    Comment: A commenter noted that this section refers to ``an
individual'' counting as engaged in work and urged us to substitute the
word ``family'' instead.
    Response: We recognize that both parents may actually be
participating and contributing to the total number of hours required to
be engaged in work, 35 or 55 hours depending on whether they receive
federally-funded child care. We used the word ``individual'' because
the statute, at section 407(c)(1)(B), uses that term. While this is not
necessarily strictly accurate, it is no more accurate to describe the
``family'' as working; a family is counted in the participation rate,
but it is one or two individuals who engage in work. We thought that
relying on the language of the statute would be less confusing in this
case.
    Comment: One commenter advised us to modify this section to
indicate that a family with a disabled parent should not be considered
a two-parent family for the purposes of the participation rate
calculations in accordance with the statute.
    Response: The commenter is correct that the statute excludes
families with a disabled parent from the two-parent participation rate
calculation. We included this provision in subpart B where we describe
the calculations for the participation rates. Please refer to
Sec. 261.24(d).

Section 261.33--What Are the Special Requirements Concerning
Educational Activities in Determining Monthly Participation Rates?
(Sec. 271.33 of the NPRM)

    Section 407(c)(2)(C) of the Act provides that a teen who is married
or the single head-of-household is deemed to be engaged in work for a
month if he or she maintains satisfactory attendance at a secondary
school or the equivalent or participates in education directly related
to employment for an average of at least 20 hours per week. Paragraph
(b) of this section paraphrases the language of this statutory
provision.
    To reinforce the emphasis on work, section 407 of the Act limits
educational activities in two ways:
    (1) An individual's participation in vocational educational
training may count for participation rate purposes for a maximum of 12
months; and
    (2) For each participation rate, not more than 30 percent of
individuals determined to be engaged in work for a month may count by
reason of participation in vocational educational training. In fiscal
year 2000 and thereafter, this 30-percent limit also includes the teens
deemed to be engaged in work by reason of maintaining satisfactory
attendance at secondary school (or the equivalent) or participating in
education directly related to employment, whom we described above.
    When PRWORA was enacted, there was substantial controversy about
precisely how the second limitation would apply. However, Pub. L. 105-
33 modified this provision, making the limitation much clearer. The
description above and the regulation at Sec. 261.33 reflect the new
provision, as amended by Pub. L. 105-33.
    Based on some of the comments we received, we have made some minor
modifications to the regulatory language as it appeared in the NPRM.
The proposed regulatory language inadvertently suggested that only
married heads-of-households, as opposed to any married teen, could be
deemed to be engaged in work by virtue of this provision. In addition
to correcting that error, we have modified the wording of the 30-
percent cap to reflect the statute more closely.
    We also want to explain the technical details of how we will
interpret the provision relating to counting teens in educational
activities for the purposes of calculating the participation rates. We
are interpreting the deeming of teens as engaged in work based on
satisfactory attendance in secondary school (or the equivalent) or 20
hours per week of

[[Page 17781]]

education directly related to employment to apply to both participation
rates. While the provision might appear at first glance to apply to the
overall rate alone, after considering Congressional intent and the
legislative history, we think it is appropriate to apply it in the two-
parent rate as well.
    Because the two-parent rate, as amended by Pub. L. 105-33, permits
the hours of the two parents to be combined to achieve the required
weekly average, we needed to determine how many, if any, additional
hours the parents would need to work in order to count in the two-
parent rate when one parent was maintaining satisfactory attendance in
high school or the equivalent. It seemed unreasonable and contrary to
the spirit of the law to count the family without any additional hours;
for example, that would allow a two-parent family to count based solely
on the attendance of one parent in a GED class. Such a policy would
support neither the educational welfare of the other parent nor the
economic self-sufficiency of the family, faced with time-limited
benefits.
    To address these concerns, our rules incorporate the following
policy for two-parent families: (1) we will consider satisfactory
attendance at secondary school or the equivalent of a single head-of-
household or married recipient under the age of 20 to equate to 20
hours per week of participation; thus, the parents would need a
combination of 15 or 35 additional average hours per week (depending on
which standard of hours applied to them) to count for the two-parent
participation rate; and (2) if both parents in the family are under 20
years of age, we will consider them to be engaged in work if both meet
the conditions of Sec. 261.33(b), that is, if both are either
satisfactorily attending school or equivalent or participating in
education directly related to employment for at least 20 hours per
week. Our rationale for equating satisfactory attendance in secondary
school with 20 hours of participation is that the statute makes the
presumption that such attendance is equivalent to 20 hours in education
directly related to employment.
    Comment: One commenter, while acknowledging the statutory origin of
the 30-percent cap, nevertheless objected to the provision as it
relates to teens in secondary education. The commenter stated that a
mandated activity cannot have a cap.
    Response: The commenter is correct that the 30-percent limitation
is required by the statute; however, we would like to address the
question of secondary education as a mandated activity. The commenter
is referring to section 408(a)(4) of the Act, which prohibits a State
from using TANF funds to assist a single parent under the age of 18 who
has not completed high school (or equivalent) unless he or she attends
high school (or equivalent) or a State-approved alternative education
or training program. Both provisions underscore the importance of basic
education for teens but are distinct in their effects within TANF. Even
if the teen populations and the activities described were identical,
which they are not, the central difference between the two provisions
is that one mandates what the teens must do and the other restricts
what a State receives credit for in the participation rate.
    Comment: We received several comments urging us to count post-
secondary education toward the participation rate and recommending that
the regulations explicitly indicate that it is a TANF work activity.
    Response: As we have indicated above, we do not have the authority
to create additional work activities beyond the 12 statutory
activities. Nevertheless, depending on whether and how the State chose
to incorporate it into its TANF structure, post-secondary education
could fit within the definition of 1 or more of the 12 activities. The
appropriateness of categorizing it as one activity versus another would
depend on the nature of the post-secondary program, such as whether it
were vocational training.
    We would also like to emphasize that States have the flexibility to
design programs that allow recipients to combine school and work. We
have suggested some possible models for this in the preamble to
Sec. 261.30 and are ready to work with States that want help in
pursuing such program designs.
    Comment: A couple of commenters objected to the limitations on
vocational educational training, both an individual's limit to 12
months and the 30-percent cap. They stressed that States should be free
to design vocational programs that are effective in moving participants
into permanent employment, which may require more than one year of
training.
    Response: States are free to design and operate vocational programs
that take longer than one year to complete; the limitation is strictly
about the period of time for which a State could receive credit for a
recipient's participation in that program. The limitation, while
potentially discouraging States from designing certain long-term
programs, is statutory and beyond our authority to modify. Again, we
would like to point out that combining vocational training with
practical experience that could count as another activity may be a
viable approach in many cases. Moreover, States should consider that
effective vocational education programs, or other programs that succeed
in moving recipients from welfare to work, will contribute to the
likelihood that the State will qualify for a high performance bonus or
caseload reduction credit. We are awarding the high performance bonuses
based on several criteria, including the number of new hires, increases
in earnings and job retention. Thus, in spite of the limits on counting
vocational educational training as participation, there are other
incentives to designing effective vocational education programs.
    Comment: A commenter noted that the proposed regulations and
preamble did not indicate that a State has discretion to determine how
to measure the 12-month limit on counting an individual's participation
in vocational educational training.
    The commenter also urged us to amend the regulations to indicate
that a State that combines education with other activities would
necessarily be able to count these activities in the participation
rates.
    Response: States have limited flexibility in this area. If a family
is included in the numerator of a participation rate for a month by
virtue of participation in vocational educational training, then that
month counts against the 12-month limit for that individual.
    If a State reports hours of participation in an activity that meet
the requirements of this subpart, then the hours would count in the
participation rates, and the month would count as a month or
participation in the activity, regardless of whether the individual
performed them in combination or separately.
    We have stressed the possibility of combining education and work
activities in part because of the statutory limits, reflected in these
regulations, on how educational activities may count for participation
purposes. In addition, we believe that encouraging recipients to
acquire new and more advanced skills after they have entered the work
world will help them attain and keep higher-paying jobs, leading to
more economic security for families.

Section 261.34--Are There Any Limitations in Counting Job Search and
Job Readiness Assistance Toward the Participation Rates? (Sec. 271.34
of the NPRM)

    Section 407(c)(2)(A)(i) of the Act limits job search and job
readiness assistance in several ways.

[[Page 17782]]

    First, an individual generally may not count as engaged in work by
virtue of participation in job search and job readiness assistance for
more than six weeks. No more than four of these weeks may be
consecutive. During our consultations prior to drafting the NPRM, we
were asked whether these limitations applied for the lifetime of the
individual, per spell of assistance, or per fiscal year.
    Based on those consultations, after an analysis of the statute, we
decided in the NPRM to interpret it as a fiscal-year limit for two
policy reasons. First, since the participation rate itself is tied to
the fiscal year, it makes sense to apply the limitation to the same
timeframe. Second, a different policy could force States to place
individuals in other, less appropriate activities just to meet the
participation rate. Moreover, research indicates that job search
activities are an instrumental component in effective work program
designs.
    The statutory language supports the fiscal-year interpretation. The
job search language at section 407(c)(2)(A)(i) of the Act limiting the
weeks of participation states that the limit is ``notwithstanding
paragraph (1).'' Paragraph (1) refers to the determination of whether a
recipient is engaged in work for a month ``in a fiscal year.'' Thus the
reference to paragraph (1) puts the job search limitation in the
context of a calculating whether an individual is engaged in work in
the fiscal year. Based on these considerations, we clarified in the
proposed rules that the six-week limitation applies to each fiscal year
and have not changed that interpretation in the final regulations.
    The legislation and our rules allow the 6-week limit on job search
and job readiness assistance to extend to 12 weeks if the unemployment
rate of a State exceeds the national unemployment rate by at least 50
percent, or if the State could qualify as a needy State for the
Contingency Fund.
    Finally, our rules paraphrase the statute (at section
407(c)(2)(A)(ii) of the Act) in allowing a State to count three or four
days of job search and job readiness assistance during a week as a full
week of participation on one occasion for the individual.
    Comment: We received many comments in support of our interpretation
that the job search and job readiness limit applies on a fiscal-year
basis. However, one commenter thought we were too specific and should
allow States to interpret the limitation.
    Response: We have not modified the regulation as the commenter
suggests. We think it is reasonable to apply one standard to all
States. Given the overwhelming support for the fiscal-year
interpretation and the statutory and policy support we provided for it
above, the final regulations maintain that policy. This policy only
limits the maximum job search and job readiness that count for
participation purposes. States still have flexibility in determining
how much an individual should actually participate in such activities,
including the flexibility to apply the job search and job readiness
limit on a lifetime basis for an individual if they so choose.
    Comment: A commenter thought that the way in which we paraphrased
the statute's limit on job search and job readiness to not more than
four consecutive weeks was confusing and urged us to use the statutory
wording.
    Response: We have modified this section to follow the statute's
language more closely. We have left the provision limiting the number
of consecutive weeks separate in paragraph (c) because we think it is
easier to follow this way, but have changed the wording within the
paragraph in response.
    Comment: A few commenters objected to limiting job search and job
readiness to four consecutive weeks, arguing that there is no rationale
for stopping at that point or that it is simply too short a period of
time to ensure that recipients will find jobs. A couple of other
commenters objected to the six-week total limitation for essentially
the same reasons and urged us to create a longer time period.
    Response: There is no limit on the amount of job search and job
readiness a State may require of an individual. However, the statute
imposes limitations on how much the activity counts toward the
participation rate. We have no authority to extend the it counts, other
than when a State meets the criteria for counting 12 weeks of job
search and job readiness assistance instead of 6.
    Comment: Two commenters recommended that we separate job search
assistance from job readiness assistance and establish separate limits
on each activity.
    Response: In determining whether an individual is ``engaged in
work'' for the participation rates, the statute provides for 12
different work activities. One of those activities is ``job search and
job readiness assistance''; the statute does not recognize them as
separate components. As we indicated in the discussion at Sec. 261.30,
we do not have the discretion to add to those activities or to separate
job search from job readiness. If a State has two different activities
as part of its TANF program, it would have to count an individual's
participation in either one toward the limits described in this
section.
    Comment: A commenter suggested that we clarify the regulations to
allow a State to apply the extended job search and job readiness
provision to a ``needy political subdivision'' as it would if the State
were a ``needy State.''
    Response: The statute is very specific in describing the two
conditions under which 12, rather than 6, weeks of job search and job
readiness can count toward the participation rates. One of those
conditions is when the ``State'' qualifies as ``needy'' under the
Contingency fund definition. That definition applies to a State as a
whole; therefore, there is no mechanism by which to apply it to a
political subdivision.
    Comment: One commenter recommended that the regulations ensure that
a State has advance notice of whether it qualifies to count individuals
for the extended 12 weeks of job search and job readiness in a fiscal
year. The commenter argued that this would let the State plan which
activities to make available to its recipients in order to meet its
work participation rate.
    Response: As we indicated above, there is no limit on the amount of
job search and job readiness a State may require of an individual; the
limitation is on how many hours of the activity count toward the
participation rates. We hope that a State would not, as the commenter
suggests, withhold access to job search and job readiness--or any
activity, if it were the most appropriate for a recipient--and require
participation in another activity, solely for the purpose of meeting
the participation rate. The participation rates represent a requirement
on the State, not a requirement on specific individuals, and the State
can inherently meet the participation rates even if every individual is
not in a countable activity.
    Further, we have no ability to make an advance determination that a
State qualifies for a 12-week job search limit because the data are not
available in advance and the statute authorizes the 12-week limit based
on a State's current situation.
    Comment: One commenter objected to the provision permitting a State
to count three or four days of job search and job readiness assistance
as a full week of participation because the data collection system in
the commenter's State does not allow it to count hours of participation
on that basis.
    Response: This provision is not a requirement. Any State that does
not wish to count three or four days of this

[[Page 17783]]

activity as a full week of participation is not required to do so. The
origin of the provision is statutory; we presume the intent was simply
to make it easier for States to receive participation credit for this
activity.

Section 261.35--Are There Any Special Work Provisions for Single
Custodial Parents? (Sec. 271.35 of the NPRM)

    Section 407(c)(2)(B) of the Act provides a special participation
rule for single parents or caretakers with young children. A single
parent or caretaker with a child under the age of six will be deemed to
be engaged in work for a month if he or she participates in work
activities for an average of at least 20 hours per week.
    This provision has little relevance in FYs 1997 and 1998, when, for
the overall rate, the required number of hours for all individuals is
20 hours per week. But, when the required number of hours rises to 25
hours per week in FY 1999 and to 30 hours per week thereafter, this
provision allows single parents or caretakers to spend time with
younger children. It also may enable those with young children to
fulfill their work obligations while their children are in preschool
activities.
    The regulations paraphrase this statutory provision.
    There were no substantive comments on this section.

Section 261.36--Do Welfare Reform Waivers in a State Affect the
Calculation of a State's Participation Rates? (Sec. 271.36 of the NPRM)

    This section is simply a cross-reference to subpart C of part 260,
which addresses welfare reform demonstration waivers. We thought it
would be helpful to include it so that readers would know to refer to
this important exception to the work activities and hours specified in
subpart C. We have changed the reference from what it was in the NPRM,
in light of our consolidation of the regulatory provisions relating to
waivers under part 260.
    There were no comments on this section.

Subpart D--How Will We Determine Caseload Reduction Credit for Minimum
Participation Rates?

Section 261.40--Is There a Way for a State to Reduce the Work
Participation Rates? (Sec. 271.40 of the NPRM)

    To ensure that States receive credit for families that have become
self-sufficient and left the welfare rolls, Congress created a caseload
reduction credit. The credit reduces the required participation rate
that a State must meet for a fiscal year. It reflects the reduction in
the State's caseload in the prior year compared to its caseload under
the title IV-A State plan in effect in FY 1995, excluding reductions
due to Federal law or to State changes in eligibility criteria.
    This provision enhances the inherent interest of States to help
families become independent. As a State reduces its caseload, its risk
of incurring a penalty lessens because lower work participation rates
are easier to achieve. This provision also increases a State's chance
of qualifying for a lower basic MOE requirement, which would reduce its
risk of incurring an MOE penalty.
    To establish the caseload base for FY 1995, we proposed using the
number of AFDC cases and AFDC Unemployed Parents reported on ACF-3637.
To avoid artificial reductions in the minimum participation rates, the
NPRM included cases in any separate State program used to meet the
maintenance-of-effort (MOE) requirement in determining the prior-year
caseload. Under the proposed rules, we would not have granted a
caseload reduction credit unless the State reported case-record
information for its separate State programs.
    Comment: Some commenters suggested that allowing States to reduce
their work participation rates emphasizes caseload reduction over the
goal of self-sufficiency. Others strongly supported the caseload
reduction concept.
    Response: By including this provision in the statute, Congress
sought to recognize State success in moving individuals off assistance.
We believe this provision comports closely with both statutory language
and intent.
    Comment: One commenter asked us to include explicit language to the
effect that we would apply the caseload reduction credit to reduce the
participation standards before evaluating State performance.
    Response: Different groups will evaluate State performance in a
variety of ways. For purposes of determining potential penalty
liability, we will compare a State's actual participation rates with
the rates that apply following any adjustments due to caseload
reduction credits.
    Comment: A substantial number of comments addressed our proposed
method of using reported AFDC data to establish the 1995 caseload
baseline. First, several correctly pointed out that the Statistical
Report on Recipients Under Public Assistance is ACF-3637, not ACF-3697.
In addition, many commenters thought that, to conform to the statute,
the base-year calculation should include not only the AFDC population,
but also recipients of assistance under the Emergency Assistance
program (EA) funded under title IV-A and cases receiving At-Risk and
transitional child care benefits.
    Other commenters suggested that two-parent families receiving TANF
assistance are not comparable to AFDC Unemployed Parent (AFDC-UP) cases
because TANF does not restrict two-parent families as AFDC-UP
eligibility rules did. They argued that, to be fair, we ought to
compare ``apples to apples'' and ``oranges to oranges.'' Most
recommended either not counting two-parent cases at all or allowing
States to adjust the base-year caseload reports to include any two-
parent cases that were not AFDC-UP cases. They also recommended
adjusting the reports to correct inaccuracies.
    Response: We agree with these commenters that, to some extent, our
proposal compared ``apples to oranges.'' In developing the NPRM, we
recognized that the calculation should reflect an unduplicated count of
cases receiving ``assistance'' under either AFDC or EA. However, from
the data reported to us, we could not unduplicate the AFDC and EA case
counts or determine which, if any, of the EA benefits constituted
``assistance.'' In our consultations, many State staff told us that
they would also not now be able to unduplicate AFDC and EA cases for
fiscal years 1995 and 1996. Thus, for consistency, we limited the base-
year data to AFDC cases reported on the ACF-3637. Based on the comments
and further internal discussion, however, we believe it would be fairer
to afford States the opportunity to adjust and correct baseline data if
they can do so because adjustments would make the base-year and prior-
year caseload figures more comparable. For example, it would not be
appropriate to include certain EA cases in the base-year caseload
because, as recipients of ``one-time, short-term'' benefits, such cases
would not be receiving TANF ``assistance'' and do not show up in the
prior-year TANF caseload. However, if there were EA cases in 1995 that
received ``assistance'' and that did not receive both EA and AFDC
benefits, it would be appropriate to include those cases in the base-
year caseload.
    To allow for more comparable caseload data, we have modified the
final rule. We will adjust the base-year case count for any State that
can provide accurate adjustment data or unduplicated case counts, for
example, through a computer match of each month's 1995 AFDC and EA
caseload and subsequent years. This includes reliable information on
the actual

[[Page 17784]]

number of two-parent cases in its AFDC caseload for applicable years.
However, we will only include EA cases to the extent that the
assistance provided under EA would meet the TANF definition of
assistance.
    Comment: Some commenters suggested that some types of cases from FY
1995, such as State General Assistance (GA) cases, are not included in
the baseline, but should be. They argued that analogous cases are
served in separate State programs and thus will be included in the
comparison year.
    Response: We appreciate the commenter's point and agree that, in
this regard, we are not comparing like cases. However, we cannot
include GA or similar cases in the base-year because the statute
specifies that we compare cases ``that received aid under the State
plan approved under part A (as in effect on September 30, 1995) during
fiscal year 1995.'' To the extent that such cases are in the prior-year
caseload, but not in the 1995 base because the State has expanded its
eligibility criteria since 1995, the net caseload decrease calculation
will adjust for this difference. Please refer to Sec. 261.42 for
additional discussion.

Section 261.41--How Will We Determine the Caseload Reduction Credit?
(Sec. 271.41 of the NPRM)

    In the proposed rule, we explained how difficult it was to develop
an appropriate methodology to quantify the different types of caseload
reductions. We had considered and rejected two alternatives, i.e., the
use of Medicaid records to estimate the effect of eligibility changes
(since Medicaid eligibility is based on the July 1996 AFDC eligibility
rules) and a computer simulation model. Neither alternative could
produce reasonably accurate estimates of the effect of eligibility
changes on the caseload size. Nor did our extensive consultations
provide a straightforward methodology that could be universally
applied.
    As a result, the NPRM proposed a caseload reduction methodology
based on State-submitted information and estimates. These regulations
incorporate the same basic approach. Under the final rules, we
determine the appropriate caseload reduction for each State using the
following process:

    Step 1--We compare 1995 AFDC and Unemployed Parent caseload data
to State-reported TANF and SSP-MOE caseload data for the prior-year.
    Step 2--The State submits a Caseload Reduction Report that
provides: a complete listing and implementation dates of State and
Federal eligibility changes since FY 1995; a numerical estimate of
the impact on the caseload since 1995 of each eligibility change; an
overall estimate of the net cases diverted from assistance as a
result of eligibility changes; an estimate of the State's caseload
reduction credit; the number and distribution of caseload closures
and application denials, by reason; a description of the methodology
for the estimate, as well as supporting data to document the
information in the report; a certification that it incorporated all
net reductions, there was an opportunity for public comment on the
content of the report, and it considered such comments; and a
summary of all public comments. (We have included the Caseload
Reduction Report form and instructions at Appendix H.)
    Step 3--We compare and analyze each State's methodology,
estimates, and data to determine whether they are plausible. We may
request that a State submit additional information within 30 days to
support the estimates. In addition, we will conduct periodic on-site
visits and examine case records to validate the information we have
received.

    Because eligibility changes often affect two-parent cases
differently from the overall caseload and the two-parent rates are
distinct, the NPRM required States to submit separate estimates and
information for the overall and two-parent rates to receive a caseload
reduction credit.
    Comment: Many comments noted how difficult it is to measure the
impacts of policy changes and achieve comparability or equity among
States. One suggested that the only accurate way to determine the
caseload impact of a policy change is to use experimental and control
groups. A few commenters suggested using a ``quality control'' model or
system based on sampling, exception criteria, and audits to establish
the estimates of policy changes.
    A number claimed that we had shifted the statutory burden and
responsibility for calculating the caseload reduction credit from us to
the States, with mixed views as to whether this was appropriate. One
cited the specific statutory language requiring that the regulations
``shall place the burden on the Secretary to prove that such families
were diverted as a direct result of differences in such eligibility
criteria.'' Some expressed concerns about the standards to which we
might hold States seeking caseload reduction credits (i.e., in
quantifying the effects of eligibility changes).
    Another suggested that the Secretary has an obligation to pay for
obtaining such data.
    Others, while expressing concern with the proposal, agreed that it
would be difficult for us to develop a sounder methodology that could
be used in every State. Several commenters noted that the methodology
imposed a tremendous burden because States may not have retained or may
never have collected the information needed to make estimates. Some
urged working with States to find a reasonable or less burdensome
method of measuring the caseload reduction. Others suggested that,
working in partnership with States, we should provide technical
assistance to help States do the required analysis.
    Response: We are glad that commenters clearly understood the
difficult dilemma posed in developing a caseload reduction methodology,
and we are sympathetic to their concerns about the burden our proposed
methodology would impose on States. However, we believe that the
specific recommendations for methodological alternatives, such as a
quality control model, ultimately would impose an even greater
information collection burden on States, without a guarantee of more
precise estimates. Therefore, in the final rule, we are retaining the
same general approach, while adopting some suggested improvements. We
have clarified that we will accept State estimates of the impact of
eligibility changes and the resultant caseload reduction credit, unless
they appear to be implausible, based on the common experience of other
States. In these situations, we will ask the State to re-examine its
estimate in light of this new or additional information.
    At the same time, we have clarified our expectation that States
provide aggregate information on the number and distribution of case
closures, by reason. At a minimum, States must provide this information
for the base year (1995) and prior year. The NPRM asked for a listing
of reasons, but did not directly say we were looking for quantitative
information that might reveal any significant shifts in the causes of
case closures that might be associated with changes in State policies.
We also decided to ask for similar data on application denials and
added an explicit requirement that States report an overall estimate of
the net number of cases diverted due to eligibility changes.
    We understand that the caseload closure and application denial
information that we are requesting may not directly measure the
caseload effects of eligibility changes, especially over time as the
effects of changes decay and reporting practices may shift. However, it
is useful information for a State to consider in preparing its Report,
it will give the public a context for assessing and commenting on the
State's methodology and estimates, and it will give us a national set
of data that will

[[Page 17785]]

enable us to judge the plausibility of individual State determinations.
    As suggested, we have consulted--and intend to work in partnership
with--States, State groups, and advocates to develop appropriate
estimates, complete the caseload reduction analysis, and refine their
estimating methodologies.
    We thought it would be helpful for States to have all forms related
to the rule published together. As a result, we have included, under
Appendix H, the Caseload Reduction Report form and instructions for
completing it. Although the form itself was not part of the NPRM, we
addressed the burden associated with the caseload reduction estimates
in our paperwork burden estimate. Anyone wishing to comment on the form
or burden should submit comments to the Office of Management and
Budget. Please refer to the section of the preamble titled ``Paperwork
Reduction Act'' for further information.
    Regarding the suggestion that the Secretary pay for obtaining such
data, we would say: (1) Congress did not appropriate funding for this
purpose; (2) many States can draw upon analyses done for other purposes
to reduce the cost burdens associated with these determinations (For
example, in proposing changes to eligibility rules, some States will
routinely prepare estimates of caseload and budgetary impacts of those
changes as part of the State budgetary and legislative process.); (3)
if such estimates are not otherwise available, because of caseload
reductions and the strong economy, in general, States have substantial
funds available to do research and analysis; and (4) we expect the
burden of these reports to diminish over time because State program
rules should become less subject to change and States will have
developed the methodological framework for producing their estimates.
    Comment: Several commenters objected to using case closure
information as the basis for estimates, because the reason for closure
is often unknown and the coded reason is sometimes incorrect. Others
suggested that different eligibility rules affect applicants as well as
recipients and thus the reason for an application denial is just as
important as case closure information in determining the effect of a
policy on the caseload.
    Response: While we recognize the deficiencies in both case closure
information and application denial information, it is generally the
most readily available State information that we can use to help assess
the impact of policy decisions. Therefore, we have retained the
requirement of submitting case closure information in the final rule.
Also, we have added a requirement to submit similar application denial
information. In addition, we have clarified that we are looking for
quantitative information.
    While we are not requiring it, a State may conduct surveys or in-
depth reviews to establish more accurate estimates of the effect of
policy changes and use this information in place of case closure and
application denial data from case files.
    Comment: Numerous comments noted that the statute does not address
whether there should be separate caseload reduction calculations for
the overall and two-parent rates. Some thought that, if the State had
achieved an overall caseload reduction, it was unfair to penalize it
for an increase in two-parent households (especially when it
implemented these policies to help keep families together--a purpose of
the statute). Some commenters recommended applying an overall caseload
reduction credit to both rates. Others liked our proposed approach of
two separate calculations, each based on reductions in the applicable
caseload. Many suggested that States should be afforded the option to
choose whether to use one or two calculations. One commenter suggested
that, if we retained the approach of two separate calculations, we
should allow a State to request and submit estimates on only one rate,
if a reduction credit were not appropriate for the other.
    Response: Resolving this issue is critically important because of
its impact on preparing families for work and self-sufficiency, the
potential penalty liability of States, and the lack of guidance on
Congressional intent. We were persuaded by the comments that providing
an option would be an appropriate way to ensure that States that adopt
policies to promote two-parent families would not be penalized. In
particular, we thought it made sense to allow States credit for success
with its total population, since the two-parent caseload is a subset of
that total, generally a very small subset. It also allows us to give
flexibility to States to accommodate differing circumstances. At the
same time, we were concerned that allowing an option could reduce the
strong Congressional mandate for two-parent families to prepare for and
engage in work, because States that were particularly successful in
achieving overall caseload reductions could reduce their target two-
parent participation rates to minimal levels.
    To help us make this decision, we analyzed caseload data for fiscal
years 1995, 1996 and 1997, TANF participation rates for FY 1997, and
preliminary participation rates for FY 1998. We determined that
providing a State option would not nullify the two-parent participation
requirements, as we had feared. In fact, our analysis showed that more
States derive a greater caseload reduction credit from calculating two
separate credits, i.e., applying the two-parent caseload reduction to
the two-parent rate.
    Based on this analysis and to accommodate State circumstances
better, we have decided to allow States an option regarding the
caseload reduction credit for the two-parent participation rate. A
State may use the overall reduction credit for its two-parent rate or
may opt to submit separate caseload reduction information on its two-
parent caseload and base the credit for the two-parent rate on
reductions in the two-parent caseload alone. States do not have the
option of applying to the overall rate a reduction credit based on
reductions in the two-parent caseload.
    Comment: A commenter asked that the rules clarify that the State's
methodology must account for the ongoing effects of an eligibility
change beyond the initial year.
    Response: We agree with this comment. The final rule requires
estimates of the effects of all eligibility changes since FY 1995.
    Comment: Many State commenters noted that it would be better to
know their caseload reduction credits earlier in the year, but then
noted that two weeks might not be enough time to provide any additional
information requested by ACF. Several suggested that States need at
least one month to provide supplemental information. Others suggested
that we negotiate an appropriate deadline with each State, based on the
information needed.
    Response: We recognize that States may need more than two weeks to
provide additional information. Therefore, we have modified the final
rule to allow a State to negotiate the information deadline or submit
it within 30 days of the request. We believe that it is important to
resolve such matters within a short timeframe so that States will know
what participation rates they must meet as soon as possible.
    Comment: Several commenters suggested that States and the
Department would both benefit from making each State's methodology and
plan available for public review and comment. That way, other
organizations would be able to provide another perspective on
eligibility changes and their impacts on the caseload. For the

[[Page 17786]]

purpose of public review and comment as well as sharing methodologies
and approaches among States, several commenters suggested that we
electronically post each State's estimates and methodology.
    Response: We agree that both States and the Department would
benefit from public input on the estimates and methodology. Therefore,
in the final rule, we have required a State to certify that it has
provided the public an appropriate opportunity to comment on the
estimates, methodology, and reductions. To allow time for public input,
we have extended the due date of the Caseload Reduction Report until
December 31 of each year. We also require a summary of the public
input. To enable us to learn effective estimating techniques from each
other, we intend to post electronically useful illustrative estimates,
techniques, and comments on the ACF World Wide Web page at http://
www.acf.hhs.gov.
    Comment: Several commenters objected to the requirement that the
Governor certify the caseload reduction figures, e.g., ``the not-too-
subtle implication that States will not be truthful is both offensive
and unnecessary.'' One suggested that surely the Governor's designee
should be able to ``certify'' that the State had taken into
consideration all reductions.
    Response: We agree with the suggestion and have made the
appropriate change in the rule.
    Comment: A number of comments suggested that we strike the
provision that requires a State to report disaggregated data on
families in separate State programs in order to qualify for a caseload
reduction. Some maintained that requirements for disaggregated data on
separate State program cases exceed our authority. As an alternative,
some suggested that aggregated caseload data should suffice.
    Response: If a State moves a family receiving TANF assistance to a
separate State program where it receives benefits meeting the
definition of assistance, this change in the family's status would
represent an eligibility change if we did not include separate State
program (SSP) cases in the caseload count. Therefore, unless we require
and receive the SSP information, it would be impossible to calculate
the appropriate caseload reduction credit. However, we point out that,
under the final rule, we have significantly reduced the amount of data
we are requesting on SSP cases. Most of this cutback is due to a
reduction in the number of programs and types of cases for which States
must report data. This is one of the effects of changing the definition
of assistance. We have also reduced the burden by changing some data
elements and changing the amount of data we expect on other individuals
in the family (i.e., those not receiving assistance).
    Comment: Some comments suggested reducing the data collection
burden by actually treating a transfer to a separate State program as a
change in eligibility and estimating the impact on the caseload
reduction.
    Response: We do not agree with this suggestion. Since we are
expecting States to report case-record information on separate State
programs, we believe actual caseload numbers will be available and
there is no reason to develop or accept estimates.
    Also, States and other commenters argued that, to the extent
possible, the methodology should compare ``apples'' with ``apples,''
not ``oranges.'' We believe that including the SSP cases in the prior-
year caseload best serves that objective. Because SSP cases will be
receiving benefits that address their basic needs, we expect that,
generally, they will be comparable to AFDC cases.

Section 261.42--Which Reductions Count in Determining the Caseload
Reduction Credit? (Sec. 271.42 of the NPRM)

    Congress enacted the caseload reduction provision to give States
credit toward participation for families that have achieved self-
sufficiency or left the welfare rolls due to work, marriage, child
support, or other means of support. The statute does not give caseload
reduction credit for Federal or State eligibility changes that deny
assistance to vulnerable families.
    In the NPRM we gave States full credit for caseload reductions,
except when those caseload reductions arose from changes in rules that
directly affect a family's eligibility for benefits (e.g., more
stringent income and resource limitations, time limits, grant
reductions, more restrictive residency, age, demographic or categorical
factors). States could take credit for the calculable effects of
mechanisms or procedural requirements used to enforce eligibility
criteria (such as fingerprinting or other verification techniques) only
to the extent that they identify or deter ineligible families under the
State's rules.
    We also proposed that, in order to qualify for a caseload
reduction, a State must report data on families in separate State
programs. Based on the type of family served or the nature of benefits
provided, we proposed that we would exclude some families in separate
State programs from this calculation, if a State demonstrated that the
cases would not have been included under AFDC or EA, based on specific
data on the family.
    Comment: In determining the Federal and State eligibility changes
that do not count for the caseload reduction credit, a number of
commenters recommended using a concept of net caseload change. They
suggested that eligibility changes that result in caseload reductions
should be offset by the positive policy choices of States that increase
the caseload. To illustrate: if a State-imposed time limit resulted in
the termination of 1,000 cases in a year, but the elimination of the
``100-hour rule'' and the ``attachment to the workforce'' requirements
to encourage two-parent family formation added 300 families, only 700
cases would not count toward the State's caseload reduction credit.
These commenters suggest that an alternative reading discourages States
from adopting proactive policies that are consistent with the intent of
the law, such as making work more attractive and encouraging and
supporting the formation of two-parent families.
    Response: Like commenters who offered these suggestions, we are
very supportive of policies that promote work, enhance family formation
and help make work pay. Given our desire to encourage family-supportive
policies, we found this proposal to mitigate caseload reduction
incentives appealing. We also think that the concept of a net
eligibility decrease, taking all eligibility changes into
consideration, provides an opportunity to improve the comparability of
caseloads, i.e., it would result in comparing ``apples to apples''
rather than ``apples to oranges.'' Many States have dramatically
increased their earned income disregards and resource limits and
eliminated various categorical requirements. Thus, many current
recipients would not have been eligible under the 1995 AFDC criteria.
To avoid penalizing States for such positive changes, we have adopted
the recommendation of using the net number of cases diverted from TANF
due to eligibility changes in determining the caseload reduction
credit.
    Two examples illustrate how the concept will actually work.
Consider a State in which the caseload was 100,000 in FY 1995 and fell
to 75,000 in FY 1997. The State estimates a caseload decrease of 15,000
due to time limits and other restrictive eligibility rules and a
caseload increase of 10,000 because of increased earnings disregards
and resource standards. In this example, the net caseload reduction due
to eligibility changes is 5,000. This means that, of the actual decline
of 25,000 cases, 20,000 count toward the caseload reduction

[[Page 17787]]

credit. Thus, the State's caseload reduction credit for FY 1998 is 20
percent, (because 20,000 is 20 percent of 100,000).
    To demonstrate what happens when caseload increases due to
eligibility changes exceed eligibility-related decreases, we simply
reverse the example above. The State's caseload fell 25 percent, from
100,000 to 75,000 between FYs 1995 and 1997. In this example, the
estimated decline attributable to time limits and other restrictions is
10,000. The estimated increase due to higher earnings disregards and
resource limits is 15,000. Because the net effect of eligibility
changes is a 5,000 increase in the caseload, there would be no net
number of cases diverted from TANF as a result of eligibility changes.
Since there is no net reduction, we do not disregard any cases from the
actual decline of 25,000. Thus, the State would be entitled to the
entire 25-percent caseload reduction credit.
    Comment: Several commenters suggested that cases analogous to some
cases in separate State programs were not included in the FY 95
baseline and therefore would improperly inflate the comparison year
caseload, if included. All these commenters asked that we exclude
families in separate State programs from the caseload reduction
calculation. Others noted that, while the statute does not directly
address this issue, there is a legitimate need to look at cases in
separate State programs in the calculation. Otherwise, a State could do
something like simply move half of its cases to a separate State
program and assert a 50-percent caseload reduction.
    Response: Congress did not intend to give States credit for
caseload reductions resulting from changes in eligibility. We believe
that when a State moves a family receiving TANF assistance to a
separate State program, it would represent an eligibility change if we
did not include it in the caseload count; therefore, we have not
modified the regulation as some of the commenters suggest. However, as
noted elsewhere, we have modified the reporting for separate State
programs. This change has the effect of reducing the number and type of
SSP cases reported by the State.
    Comment: Commenters objected that we had inappropriately retained
discretion (by using the language, ``we will consider excluding
cases'') to exclude cases in separate State programs that duplicate
TANF cases or were made ineligible for Federal benefits by Pub. L. 104-
193. Several found that the three categories of exclusions appear to be
more ambiguous and discretionary than appropriate. Some commenters
thought the third category--cases receiving tax credits, child care or
transportation subsidies or other benefits for working families that
are not directed at their basic needs--was particularly confusing. Most
recommended that we explicitly exclude from the caseload reduction
calculation, cases in separate State programs that: (1) duplicate cases
in the TANF caseload; (2) provide assistance to immigrants made
ineligible for Federal benefits; and (3) provide income support or
services to low-income, working families for whom employment provides
the primary source of income.
    Response: Generally, we agree with the comments and have made
appropriate changes in the final rule. If a State provides
documentation on cases in separate State programs that meet either of
the following conditions, we will exclude them from the caseload count:
(1) cases that duplicate TANF cases; or (2) cases made ineligible for
Federal benefits by PRWORA and that are receiving only State-funded
cash assistance, nutrition assistance, or other benefits. We did not
include the third exception suggested by commenters since these cases
are no longer reported as SSP cases under the revised definition of
assistance.
    However, we note that these are the only circumstances under which
we will exclude separate State program cases from the caseload
reduction calculations. As we have indicated already, we believe that
moving a family receiving TANF assistance to a separate State program
where they are receiving assistance would represent a change in
eligibility criteria if we did not include such programs in the
caseload reduction calculation.
    Comment: We had wide-ranging and divergent comments on the
methodology and supporting data required of States. Several commenters
noted that a State policy that denies assistance when an individual
does not comply with work requirements, child support cooperation
requirements, or other behavioral requirements is the same as any other
eligibility requirement--it defines the categories of families that do
or do not qualify for assistance. Some commenters suggested that
enforcement mechanisms such as fingerprinting also deter eligible
families. They recommended that States receive credit only to the
extent that the number of families removed exceeds the number wrongly
denied, deterred or removed. Several commenters requested that the
final rule explicitly consider full-family sanctions, burdensome
verification requirements, and requirements that applicants engage in
certain activities to be changes in conditions of eligibility. Most
recommended that a State should not receive credit for any such changes
in its policy. Others suggested just the opposite, that full caseload
reduction credit is appropriate for all denials of assistance for
failure to comply with a behavioral requirement.
    Response: Under the final rules, we consider behavioral
requirements that divert families to be eligibility changes, and we
exclude them from assistance from the caseload reduction credit. We
believe it is appropriate to treat both full-family sanctions and
behavioral requirements as eligibility changes. Based on the comments,
we have tried to clarify explicitly that no type of Federal or State
eligibility change since FY 1995 that directly affects a family's
eligibility for assistance will count in a State's caseload reduction
credit. These changes include more stringent income and resource
limitations, time limits, full-family sanctions, and other new
requirements that deny families assistance when an individual does not
comply with work requirements (e.g., applicant job search), cooperate
with child support, or fulfill other behavioral requirements. A State
may count the reductions attributable to enforcement mechanisms or
procedural requirements that are used to enforce existing eligibility
criteria (e.g., fingerprinting or other verification techniques) to the
extent that such mechanisms or requirements identify or deter families
otherwise ineligible under existing rules.

Section 261.43--What Is the Definition of a ``Case Receiving
Assistance'' in Calculating the Caseload Reduction Credit? (Sec. 271.43
of the NPRM)

    To determine the caseload reduction credit, we proposed to consider
caseloads in both TANF and in any separate State programs that are used
to meet the maintenance-of-effort (MOE) requirement. Using the
definition of assistance proposed under part 270, we proposed to base
the calculation on all cases in the State receiving AFDC in FY 1995 and
TANF assistance for all other fiscal years.
    Comment: Several commenters asked us not to use the definition of
``assistance'' to calculate caseloads for periods prior to the State's
implementation of TANF. They argued that, since there was no definition
of assistance similar to the TANF definition, many States granted
assistance based on broader criteria. In particular, they pointed out
that EA cases often did receive one-time, short-

[[Page 17788]]

term assistance. Since they were legitimate IV-A cases, the commenters
maintained that the cases should be included in the number of cases
receiving assistance in 1995.
    Response: In the NPRM, we specified that the definition of
``assistance'' should be applied to the caseload count, but our
methodology did not actually allow a State to use the definition until
it had implemented the TANF program. The caseload information reported
by States on ACF-3637, which applied to a State until it implemented
the TANF program, reflected the AFDC and EA definitions. Under the
final rule, to get caseload data that are comparable to TANF, we adjust
the baseline AFDC and EA data, as appropriate, to estimate the
unduplicated cases receiving benefits under State programs in those
years that would have met the TANF definition of ``assistance.''
    We point out that this final rule does not dictate the
determination of caseload reduction credits for fiscal years 1997,
1998, or 1999. Thus, it does not cover the determination of credits for
periods when States were still operating AFDC and EA programs. For such
earlier periods, it would be appropriate to keep all unduplicated AFDC
and EA cases in the calculations because the base-year and prior-year
caseload figures would be comparable.
    These rules cover caseload reduction credits that apply in FY 2000
and thereafter--after States had converted to TANF. Since the
definition of ``assistance'' determines the prior-year caseload
numbers, it is appropriate to adjust the 1995 caseload numbers to
mirror the TANF definition of assistance, in order to compare ``apples
with apples.'' In some instances, that could mean that EA cases should
not be part of the 1995 base.
    Comment: One commenter recommended that we clarify this section to
ensure that, in calculating the caseload reduction credit, we only
include a percentage of the separate State program cases that equals
the State's MOE requirement (either 75 or 80 percent). Otherwise, the
commenter argued, the policy would discourage States from investing
more than the required MOE amount.
    Response: We agree with the comment and have revised the final rule
accordingly.

Section 261.44--When Must a State Report the Required Data on the
Caseload Reduction Credit? (Sec. 271.44 of the NPRM)

    Under the NPRM, we required a State to submit its caseload report
and estimates for each fiscal year by November 15. We proposed to
approve or reject a State's estimated reduction credit within 90 days,
that is, by February 15.
    Comment: Commenters expressed mixed feelings about the timeframes.
On the one hand, nearly everyone wanted States to receive the caseload
reduction credit and net participation requirement as early as
possible. On the other hand, most commenters, and especially State
commenters, suggested that the timeframes for responding to additional
information requests from the Department and for resolving issues were
not sufficient.
    Response: To give the public an opportunity to comment on State
estimates and ensure that States have adequate time to provide
additional requested information, we have modified the final rule. The
caseload reduction report and estimates are now due from States on
December 31. States may negotiate the deadline for additional
information or submit it within 30 days. As a result, we will provide
States with their caseload reduction credits no later than March 31.
Any extensions for submitting additional data that we grant to States
must be consistent with this deadline.

Subpart E--What Penalties Apply to States Related to Work Requirements?

    While PRWORA embodies State flexibility in program design and
decision-making, it also embodies the principle of accountability.
Where a State does not live up to the minimum standards of performance,
it faces serious financial penalties. One of the principal areas of
accountability is the State's provision of work and work-related
activities to promote employment and self-sufficiency. The work
participation rates are demanding, but designed to ensure that
recipients move as quickly as possible into work and toward
independence. This is especially important given the time-limited
nature of Federal TANF benefits.
    In structuring this part of the regulations, we have attempted to
balance the imperative of State accountability in the work
participation rates with the knowledge that each State enters TANF from
a different standpoint and with different ideas about the best way to
help its recipients.

Section 261.50--What Happens if a State Fails To Meet the Participation
Rates? (Sec. 271.50 of the NPRM)

    In accordance with section 409(a)(3) of the Act, as amended by Pub.
L. 105-33, if we determine that a State has not achieved either or both
of the minimum participation rates in a fiscal year, we must reduce the
SFAG payable for the following fiscal year. The initial penalty is five
percent of the adjusted SFAG and increases by two percentage points for
each successive year that the State does not achieve the participation
rates. We reduce the penalty amount based on the degree of
noncompliance, as discussed at Sec. 261.51. The total work
participation penalty can never exceed 21 percent of the adjusted SFAG.
(See Sec. 262.1(d) for a discussion of the total penalty limit under
TANF.)
    If a State fails to provide complete and accurate data on work
participation, as required under section 411(a) of the Act and
Sec. 265.8 of the regulations, we may determine that a State has not
achieved its participation rates, and the State will be subject to a
penalty under this part. We also have the authority to penalize a State
that does not report its work participation data for failure to report
(under section 409(a)(2) of the Act). However, in this case, we thought
it would be more appropriate to penalize the State for failure to meet
its work rate. First, this policy is consistent with the approach we
are taking when a State fails to report information related to other
penalty determinations. Also, we did not want to create a situation
where nonreporting States would face lesser penalties than reporting
States, and we did not believe duplicate penalties were warranted.
    We received some comments regarding the year in which we will
impose a penalty. We have addressed these comments at Sec. 262.1 of
this chapter.
    Comment: We received quite a few comments concerning our preamble
language indicating that we would impose a penalty for failure to meet
the work participation rates if a State failed to report complete and
accurate data on the work participation rates. Some commenters objected
to the policy altogether. Others suggested that we should only impose
the work participation penalty where, as a result of incomplete or
inaccurate data, we are unable to determine whether the State failed
the participation rates. Another commenter suggested that we specify
what ``complete and accurate'' means for the purposes of calculating
the participation rates.
    Response: Our intent in including this policy in the preamble was
only to impose a work participation penalty based on a State's failure
to report complete and accurate data if the lack of data impeded our
ability to determine whether the State actually achieved the required
rates. In fact, at Sec. 262.3 of this chapter, we indicate that this is
our

[[Page 17789]]

policy, stating that we will impose the participation rate penalty ``if
we find information in the reports * * * to be insufficient or if we
determine that the State has not adequately documented actions
verifying that it has met the participation rates.'' For clarity, we
have changed the wording above to indicate that we ``may'' impose such
a penalty, and we will implement the policy as explained in Sec. 262.3.
    Comment: One commenter urged us to penalize a State where an entity
with jurisdiction or group of people affected finds a systemic
violation of any applicable Federal law (e.g., title VI of the Civil
Rights Act).
    Response: We think it is appropriate to defer to the entity that
enforces a given Federal law to penalize a State that violates that
law. In general, the laws the commenter alludes to include specific
remedies for individuals that are adversely affected. At the same time,
we encourage States to make sure recipients are informed of their
rights to remedies under Federal, State and local laws.
    If, at a later date, we learn of a specific problem in this regard,
we will consider further action, but we think it is unnecessary to
include such penalties in the regulation at this time.

Section 261.51--Under What Circumstances Will We Reduce the Amount of
the Penalty Below the Maximum? (Sec. 271.51 of the NPRM)

    The statute requires us to reduce the amount of the penalty based
on the degree to which the State is not in compliance with the required
participation rate. The required rate for a State is the rate at
Sec. 261.23, adjusted for any applicable caseload reduction credit;
however, it specifies neither the measures of noncompliance nor the
extent of reduction. The statute also gives us the discretion to reduce
the penalty if the State's noncompliance resulted from certain specific
causes; we address this latter issue separately, in the section
entitled ``Discretionary Reductions.''
    As we indicated earlier, we have not included in the final
regulations the NPRM proposals that would have linked a State's
decisions about implementing separate State programs to its eligibility
for penalty relief. Thus, we have removed from Sec. 261.51 the
provision that would have denied penalty reduction to a State that
diverted cases to a separate State program for the purpose of avoiding
the work participation requirements. Please refer to the section
entitled ``Separate State Programs'' for a discussion of this policy
and the comments that we received relating to it.
Required Reduction
    We have significantly modified this part of the penalty reduction
section after considering the comments we received. In the NPRM, we
defined degree of noncompliance first by which of the rates a State
missed and second by how far it came from meeting the required rate.
Thus, if a State missed only the two-parent participation rate, we
proposed imposing a penalty that equaled, as a percentage of the
maximum possible penalty, no more than the State's percentage of two-
parent cases. Second, if the State missed the overall rate (or both
rates), we proposed reducing the penalty only if the State achieved a
threshold of 90 percent of the required rate. Above 90 percent, the
reduction was to be proportional.
    The final regulations use five basic criteria to measure the degree
of noncompliance: which participation rate the State failed; the amount
by which it failed; how well it succeeded in increasing the number of
recipients engaged in work (despite failing the participation rate(s));
the number of consecutive years in which the State failed the rates;
and the number of rates that the State failed.
    First, as in the NPRM, we will measure noncompliance on the basis
of whether the State failed one or both rates for the fiscal year and
which participation rate it failed, if only one. We believe that a
State that fails the two-parent rate should be subject to a smaller
penalty than a State that fails the overall rate or both. In addition,
we believe that it is appropriate to consider the size of the two-
parent caseload in deciding how much weight to give a failure of only
the two-parent rate.
    In looking at the data for FY 1996, we noted that the two-parent
participation rate, on average, affected a very small percentage of a
State's entire caseload--the mean State percentage was about 6.6
percent, but the median was only about 2.4 percent. We think a State
that failed with respect to only a small percentage of its cases should
not face a huge penalty. At the same time, we want to ensure that
States make adequate commitments to achieving the two-parent
participation rate and that our policies support State efforts to
extend benefits to two-parent families. We have attempted to balance
these goals.
    Under this rule, the maximum penalty a State could face for failure
to meet only the two-parent rate depends directly on how much of the
State's total caseload consisted of two-parent families. We have not
created a similar proportional reduction for a State that fails only
the overall rate because all cases, including two-parent cases, are
reflected in the overall rate.
    Second, we measure noncompliance on the basis of the severity of a
State's failure to achieve the required rate. In drafting the
regulation, we wanted to strike the right balance between the
importance of work and the requirement to reduce the penalty based on
the degree of noncompliance. Although our first inclination was to make
reductions in direct proportion to the State's achievement toward the
required rate, our experience in the JOBS program led us to consider
creating a threshold below which we would grant no reduced penalty. We
were concerned that, as in the JOBS Unemployed Parent participation
rates, there would be States with negligible levels of achievement,
particularly with respect to the two-parent caseload, and thus did not
merit a reduced penalty. Given that experience, we thought it was
essential to have a threshold.
    In the NPRM, we set the participation threshold at 90 percent, in
an effort to support the emphasis in the statute on making the work
penalty meaningful. In particular, Pub. L. 105-33 amended the work
penalty provision so that the amount was fixed, removing the discretion
we had under PRWORA to set a lesser penalty amount. We thought (and
continue to think) that this shows Congressional intent to provide a
work penalty of consequence. To avoid undercutting this intent, our
proposed rules required that a State make substantial progress in
meeting the target rates before we would consider a reduction. We
continue to believe that a threshold is a key part of the penalty
structure.
    We received extensive comments about the proposed 90-percent
threshold. Some commenters accepted our reasoning for creating a
threshold, but virtually all found a 90-percent standard to be
excessively high. They argued that it bases large fiscal consequences
on small and hard-to-measure differences in reported data. While the
NPRM maintained that we did not want to give relief to States with
negligible levels of achievement, thus leading us to a threshold,
commenters asserted that a high threshold treats achievers and
nonachievers the same. For example, a State that reaches 2 percent of
the required rate and one that reaches 88 percent of that rate are
subject to the same penalty. This, they argued, gives States a strong
incentive not to serve families with significant barriers. Further,
they pointed out that

[[Page 17790]]

it would also subject States with similar achievement levels to very
different penalties. For example, with a 40-percent participation rate,
a State that reaches 35 percent would be subject to a full penalty, but
a State that reaches 38 percent would be subject to less than half the
penalty. Although most commenters opposed having a threshold at all,
believing that any threshold is arbitrary, many suggested that if we
found it essential to have one, it should be set significantly lower.
Most recommended a threshold of 50 to 75 percent. A few commenters
suggested a lower threshold for the two-parent rate than for the
overall rate.
    After reviewing those comments and analyzing preliminary data, we
have set the threshold at 50 percent. We chose this threshold both
because it was the most widely recommended alternative level and
because we believe it is a logical standard. Requiring States to reach
at least half of the target rate draws a clear line between achievers
and nonachievers. We think it is reasonable not to grant penalty
reduction to States that are closer to a participation rate of zero
than they are to achieving the requirement.
    Under the final rules, we will reduce the penalty for any
qualifying State in direct proportion to the State's level of
achievement above a threshold of 50 percent. To achieve this, we will
compute a ratio whose numerator is the difference between the
participation rate a State actually achieved and the applicable
threshold rate and whose denominator is the difference between the
applicable required participation rate and the applicable threshold
rate.
    In the final rule, we have also clarified that the applicable
required participation rate and the applicable threshold both reflect
any caseload reduction credit that the State receives pursuant to
subpart D of part 261. In other words, the standard against which we
judge the degree of noncompliance recognizes that Congress wanted
States to get credit for the caseload reductions they achieve, as long
as they are not due to eligibility changes. If we did not include this
clarification, the threshold standard for some States could actually be
higher than the target (i.e., full compliance) rate provided under the
statute.
    For example, assume a State's adjusted target rate (i.e., after
applying its caseload reduction credit) equals 30 percent. Further
assume the State achieved 18 percent, which exceeds the threshold of 15
percent (one half of 30 percent) by 3 percentage points. The 3
percentage points equal 20 percent of 15 percent, the difference
between the required rate and the threshold. Therefore, we would reduce
the penalty amount by 20 percent.
    Commenters also urged us to consider a wide range of alternative
means of measuring noncompliance. On the whole, they urged us to give
States credit for their level of effort, rather than looking
specifically to a percentage of the participation rate. One commenter
offered that, if the purpose of penalties is to give States a strong
incentive to take the requirements seriously rather than to punish
those that fail, then a broader view of State achievement is in order.
In particular, several commenters suggested variations of the following
alternative factors for determining penalty reduction:
    <bullet> An increase in a State's caseload (in either the current
year or prior year);
    <bullet> Improvement in a State's performance over the prior year;
    <bullet> Increase in the number of participants in countable work
activities in a State, or in the number of participants in work
activities but below the required number of hours to count for
participation; and
    <bullet> The extent to which a State exceeded the overall rate,
even though it missed the two-parent rate.
    Some also suggested that we should recognize a combination of
alternatives, perhaps without even specifying a comprehensive list in
the regulation.
    One set of extensive comments on this issue put forward an argument
for treating any penalty reduction factors that we adopt in a formulaic
way so that a State's penalty liability is clear. Although this can
make for a complex provision, we have responded to this concern by
adding some detail to the final rule. We believe that this formula will
help States foresee the possible fiscal consequences of their policy
decisions.
    We considered all these alternatives measures from the perspective
that our primary interest in the participation rates is to encourage
work. As a result, we have modified the regulations to include as our
third measure of noncompliance an adjustment factor that reflects a
State's success in engaging additional recipients in countable work
activities. The factor rewards a State that increases the number of
individuals it engages in work by at least 15 percent over the previous
fiscal year. If the number of individuals engaged in work decreases,
the State would not be eligible for a penalty reduction, beyond the
proportional reduction for failing only the two-parent rate. For this
calculation, we will use the average monthly participation data, just
as we do in calculating the participation rates themselves.
    We calculate the adjustment factor by dividing the change in the
average number of individuals the State has engaged in work by 15
percent of the number it engaged in work the prior year. For example,
if the State engaged an average of 2,000 individuals each month in the
prior year, and 2,400 individuals in the current year, we would divide
400 (the change) by 300 (15 percent of 2,000, the prior year's average
monthly number engaged in work). This would result in an adjustment
factor of 1.33. In other words, in the example, the State's increase in
participants exceeded 15 percent of the prior year's level by one
third. Thus, under these rules, the State's penalty reduction would
increase by one third, compared to the reduction it would have received
if it had achieved only a 15-percent increase.
    We chose to tie the adjustment factor to a 15-percent increase to
approximate the average annual increase in the overall participation
rate.
    We based the adjustment factor on an increase in the number of
participants in work instead of on an increase in the percentage of
participants in work for two reasons. First, the proportional reduction
above the threshold already takes a percentage of participants into
account through the increase in the participation rate. Second,
commenters made a persuasive argument that measuring individuals would
reward States that actually showed greater success with work, where
participation percentages would be affected by caseload changes that
might have nothing to do with work or the State's efforts to engage
individuals in work.
    Readers will note that, in addition to the threshold, the
adjustment factor also serves as a trigger for penalty reduction; the
State must have an adjustment factor above zero to qualify for penalty
reduction beyond the proportional reduction for failing only the two-
parent rate. We needed to cut off the adjustment factor at zero because
a negative number would actually increase the penalty above the amount
described in Sec. 261.50, which we have no authority to do. We then
linked the presence of an adjustment factor to further penalty
reduction because we did not want to reward a State with a decrease in
the number of working recipients more than a State with a small
increase (under 15 percent) in the number engaged in work.
    Finally, we adjust the penalty reduction on the basis of whether
the State failed both participation rates in

[[Page 17791]]

the current year and how many consecutive years it failed them. If the
State met both participation rates in the previous year and only failed
one rate in the penalty year, we will apply the full reduction to the
penalty. If it failed both rates, but failed none the previous year, we
will decrease the penalty reduction by one half.
    For the second consecutive year of penalty liability, we will
prorate the penalty reduction by 50 percent if the State failed just
one rate; if it failed both rates, it is entitled to a 25 percent
reduction.
    If the State fails to meet the participation rates for three or
more years in a row, we will not reduce the penalty at all. We think
that this is a fair and reasonable approach to avoid rewarding a State
that has not successfully addressed a persistent problem and that
repeated failures is an appropriate indicator of the degree of
noncompliance. A State with successive failures could still claim a
discretionary work penalty reduction (as discussed below), claim a
reasonable cause exception, or enter the corrective compliance process.
    We have also added a paragraph indicating that we will adjust the
calculations in this section to exclude cases for which a State has
granted federally recognized good cause domestic violence waivers.
Based on the comments we received in this area, and given our
reasonable cause exception policy with respect to cases with federally
recognized good cause domestic violence waivers, we thought these
waivers should play a similar part in penalty reduction. For comments
about domestic violence waivers, please refer to the preamble section
entitled ``Treatment of Domestic Violence Victims.''
    To summarize the entire penalty adjustment process, we begin with
the proportional reduction based on the amount by which the State
exceeded the 50-percent threshold. Second, we calculate the adjustment
factor for increasing the number of individuals working and multiply
the reduction by the adjustment factor if it is positive, arriving at
an adjusted reduction. If the adjustment factor is zero or negative,
there is no adjusted reduction. Then we multiply the adjusted reduction
by the applicable penalty percentage, derived from whether the State
failed just the two-parent rate. Finally, we adjust the penalty based
on whether the State failed both rates and on the number of consecutive
years of failure.
    As we stated earlier, if a State does not qualify for an adjusted
reduction, it still may be eligible for the proportional penalty
reduction for failing only the two-parent rate.
    In spite of our desire to make this regulation as simple as
possible, we realize that this process is more complex than the
approach we adopted in the NPRM. We have taken very seriously the
commenters criticism that the proposed penalty reduction provision did
not look broadly enough at State success in work. We think the new
provision treats States more fairly and will be more effective at
encouraging work. Factoring in multiple ways of looking at such success
naturally makes the new methodology more complicated. In fact, we
considered several of the other alternatives that commenters suggested,
but ultimately decided that additional factors would make the
calculation too convoluted, without adding to the balance or the work
focus.
    Comment: We received a great many comments about linking the size
of the penalty for missing only the two-parent participation rate to
the proportion that two-parent cases make up of the State's total
caseload. Nearly all agreed with our approach; however, commenters put
forward two additional ideas. First, one commenter suggested linking
the size of the two-parent penalty to the national two-parent
proportion, rather than varying the penalty based on each State's two-
parent caseload. The second idea was to provide penalty relief for
States that have made policy choices that have expanded the 2-parent
caseload.
    Response: We have not adopted either of these recommendations.
While using a national caseload proportion would remove a possible
inadvertent incentive for a State to reduce the size of its two-parent
caseload, or a disincentive to expand eligibility, two-parent cases are
not distributed evenly across the States. Moreover, we think the
difference in penalty amounts would not be enough of an incentive to
drive State policy regarding two-parent cases.
    Regarding the issue of policies that increase the two-parent
caseload, we think that our policy of adjusting the penalty base to
reflect the two-parent caseload is the appropriate mechanism for
helping States with the two-parent participation rate. (We also
considered this issue in the context of the caseload reduction factor,
as addressed above. Please refer to subpart D of this part for further
discussion.)
    Comment: As we indicated above, we received many comments
suggesting alternative measures to use in penalty reduction. We listed
above the ones that were most persuasive or appeared most frequently.
The comments included others that we have not listed.
    Response: We think the new penalty reduction methodology we have
adopted gives States credit fairly for making substantive progress in
reaching the participation rates and supports State efforts to engage
recipients in work. It should be viewed as a whole because its various
components are designed to work in combination to achieve a balanced
result. While there are other factors that might also have worked well,
we believe that we have selected elements that would achieve these
goals and are easily calculable.
    Comment: A few commenters urged us to require a State to have a
system for monitoring and enforcing compliance with Federal employment
laws within its TANF program in order to qualify for a penalty
reduction.
    Response: As we have indicated earlier, we fully expect States to
conduct programs that are lawful and to uphold employment laws that
apply to working welfare recipients. We have chosen not to adopt this
suggestion out of deference to the enforcement mechanisms already
available under Federal law. However, we have created a new regulatory
section at Sec. 260.35 to reference existing employment and recipient
protections. Please refer to the section entitled ``Recipient and
Workplace Protections'' for a more detailed discussion of this issue.
Discretionary Reductions
    The final regulations reflect the discretion that we have to reduce
the amount of the penalty if the State could qualify as a needy State
for the Contingency Fund. The definition of ``needy State'' at
Sec. 260.30 is based on especially high unemployment or large numbers
of Food Stamp recipients in the State. (See subpart B of part 264 for
more discussion of how a State qualifies for the Contingency Fund.)
    Pub. L. 105-33 gave us the added discretion to reduce the penalty
if the State failed to meet the participation rate due to extraordinary
circumstances such as a natural disaster or regional recession. We have
modified this provision from the NPRM to include substantial caseload
increases among the examples of extraordinary circumstances. Although
this criterion is not given as an example in the statute, based on the
comments we received, we believe such a condition could constitute an
extraordinary circumstance and think it is appropriate to include it.
    To ensure that we take any such circumstances into consideration,
States should submit information describing

[[Page 17792]]

the extraordinary circumstances and their effects on the ability of the
State to meet the participation rates. We must provide a written report
to Congress to justify any penalty reductions that we grant under this
provision.
    One criterion for discretionary reductions is similar to the
criterion at Sec. 262.5(a)(1) for granting a reasonable cause exception
to a penalty due to a natural disaster. We will evaluate any
information a State submits concerning the effects of a natural
disaster on its ability to achieve the participation rates. If the
material does not support granting a reasonable cause exception, we
will consider whether it is sufficient for penalty reduction purposes.
For example, if the disaster caused a failure in only one small area of
the State, but the State missed the required participation rate by a
significant amount, we would not grant a reasonable cause exception,
but we might reduce the penalty in proportion to the TANF caseload in
that area. We intend to use a similar approach to evaluating the
effects of a regional recession.
    Comment: Some commenters urged us to add other factors to the
examples of discretionary reductions. Some suggested an open-ended
example such as ``other circumstances beyond the State's control''
while others gave specific suggestions, including: caseload increase;
sub-state recessions; widespread economic disruption from the closing
of a plant or significant numbers of lay-offs; chronic unemployment;
bad weather; and mismatch between available jobs and skills of
recipients.
    Response: As we indicated above, we have added substantial caseload
increases to the list of examples of extraordinary circumstances;
however, it is simply a list of examples. We believe the provision
leaves the flexibility for a State to make a claim of ``extraordinary
circumstances'' based on other factors. The final regulation indicates,
as did the NPRM, that we will consider the objective evidence of
extraordinary circumstances that a State submits. We have not specified
the basis on which we will evaluate that evidence or apply a reduction.
We believe this responds to the recommendation of commenters that we
should have flexibility under our rules to address situations that we
could not foresee at this writing. Since the extraordinary
circumstances are likely to be different in each case, we think it is
most appropriate to use the discretion available to us to evaluate the
materials that a State submits to determine whether its claim warrants
a reduction in penalty amount.

Section 261.52--Is There a Way To Waive the State's Penalty for Failing
To Achieve Either of the Participation Rates? (Sec. 271.52 of the NPRM)

    Section 409(b) of the Act creates a reasonable cause exception to
the requirement for certain penalties, including failure to meet the
minimum participation rates. If we determine that a State has
reasonable cause, we cannot impose a penalty.
    We have included general reasonable cause criteria at Sec. 262.5.
These apply to any of the penalties for which there are reasonable
cause exceptions. The preamble to Sec. 262.5 discusses how we arrived
at these criteria, as well as our general philosophy about the role of
reasonable cause exceptions.
    For the work participation rate penalty, two additional, specific
reasonable cause exceptions apply. Under the regulation at Sec. 261.52,
a State may demonstrate that its failure can be attributed to its
granting of federally recognized good cause domestic violence waivers
under the Family Violence Option. In this case, the State must show
that it would have achieved the required work rates if cases with these
waivers were removed from both parts of the calculation (i.e., from the
numerators described in Secs. 261.22(b)(1) and 261.24(b)(1) and the
denominators described in Secs. 261.22(b)(2) and 261.24(b)(2)). A State
must grant domestic violence waivers in accordance with criteria in
subpart B of part 260 to be eligible to qualify as federally recognized
good cause domestic violence waivers and to receive a reasonable cause
exemption on these grounds. We have explained this policy and responded
to comments on this subject in subpart B of part 260.
    The regulation also provides that a State may receive a good cause
exemption if it demonstrates that its failure to achieve the work
participation rates can be attributed to the provision of assistance to
refugees in a federally-approved alternative project.
    Finally, this section of the regulation indicates that States may
dispute our findings that they are subject to a penalty.
    Comment: Many commenters urged us to expand the reasonable cause
exceptions specifically available for failure to meet the work
participation rates. They suggested a variety of additional criteria,
such as a high incidence of recipients with severe employment barriers,
a significant refugee population, correcting unlawful employment
discrimination, conflicts with other Federal requirements (including
the FLSA) or litigation, and enforcing the nondisplacement provisions.
Some commenters, paralleling the domestic violence exception, suggested
that the provision of targeted services to other groups of recipients
with significant barriers to employment should entitle a State to a
reasonable cause exception. Others recommended many of the same
criteria suggested for reducing a participation rate penalty, including
caseload increases, economic downturns, and increases in the number of
recipients the State engages in work or places in countable activities
but below the hours standard. Many also suggested granting a reasonable
cause exception for a combination of factors. Also, a number of
commenters urged us to leave the reasonable cause criteria in this
provision open-ended so that a State could present its arguments for an
exception as situations arise and each could be evaluated on its own
merits.
    Response: Although these comments appear in the context of
exceptions to the work participation rate penalty, many commenters made
the same arguments regarding the general reasonable cause criteria at
Sec. 262.5. We have addressed comments that apply broadly to reasonable
cause exceptions in that section.
    We continue to believe that the best way to address a State's
difficulty in meeting a program requirement is through the corrective
compliance process. This holds true for the participation rates as much
as, if not more than, any other requirement. Families, States, and the
Federal government are better served by solving the problem than by
forgiving it, or by imposing a penalty. It is for this reason that we
have chosen to limit reasonable cause exceptions, particularly those
that relate to a specific provision, as in the case of the
participation rates, and have not added the criteria suggested.
Nevertheless, under Sec. 262.5, a State may present a case for a
reasonable cause exception outside the ones specifically listed. We
think that the revised language in this section, together with a
State's ability to dispute our finding of a penalty, the corrective
compliance process, and the opportunities for work penalty reduction,
sufficiently recognize the difficulties States may face in meeting the
participation rates.

Section 261.53--May a State Correct the Problem Before Incurring a
Penalty? (Sec. 271.53 of the NPRM)

    The process for developing a corrective compliance plan does not
differ from one penalty to the next,

[[Page 17793]]

although the content of the plan naturally would. Thus, the regulation
refers to Sec. 262.6, the general section on submittal of a corrective
compliance plan for any penalty.
    Readers should note that Sec. 262.6(e) establishes a maximum
corrective compliance period for failure to meet the work participation
requirements. Since we measure participation annually, we will measure
compliance based on performance during the fiscal year that ends at
least six months after we receive the State's corrective compliance
plan.
    In this section, we establish a specific threshold that States must
achieve in order to be considered for a reduced work penalty under
Sec. 262.6(j) for making significant progress toward achieving
compliance. A State must increase its participation rate during the
compliance period enough to fill at least half the gap between the
participation rate it achieved in the penalty year and the required
rate for the compliance period. In other words, we will divide the
difference between the rate achieved during the compliance period and
the rate achieved during the penalty year by the difference between the
required rate for the compliance period and the rate achieved during
the penalty year; a result of at least 0.50, qualifies the State for a
possible reduction.
    You should note that, in this final rule, the required rate for the
compliance period reflects any caseload reduction credit that the State
receives under subpart D of part 261. We believe that this adjusted
rate reflects the performance standard that Congress intended would
apply to States.
    We also believe that making more progress toward the rate than
failure--that is, achieving at least 50 percent--is a reasonable
standard for significant progress. Thus, at the point at which a State
reaches this threshold, we may reduce its work penalty under the
corrective compliance provision.
    This approach is similar to the one taken in Sec. 261.51, with
respect to potential reductions in work penalties based on degree of
noncompliance. In both cases, we expect significant compliance in order
to merit a reduced penalty. However, we look at performance over
different periods in the two provisions.
    Comment: Several commenters thought that the 50-percent standard of
achievement measured against the ``new'' rate was restrictive and
arbitrary. Commenters proposed two basic alternatives. Many urged us to
set a threshold of achievement based on a particular State's
circumstances or to negotiate a State's threshold in the corrective
compliance plan process. Some thought that we should consider a State
to be in compliance if it achieves the participation rate associated
with the year for which it was subject to the penalty. (Presumably, if
we were to use a threshold to reduce the penalty in this scenario, it
would be applied against the latter rate.) One commenter thought that
we should link the threshold to the average increase among States with
corrective compliance plans, and another suggested that States should
be able to show significant improvement by means other than reaching
the threshold. Another commenter remarked that we made no provision for
circumstances arising in the year following the penalty year that
prevent a State from reaching the threshold.
    Response: We note that this provision applies a second reduction to
a State's penalty amount, the first (described at Sec. 261.51) having
been significantly expanded over the original proposal. This reduction
follows a corrective compliance period in which the State should have
been applying the steps of its plan to resolve the participation rate
problem. Given these circumstances, we think it is appropriate to
maintain a fairly rigorous standard for reducing a penalty still
further. Moreover, we do not think that a 50-percent threshold is
overly demanding--it simply requires a State to be more successful,
rather than less successful, in coming into compliance. We measure
progress against the ``new'' rate (i.e., the one that applies for the
corrective compliance plan year) because to do otherwise would suggest
that the State is not being held to the same standard as all the others
for that year. Otherwise, we would effectively give a State an extra
year to achieve the minimum participation rate. We expect a corrective
compliance plan to allow a State to come into compliance with the
applicable rates. Thus, the penalty reduction associated with
corrective compliance should use that standard.
    If circumstances arise during the corrective compliance plan period
that prevent the State from achieving the threshold, it is free to
claim a reasonable cause exception or develop a corrective compliance
plan for the penalty year, but we do not think it is appropriate to
reduce the prior penalty on that basis. In addition, the State might
qualify for penalty relief under Sec. 262.6(j)(2), relating to natural
disasters or regional recessions during the compliance period.

Section 261.54--Is a State Subject to Any Other Penalty Relating to Its
Work Program? (Sec. 271.54 of the NPRM)

    In accordance with section 409(a)(14) of the Act, as amended by
Pub. L. 105-33, if we determine that a State has violated 407(e) of the
Act in a fiscal year, which relates to when a State must impose
penalties on individuals who refuse to engage in required work, we must
reduce the SFAG payable for the following fiscal year by between one
and five percent of the adjusted SFAG.
    Comment: One commenter thought we did not provide adequate guidance
concerning the means by which we will judge whether a State has
violated the sanctioning requirement. Without such guidance, the
commenter thought that States might sanction families more severely
than necessary to avoid a potential penalty.
    Response: As we indicated at Sec. 262.3, we will use the single
audit to assess whether a State is complying with section 407(e) of the
Act and thus whether it is liable for a penalty under this provision.
We expect that, if there are widespread problems with States'
sanctioning practices, our data collection and the audits will help
identify them.
    While we understand the commenter's concern that States will
``over-sanction'' to avoid this penalty, it is important to understand
that this penalty applies both to a State's failure to sanction when it
should have and to its imposition of a sanction when it should not have
imposed one. Thus, a State that overreacts by sanctioning too readily
could be equally liable for a penalty. If the commenter is concerned
that States will impose larger sanctions than they would otherwise, we
would point out that States have the explicit authority, independent of
this penalty provision, to impose sanctions that are greater than pro
rata reductions, up to and including terminating assistance to the
case.
    Comment: One commenter objected to our intention to collect
sanction policy information via Sec. 265.9, stating that such
information was available in the TANF State plans.
    Response: While some States may have included sanctioning policies
in their TANF plans, the statute does not require it. Thus, we cannot
count on obtaining this information through the plans. Also, at best,
the plan information would only tell us about State policy, not State
practice (e.g., the nature and scope of sanctions imposed).

[[Page 17794]]

Section 261.55--Under What Circumstances Will We Reduce the Amount of
the Penalty for Not Properly Imposing Penalties on Individuals?
(Sec. 271.55 of the NPRM)

    The statute requires us to reduce the amount of the penalty based
on the degree to which the State is not in compliance with the section
407(e) of the Act.
    In determining the size of any reduction, we will consider two
factors. First, we will examine whether the State has established a
control mechanism to ensure that the grants of individuals are reduced
for refusing to engage in required work. Second, we will consider the
percentage of grants that the State has failed to reduce in accordance
with the statute.
    As we indicated in the preamble to Sec. 261.14, States have the
discretion to define the term pro rata reduction. Under Sec. 265.9, as
part of the annual report we require each State to provide us with a
description of how it will carry out a pro rata reduction. This
information will help us determine whether States are taking sanctions
appropriately. Also, these definitions will help us determine whether
States face an equitable and level playing field under this penalty
provision.
    Some commenters noted that the proposed rules incorrectly specified
that reasonable cause and corrective compliance did not apply to this
penalty. We have deleted the provision that included this inadvertent
error.
    Comment: One commenter urged us to clarify what we mean by control
mechanisms.
    Response: We did not want to limit a State's range of possible
control mechanisms by creating a single definition. However, one
example of a possible control mechanism would be a system that
identifies cases in which an individual refused to participate, then
cross-checks those cases against information on sanction actions, and
corrects any errors in sanctioning.
    Although we did not define a control mechanism in the regulation,
there are some additional elements that we expect a State to include in
a control mechanism to ensure appropriate sanctioning of recipients.
Section 402(a)(1)(B)(iii) of the Act provides that a State must set
forth objective criteria for fair and equitable treatment of
recipients, including an explanation of how the State will provide an
opportunity for recipients who have been adversely affected to be heard
in a State administrative or appeal process. We think that any State
mechanism that controls whether sanctions have been imposed properly
should ensure that recipients are informed of their rights to fair
hearings and advised of the process for invoking that right. In
addition, we encourage States to consider adding procedures to advise
recipients of their rights to pursue other remedies that might be
available under State and local laws.
    Comment: A commenter, citing the fact that States have a right
under the regulations at Sec. 262.7 to appeal a finding that it is
subject to a penalty, urged us to ensure that individuals are accorded
a similar right.
    Response: As we explained in the previous comment, section
402(a)(1)(B)(iii) of the Act accords recipients the right to appeal
adverse actions. While we are not regulating this provision itself, we
do expect that States will address this requirement as part of their
sanctioning control mechanisms, and we will take it into consideration
in determining any reduction to the amount of this penalty.
    Comment: A commenter expressed concern that examining only
sanctioning data, without data from cases not sanctioned, as the basis
for the penalty would lead to unnecessarily harsh sanctions. The
commenter recommended basing the penalty determination solely on
whether the State has established control mechanisms.
    Response: As we indicated in the previous section, this penalty
applies to all violations of the sanctioning requirement, whether
failing to sanction inappropriately or imposing sanctions
inappropriately. For example, we anticipate sampling sanctioned cases
to determine whether a State has imposed sanctions without evidence of
a recipient's refusal to participate. Thus, a State has just as much
incentive to exercise restraint in sanctioning as to impose sanctions
too readily. At the same time, States may impose sanctions that are
greater than pro rata reductions without violating section 407(e) of
the Act.
    Comment: A commenter urged us to base the penalty amount on the
amount of the sanctions that should have been imposed, as a percentage
of the total amount of grants the State awards, or as a percentage of
the total grants that should have been reduced but were not.
    Response: This approach seems overly complex to us. We see no
advantage to basing the reduction on dollar amounts instead of case
percentages.
    Comment: A commenter recommended that we allow a tolerance for
errors before imposing a penalty under this provision.
    Response: We have built a tolerance for errors into the reasonable
cause exceptions at Sec. 262.5. In addition, States have the right to
dispute our determination that it is subject to a penalty, in
accordance with the provision at Sec. 262.4.
    Comment: A commenter urged us to deem 80-percent compliance as full
compliance with the requirement because the penalty amount must be
between 1 and 5 percent.
    Response: We have not established a specific formula for
determining and reducing the amount of the penalty. We will factor in
objective evidence of whether the State has established a control
mechanism, as discussed above, and of how many cases have been
improperly sanctioned.

Section 261.56--What Happens if a Parent Cannot Obtain Needed Child
Care? (Sec. 271.15 of the NPRM)

    Readers will note that we have moved the substance of this section
from Sec. 271.15 of the NPRM to Sec. 261.56 of the final rule. The
proposed rules contained two sections dealing with the question of
sanctions for parents of young children who refuse to work because they
cannot find needed child care. The first section specifically addressed
the statutory protections from sanctioning available to such
individuals who could not obtain child care; the second dealt with the
penalties that a State would face if it sanctioned individuals in
violation of the exception. Because of the close interrelationship
between these two provisions and the number of comments we received on
them, we thought that putting the regulatory sections adjacent to one
another would make the provisions easier to follow. We have retained
Sec. 261.15 to ensure that subpart A, which relates to the
responsibilities of individuals under TANF, continues to discuss the
child care exception.
    To support the intent of the statute to move people to work,
section 407(e) of the Act requires that States reduce or terminate
assistance to individuals who refuse to engage in work required by
section 407 of the Act. However, as we discussed in the preamble to
Sec. 261.15, a State may not reduce or terminate assistance to a single
custodial parent caring for a child under age six for refusing to
engage in required work, if the parent demonstrates an inability (as
determined by the State) to obtain needed child care. This exception
applies to penalties the State imposes for refusal to engage in work in
accordance with either section 407 or section 402(a)(1)(A)(ii) of the
Act. The

[[Page 17795]]

parent's demonstrated inability must be for one or more of the
following reasons:
    <bullet> Appropriate child care within a reasonable distance from
the individual's home or work site is unavailable;
    <bullet> Informal child care by a relative or under other
arrangements is unavailable or unsuitable; or
    <bullet> Appropriate and affordable formal child care arrangements
are unavailable.
    Refusal to work when the State determines an acceptable form of
child care is available is not protected from sanctioning.
    Because each State has the authority to determine whether the
individual has adequately demonstrated an inability to obtain needed
child care, we expect the State to define the terms ``appropriate child
care,'' ``reasonable distance,'' ``unsuitability of informal care,''
and ``affordable child care arrangements.'' The State must also provide
families with the criteria (including the definitions) that it applies
in implementing the exception and the means by which a parent can
demonstrate an inability to obtain needed child care.
    To keep families moving toward self-sufficiency and to promote
State compliance with this penalty exception, our rules provide that
States must have processes or procedures in place that: (1) enable a
family to demonstrate its inability to obtain needed child care; (2)
inform parents that the family's benefits cannot be reduced or
terminated when they demonstrate that they are unable to work due to
the lack of needed child care for a child under the age of six; and (3)
advise parents that the time during which they are excepted from the
penalty will still count toward the time limit on Federal benefits at
section 408(a)(7) of the Act, if applicable.
    In response to numerous comments, as discussed below, the language
in Secs. 261.56 and 261.57 reflects these expectations. In this
section, which focuses on the responsibilities of the State to inform
parents, we also require that the information States provide must
include the definitions or criteria that the State uses in its
determination process.
    The regulations for the Child Care and Development Fund (CCDF)
reinforce the importance of providing this vital information to parents
by also requiring the child care lead agency, as part of its consumer
education efforts, to inform TANF parents seeking child care in the
CCDF system of the existence of the child care exception and how to
demonstrate an inability to obtain needed child care.
    The CCDF rule requires the lead agency for child care to coordinate
with the TANF agency in order to understand how the TANF agency defines
and applies the terms of the statute regarding the penalty exception
and to include the definitions of the terms or criteria in the CCDF
plan.
    We took this child care rule into consideration in drafting our
proposed rule. Under Sec. 271.15, we required that the definitions and
criteria be submitted, but did not specifically require that the TANF
agency submit them. Our goal was to ensure that this information was
available for audit and penalty purposes and that it be part of the
public record, not to create an unnecessary burden for States. We have
not altered this policy in these final regulations.
    We received many comments on the provisions in this section and
made changes as discussed below.
    Comment: Most commenters objected to having the responsibility for
informing families about the child care exemption in the hands of the
child care lead agency and urged that we give the responsibility to the
TANF agency.
    Response: In the NPRM, we did not specifically require the TANF
agency to inform clients about the exception to sanctioning because the
CCDF NPRM (now the CCDF final rule) already required it. In the NPRM
preamble, we stated our expectation that States would inform clients,
but did not name the entity responsible. Our intent was to avoid
imposing an additional Federal burden on the States where the CCDF
requirement addressed the situation adequately. However, advocates and
States alike made a compelling argument that not all TANF clients
covered by this protection would necessarily be referred to the child
care lead agency. Therefore, we have revised the regulatory language at
Sec. 261.56. In the final rule, the TANF agency must inform clients of
the existence of the child care exception to sanctions and how to
demonstrate an inability to obtain needed child care. This requirement
is in addition to the requirement, in the CCDF rules, that the CCDF
agency inform TANF parents about the exception.
    Comment: Many States objected to our requiring criteria and
definitions, arguing that we had shifted the burden of proof from the
individual to the State. We also received a few general comments to the
effect that our rules did not adequately protect individuals from harsh
State policies.
    Response: We do not believe that requiring States to inform parents
of their rights, including the definition of key terms in those rights,
shifts the burden of proof to States. The individual needs to know how
the State defines key terms to determine whether the exception applies
to his or her case.
    Regarding the concern over harsh State policies, States have
considerable latitude in implementing the child care protections. We
think the final regulations protect families as much as possible, given
the regulatory restraints of section 417.
    Comment: A few commenters urged us to require States to inform
recipients about available child care subsidies and to assist them in
obtaining appropriate and affordable child care.
    Response: While we agree that assisting recipients locate child
care is a reasonable expectation, the statute at section 417 limits our
ability to regulate in this area. Given that child care is widely
recognized as a fundamental supportive service, necessary for
recipients to obtain and maintain employment, we are confident that
States will adopt practices that inform recipients about available
child care providers. States understand the importance of employment
retention and career advancement for recipients. In fact, the
publication ``Working Out of Poverty'' by the NGA Center for Best
Practices, recognizes the need to inform recipients of the availability
of transitional supports such as child care and transportation
assistance early, for example, during eligibility determinations and
assessments, and as part of job search and job readiness programs.
    Comment: Some commenters were concerned that the NPRM left room for
a parent who wishes to use a particular type of child care that is not
available to refuse appropriate available child care arrangements,
without risk of a penalty. For example, they feared that a parent who
wants only informal relative care, but has no relative available to
provide care, could refuse affordable, suitable center-based care.
States argue that this result would be contrary to Congressional intent
and the goals of the Act. They urged us to make clear that refusing
work under such circumstances is not protected under the child care
exception to a sanction.
    Response: This issue stems from an interpretation of the wording of
the statute, which uses the phrase ``one or more'' in describing the
reasons for a parent's demonstrated inability to obtain needed child
care. However, we agree with the commenters that such a result would be
contrary to Congressional intent, which was to protect individuals from
sanction when there was no appropriate child care, not

[[Page 17796]]

to give families a loophole to avoid work requirements. Further, such
an interpretation would be contrary to the best interest of the family,
because the TANF clock continues to run during such a period.
Therefore, we have revised the regulatory language at Sec. 261.56 to
clarify that refusing to work when an acceptable form of child care is
available is not protected from sanctioning.
    Comment: One commenter was concerned that the NPRM, as written,
might create a larger problem of inadequate child care due to informal,
uncertified or unlicensed child care providers. The commenter was
concerned that this would result in caregivers with inadequate training
in child development or basic life-saving skills, poor or no
curriculum, or no health or dental care referrals.
    Response: The statute, as reflected in the NPRM, intended to give
parents some choice in child care arrangements. Informal care is only
one possible type of child care arrangement that families could use. If
the State uses CCDF funds to provide child care, the regulations
governing the CCDF program require States to have standards for
informal providers, as well as those providers who are licensed. Under
TANF, we do not have the authority to regulate child care providers.
Accordingly, we have not amended the rules in response to the comment.

Section 261.57--What Happens if a State Sanctions a Single Parent of a
Child Under Six Who Cannot Obtain Needed Child Care? (Sec. 274.20 of
the NPRM)

    As we discussed in the prior section, the statute at section
407(e)(2) protects single custodial parents of children under age six
from sanction for refusing to work when they cannot obtain needed child
care. They must demonstrate that they could not obtain child care for
one or more of the following three reasons: (1) Appropriate child care
was not available within a reasonable distance from the parent's home
or work site; (2) informal child care, by a relative or under other
arrangements, was unavailable or unsuitable; and (3) appropriate and
affordable formal child care arrangements were unavailable. However,
refusal to work when an acceptable form of child care is available is
not protected from sanctioning.
    Section 409(a)(11)(A) of the Act directs the Secretary to reduce by
no more than five percent of the adjusted SFAG, the SFAG payable to a
State that violates this sanctioning protection. To determine that a
State is liable for a penalty, we must find that the State reduced or
terminated assistance to a parent who qualified for a sanctioning
exception under the definitions or criteria that the State developed
regarding a parent's ``demonstrated inability'' to obtain needed child
care.
    We will consider the following factors in determining whether a
State has violated the exception to the sanctioning requirement at
section 407(e)(2) of the Act:
    <bullet> Whether the State informs families about the exception to
the penalty for refusing to work, including the fact that the exception
does not extend the time limit on benefits;
    <bullet> Whether the State informs families about the process or
procedures by which they can demonstrate an inability to obtain needed
child care;
    <bullet> Whether the State has defined ``appropriate child care,''
``reasonable distance,'' ``unsuitability of informal care,'' and
``affordable child care arrangements,'' and informed parents of these
definitions;
    <bullet> Whether the State notifies the parent of its decision to
accept or reject the parent's demonstration in a timely manner;
    <bullet> Whether the State has developed alternative strategies to
minimize the amount of time parents are excepted from work requirements
due to their inability to obtain needed child care. For example, a
State that uses the services of a child care resource and referral
office might grant ``good cause'' based on a statement from that office
attesting to the unavailability of appropriate or affordable child
care. However, it could implement a system for automatically rechecking
the availability of care every few weeks. If the inability to work were
due to difficulty in arranging transportation, the State could use bus
and rail rates and schedules to help the recipient find appropriate
child care within a reasonable distance.
    We are not specifying the process or procedures that States should
develop or the documents, if any, States should require. However, we
suggest that, if States plan to require documents, they select ones
that are readily available to families. We recommend that the process
or procedures be simple and straight-forward. In addition, we recommend
frequent contact with parents, since the penalty exception does not
stay the time limit and there may be fluctuations in the availability
of child care services.
    We will impose the maximum penalty if a State does not have a
process or procedure in place that enables families to whom this
provision applies to demonstrate that they have met the guidelines
provided by the State. Additionally, we will impose the maximum penalty
if there is a pattern of substantiated complaints from parents or
organizations verifying that a State has reduced or terminated
assistance in violation of the requirement at section 409(a)(11) of the
Act. We may impose a reduced penalty if the State demonstrates that the
incidents were isolated or that a minimal number of families were
affected.
    States faced with a penalty under this provision may claim
reasonable cause and/or submit a corrective compliance plan as
described in part 262.
    We expect that, because of the interrelationship between TANF and
CCDF, TANF staff will work in close coordination with the lead agency
for child care. Our expectation is that TANF staff will provide
families with information about the penalty exception and the process
and procedures developed by the State to demonstrate an inability to
obtain needed child care. Under the CCDF rule, ACF requires that the
lead agency for the CCDF program provide the same information to TANF
parents who are seeking child care in the CCDF system. In addition, ACF
requires the lead agency for child care to include in the CCDF plan the
TANF agency's definitions for ``appropriate child care,'' ``reasonable
distance,'' ``unsuitability of informal care,'' ``affordable,'' and
``child care arrangements.'' Thus, we expect the State TANF agency to
share its definitions of these terms with the child care agency. Both
agencies will then be able to share them with families whom they may be
assisting with child care arrangements.
    We received few comments on this section. They are discussed below.
We also made one minor editorial change to Sec. 261.57(c); the word
``will'' was changed to ``may'' in recognition of the variables that we
need to consider in a decision to impose a reduced penalty.
    Comment: One commenter suggested that we should review a sample of
cases of sanctioned individuals to ensure that they were actually
informed of their rights and that the State did not disregard a
demonstration of the lack of availability of care.
    Response: We agree. Since the primary vehicle for monitoring the
requirement will be the single State audit, we are developing
procedures that include the review of a sample of cases in which
benefits have been reduced or terminated due to a parent's failure to
comply with the work requirements.

[[Page 17797]]

    Comment: One commenter disagreed with the proposed regulation
because States are threatened with penalties based on isolated
instances when they do not follow the procedures they have reported to
us. The commenter argued that imposing a penalty for isolated
noncompliance would have a chilling effect on enforcing work-related
sanctions.
    Response: We disagree with the comment. In the proposed rule, we
stated that we would impose the maximum penalty of five percent if: (1)
The State did not have a statewide process in place that enables
families to demonstrate their inability to obtain child care (although
the State's process does not need to be uniform statewide, there simply
needs to be a process in all areas of the State); and (2) there were a
pattern of substantiated complaints that verifies that a State had
terminated assistance in violation of the requirement. A ``pattern of
substantiated complaints'' does not include isolated cases that affect
few families and occur in relatively few jurisdictions. This means that
we will not impose a maximum penalty based on a few aberrant situations
when it is clear that the State established a statewide procedure.
Accordingly, we have not modified the final rules in this regard.

Subpart F--How Do Welfare Reform Waivers Affect State Penalties?

Section 261.60--How Do Existing Welfare Reform Waivers Affect a State's
Penalty Liability Under This Part? (Sec. 271.60 of the NPRM)

    Based on our changes to the regulatory provisions relating to
waivers, we have modified this section. Under the NPRM, this section
described how welfare waivers affected the participation rates. In the
final rule, it merely cross-references subpart C of part 260, which
addresses welfare reform demonstration waivers comprehensively.
    We have responded to all comments relating to waivers in the
preamble section entitled ``Waivers.''

Subpart G--What Nondisplacement Rules Apply in TANF?

Section 261.70--What Safeguards Are There To Ensure That Participants
in Work Activities Do Not Displace Other Workers? (Sec. 271.70 of the
NPRM)

    The regulations incorporate the statutory prohibition against
allowing an individual participating in TANF work activities from
displacing another employee. A participant in a work activity may not
fill a vacancy that exists because another individual is on layoff from
the same or equivalent job. Also, a participant may not fill a vacancy
created by an involuntary reduction in workforce or by the termination
of another employee for the purpose of filling a vacancy with a
participant.
    The statute and the final rule also require States to establish and
maintain grievance procedures for resolving complaints of alleged
violations of the restrictions on displacing workers. Readers should
note that we have added a new reporting requirement at
Sec. 265.9(b)(7), under which each State must provide us with a
description of its grievance procedures for resolving complaints of
displacement as part of its annual report if it has not included a
description in its State TANF plan.
    We encourage States to take aggressive steps to ensure that the
current work force is not harmed or their employment jeopardized in any
way by a State's efforts to place welfare recipients in employment or
work-related positions. Our ultimate goal, and that of States, is to
increase the ranks of the employed, not to substitute one group of job-
seekers for another. Displacing current workers is counter-productive
and damages the overall stability of the labor force. We are confident
that States will develop procedures for working with employers to
protect against displacing other employees.
    Comment: A few commenters urged us to establish minimum standards
for State grievance procedures and to require that a State notify
workers of those procedures and of the remedies available to displaced
workers. Similarly, another commenter urged us to create standards for
other aspects of this provision. At least one commenter recommended
that, if we thought we did not have the authority to impose such
requirements, then instead we should deny penalty reduction to States
that do not establish effective grievance procedures or ensure
widespread notice of their procedures.
    Some commenters urged us to reference the WtW interim rules, which
included more extensive nondisplacement provisions, and to recommend
that States use one set of grievance procedures for both programs.
    Response: Section 417 of the Act limits the authority of the
Secretary to regulate the conduct of States or enforce TANF provisions,
except where specifically provided for in the statute. Thus, it is not
consistent with the principle of State flexibility embodied in PRWORA
for us to regulate a State's administrative procedures. In particular,
in this provision, there is an explicit expectation of deference to
State and local laws, which we have reflected in paragraph (c) of this
section. Moreover, we do not have penalty authority with respect to the
enforcement of the nondisplacement provision and would be reluctant to
create a structure that duplicates or conflicts with existing
enforcement mechanisms that have a clear foundation under law. For
these reasons, we have not modified the regulation to establish minimum
standards for grievance procedures or to deny access to penalty
reduction.
    Using one set of grievance procedures for both programs should
prove easier for States, employers, and workers alike. We urge States
to consider adopting this approach. However, we note that not all
States have established WtW programs, and there may be reasons that a
unified grievance procedure would not be appropriate.
    Comment: One commenter urged us to add several provisions to the
nondisplacement section in order to prevent displacement more broadly.
The suggested additions included prohibiting filling a position that:
would otherwise be a promotional opportunity for a current employee;
did not comply with applicable personnel procedures; was caused by a
strike or other labor dispute; or was an established unfilled public
agency position, unless unfunded in the budget.
    Response: The nondisplacement provisions in the statute are very
explicit. Under PRWORA, we do not have the authority through
regulations to expand the definition of nondisplacement, even if we
support the commenter's suggestions. However, expanded definitions may
be available under State law or policy.
    Comment: One commenter asked us to explain how we would educate
State welfare administrators regarding compliance with the
nondisplacement provisions.
    Response: The section entitled ``Recipient and Workplace
Protections'' describes initiatives by various agencies within our
Department and elsewhere in the Federal government to inform State
agencies about the requirements of Federal employment laws. Please
refer to that section for further information on these efforts.

VII. Part 262--Accountability Provisions--General (Part 272 of the
NPRM)

    As we noted earlier in the preamble under our discussion of
waivers, we moved the waiver provisions of Sec. 272.8 of the NPRM to
subpart C of part 260. You will find the comments that we received on
Sec. 272.8 there.

[[Page 17798]]

Section 262.0--What Definitions Apply to This Part? (Sec. 272.0 of the
NPRM)

    This section cross-references the general TANF regulatory
definitions established under part 260.
    We received no comments on this section.

Section 262.1--What Penalties Apply to States? (Sec. 272.1 of the NPRM)

    Section 409 includes 15 penalties that may be imposed on States.
This rule covers 14 of the 15. This rule does not include the specific
penalty dealing with substantial noncompliance with requirements under
title IV-D (section 409(a)(8)). Our Office of Child Support Enforcement
is addressing this penalty in a separate rulemaking. However, since the
penalty is one of the TANF penalty provisions, the general procedures
and the appeal process in this rulemaking will apply.
    The penalties that we are regulating are:
    (1) A penalty for using the grant in violation of title IV-A of the
Act, as determined by findings from a single State audit and equal to
the amount of the misused funds;
    (2) An additional penalty of five percent of the adjusted SFAG,
based on our determination that such misuse was intentional;
    (3) A penalty of four percent of the adjusted SFAG for the failure
to submit an accurate, complete and timely required report;
    (4) A penalty of up to 21 percent of the adjusted SFAG for the
failure to satisfy the minimum participation rates;
    (5) A penalty of no more than two percent of the adjusted SFAG for
the failure to participate in the Income and Eligibility Verification
System (IEVS);
    (6) A penalty of no more than five percent of the adjusted SFAG for
the failure to enforce penalties on recipients who are not cooperating
with the State Child Support Enforcement agency;
    (7) A penalty equal to the outstanding loan amount plus interest
for the failure to repay a Federal loan provided for under section 406;
    (8) A penalty equal to the amount by which qualified State
expenditures fail to meet the appropriate level of historic effort in
the operation of the TANF program;
    (9) A penalty of five percent of the adjusted SFAG for the failure
to comply with the five-year limit on Federal funding of assistance;
    (10) A penalty equal to the amount of contingency funds that were
received for a fiscal year, but were not remitted by a State, if the
State failed to maintain 100 percent of historic effort in the
operation of its TANF program in that year;
    (11) A penalty of no more than five percent of the adjusted SFAG
for the failure to maintain assistance to an adult single custodial
parent who cannot obtain child care for a child under age six;
    (12) A penalty of no more than two percent of the adjusted SFAG,
plus the amount a State has failed to expend of its own funds, to
replace the reduction to its SFAG due to the assessment of penalties
under Sec. 262.1 in the fiscal year that immediately succeeds the year
in which the reduction was made;
    (13) A penalty equal to the amount of the State's Welfare-to-Work
formula grant for failure to maintain the required historic effort
during a year in which a State receives this formula grant; and
    (14) A penalty of not less than one percent and not more than five
percent of the adjusted SFAG for failure to impose penalties properly
against individuals who refuse to engage in required work in accordance
with section 407 of the Act.
    If applicable, in calculating the amount of the penalty, we will
use the adjusted SFAG as defined in Sec. 260.30. Except for the penalty
at Sec. 262.1(a)(12), all penalties are either a percentage of the
adjusted SFAG or a fixed amount. In calculating the amount of these
penalties, we will add all applicable penalty percentages together, and
we will apply the total percentage reduction to the amount of the
adjusted SFAG that would have been payable if we had assessed no
penalties against the State. As a final step, we will subtract other
(fixed) penalty amounts.
    The penalty at Sec. 262.1(a)(12) requires that we reduce a State's
adjusted SFAG if, in the fiscal year immediately following the fiscal
year when we have taken a penalty under this section, a State does not
expend its own funds on the State's TANF program in the amount of the
penalty (i.e., the amount by which we reduced the adjusted SFAG).
Unlike the other penalties, this penalty represents both a percentage
of the adjusted SFAG (up to two percent) and a fixed amount (the amount
of the reduction a State has failed to expend replace with its own
funds). We believe it is appropriate to calculate the amount of this
penalty by including the amount of the penalty based on a percentage
with other applicable penalty percentages. We will then subtract the
fixed amount of this penalty with the other fixed-amount penalties.
Finally, we will add the amount based on the percentage for this
penalty and the fixed amount for this penalty to determine the total
amount of this penalty.
    We will not reduce a State's quarterly grant by more than 25
percent. If the 25-percent cap prevents us from recovering the full
penalty imposed on a State all at once, we will apply the remaining
amount to the SFAG payable for the immediately succeeding quarters
until we have finally taken the penalty in full.
    In preparing this final document, we noticed a few places where we
should revise the regulatory text to be clearer.
    <bullet> In both the preamble discussion and the regulations of the
NPRM, we may not have described the Contingency Fund MOE penalty and
the penalty for failure to replace penalty amounts clearly enough.
Accordingly, we have clarified the regulation at Sec. 262.1(a)(10) to
say that we may penalize a State for failure to remit contingency funds
if it does not incur State TANF expenditures (i.e., State expenditures
within its TANF program) equal to at least 100 percent of its
historical State expenditures. In determining Contingency Fund MOE
requirements, historical State expenditures do not include expenditures
under the IV-A child care programs.
    <bullet> At Sec. 262.1(a)(12), we have clarified that States must
replace penalty amounts in the year after we actually take the
penalties.
    <bullet> At Sec. 262.1(a)(2), we have clarified that the penalty
for intentional misuse is in addition to the penalty for misuse.
    We received some comments on the provisions in this section and
have made a few changes to the regulations, as noted in our responses
to the comments below.
    Comment: Some commenters expressed the view that the regulations
placed too much emphasis on penalties and included too many penalties.
Another commenter mentioned that these provisions will lead to an
adversarial relationship reminiscent of the one that previously
surrounded quality control penalties under AFDC.
    Response: The statute mandates all of the penalties included in
these regulations. As we mentioned in the NPRM, it is clear that
Congress intended for State flexibility to be balanced with State
accountability. To assure that States fulfilled their new
responsibilities under the TANF program, Congress established a number
of penalties and requirements under section 409(a). The penalties
indicate the areas of State performance that Congress found most
significant and for which it gave us clear enforcement authority. While
we want to maintain supportive partnerships with States, we cannot
avoid our responsibilities under the statute. Although the regulation
may seem unduly slanted toward penalties,

[[Page 17799]]

this is because we have limited authority to regulate outside the
penalty provisions. Most of program policy and design is up to the
States and is not the subject of regulations.
    Comment: One commenter asserted that only one penalty, the one that
will be imposed if a State fails to maintain assistance to an adult
single custodial parent who cannot obtain child care for a child under
age six, focuses on protecting and serving families and children.
    Response: We do not agree with this observation. All of the
penalties have been enacted to assure that States operate programs that
promote the goals of the legislation. Many are designed to ensure that
States use Federal and State funds appropriately to provide assistance
to needy families and end dependence by promoting work and self-
sufficiency. Even the penalty for failure to submit an accurate,
complete and timely report supports program goals in that it requires
States to submit information about what is happening to needy families
and whether specific requirements are being met. Also, as we have said
elsewhere in this preamble, the penalty system is part of a much
broader structure that helps to protect families and promotes positive
State responses to the opportunities under TANF.
    Comment: A few commenters pointed out that some of the penalties
are inter-related and can have an escalating impact on States, i.e., if
a State fails one provision, it is likely to fail one or two others. A
commenter suggested that instead of imposing penalties and requiring
States to replace funds lost due to penalties, we should require States
to reduce claims for disallowed costs. Another argued that States
should reinvest penalty amounts since withholding funds may have the
effect of making it more difficult for the States to achieve the goals
of the program.
    Response: In establishing this new block grant program, Congress
wanted to give States flexibility to design programs that would best
serve their families. It enacted the penalty provisions in order to
assure that States use funds to achieve TANF program goals. The law
requires States to replace penalty amounts with their own funds so that
they will continue to serve needy families and meet the requirements of
the Act. Congress also enacted a maximum on the total penalty amount
that can be taken in any year in order to protect the interests of
needy families and children in the State.
    Comment: A commenter suggested that we should not design a system
that perpetuates failure based upon failure, but, instead, we should
design a system that rewards States for excellence.
    Response: Although it felt the penalties were necessary to focus
State performance, Congress did not rely solely on penalties to ensure
that States work towards achieving program goals. As we previously
discussed, it also enacted provisions to reward States for excellence
when it established bonuses for high performance and for decreases in
out-of-wedlock births.
    Comment: One commenter noted that the statute specifies that the
penalty for failure to meet the basic MOE requirement applies for
fiscal years 1998 through 2003 and suggested that we include this limit
in our regulations.
    Response: Since the TANF program is currently funded only through
fiscal year 2002, we did not think it was necessary to include this
limitation in our regulations. When Congress re-authorizes TANF, it
could well extend this provision in the statute. As the rules are
written, we would not need to reissue regulations to keep them current.
If the provision were not extended, the penalty would no longer be in
effect, and we would consider making conforming changes to the rules.
    Comment: A commenter asked if there is a penalty that applies when
a State fails to screen applicants and recipients and thus fails to
deny assistance to fleeing felons, or parole or probation violators.
    Response: The statute, at section 408(a)(9), prohibits States from
using their Federal TANF funds to provide assistance to fugitive felons
and probation and parole violators. While there is no specific penalty
covering this prohibition, the penalty for misuse (or intentional
misuse) of funds will apply if States provide TANF assistance to such
individuals.
    Comment: In the NPRM, we based penalties on the amount of the SFAG
minus any reductions due to the implementation of a Tribal TANF
program, without consideration of any transfers of funds to the
Discretionary Fund of the Child Care and Development Fund (CCDF) and/or
the Social Services Block Grant (SSBG). While one commenter expressed
appreciation for the fact that we assessed penalties against the
adjusted SFAG, other commenters asked that, for the sake of consistency
and fairness (since we subtracted transferred amounts before applying
the administrative cost cap), we should consider transfers of funds to
the CCDF and/or the SSBG in determining the adjusted SFAG.
    Response: As we discuss elsewhere, we have revised the definition
of the adjusted SFAG to remove any funds transferred to the
Discretionary Fund of the CCDF and/or the SSBG. The adjusted SFAG will
be the same as the SFAG for States without Tribal grantees and with no
transfers of funds to the Discretionary Fund of the CCDF or the SSBG.
You can find additional discussion of this issue in the preamble
discussions for Secs. 260.30 and 263.0.
    Comment: We received some comments about our interpretation of the
statutory language that requires penalties to be imposed ``for the
immediately succeeding fiscal year'' or the ``immediately succeeding
fiscal quarter.'' Commenters pointed out that we did not follow the
statute precisely, but did not express opposition to our
interpretation.
    Response: We are applying penalties for the fiscal year (or
quarter) immediately following our final decision in order to establish
a practical method for implementing the statute. This method allows us
to give States the opportunity to plead reasonable cause and to correct
violations under corrective compliance plans before we actually take a
penalty. Consequently, as one commenter noted, it is possible that a
State might incur a violation in FY 1998, be determined to be subject
to a penalty in FY 1999, and actually have its funding reduced in FY
2000.
    Comment: A commenter pointed out that, rather than limiting penalty
reductions to a State's grant to 25 percent during a fiscal year, the
statute prohibits us from reducing any quarterly payment by more than
25 percent. Another commenter asked us to clarify this provision.
    Response: The commenter is correct that the statute does not permit
us to reduce any quarterly payment by more than 25 percent. While on an
annual basis, capping each quarter's reduction at 25 percent would be
the same as capping the annual reduction at 25 percent, there could be
a difference when penalty reductions begin mid-year, as provided under
Sec. 262.1(c)(1). We have modified the language at Sec. 262.1(d)
slightly to clarify that we will not withhold more than 25 percent of a
State's quarterly grant.
    Comment: A commenter asked that we assess the penalties in four
equal quarterly installments during the year.
    Response: The statute requires us to take some penalties by
reducing the SFAG payable for the quarter that immediately follows our
final decision. In these cases, if the amount exceeds 25 percent of the
SFAG payable for that quarter, we will take the remaining amount from
the next quarter's SFAG.

[[Page 17800]]

    The statute requires us to take the majority of penalties by
reducing the SFAG payable for the fiscal year that immediately follows
our decision. In these cases, if taking the penalty in a single quarter
would have an adverse impact on the State's ability to administer the
TANF program, the State may ask that we take the penalty in two, three,
or four quarterly installments in the fiscal year. However, we must
take the full amount during that fiscal year unless we are prevented
from doing so by the 25-percent cap. Also, we would take a minimum of
the pro-rata share of the penalty amount from each quarter's grant; in
other words, we would not allow States to defer a disproportionate
share of the penalty amount to the latter part of the fiscal year.
    Comment: A couple of commenters noted that, in paragraphs
272.1(c)(1) and (c)(2), we incorrectly categorized when we would take
two of the penalties.
    Response: The commenters are correct. We made errors in listing
when we would take the penalties for failure to repay a Federal loan or
to enforce child support penalties. We have corrected paragraphs
Sec. 262.1(c)(1) and (c)(2) of the regulations to indicate that we will
take penalties for failure to repay a Federal loan by reducing the SFAG
payable for the quarter that immediately follows our final decision and
penalties for failure to enforce child support penalties by reducing
the SFAG payable for the fiscal year that immediately follows our final
decision.
    Comment: Another commenter argued that, if we take penalties in the
quarter following our final decision, it will be difficult for States
to fill in with their own funds.
    Response: The statute requires us to take any penalties for misuse
of funds and failure to repay a Federal loan by reducing the SFAG
payable for the quarter that immediately follows our decision.
Generally, however, States will have an early indication that these
penalties are likely to occur and will be able to plan accordingly.

Section 262.2--When Do the TANF Penalty Provisions Apply? (Sec. 272.2
of the NPRM)

    Congress recognized that, in certain circumstances, States should
face the consequences for failing to meet the requirements of the
penalty provisions from the first day the State operates the TANF
program. It also recognized, however, that States needed some lead time
in implementing other TANF requirements.
    Section 116(a)(2) of PRWORA delayed the effective date of some of
the penalty provisions in title IV-A. For those provisions where it did
not delay the effective date, we believe that Congress intended that a
State could be subject to a penalty from the first day it began to
operate TANF.
    During the interim period between publication of the NPRM and the
effective date of final rules, we required States to implement the TANF
provisions in accordance with their own reasonable interpretations of
the statute. In the NPRM we stated that we would not impose a penalty
if we were to find that a State's actions were inconsistent with the
final regulations, but consistent with a reasonable interpretation of
the statute. However, if we were to find that a State operated its TANF
program in a manner that was not based on a reasonable interpretation
of the statute, we would penalize the State.
    We received a few comments in support of these provisions and a
couple of other comments as discussed below. We made no changes to this
section of the regulations.
    Comment: In addition to the supportive comments, one commenter
expressed the view that the penalties should not apply until the final
regulations are adopted and the States have a reasonable period of time
to adjust to the new provisions. Another commenter asked that we give
States a hold-harmless period and not subject them to penalties while
they implement the regulations.
    Response: We have followed the statutory requirements for
determining when penalties apply. We do not have the authority to delay
the penalties. However, as we discussed in the preamble to Sec. 260.40,
prior to the effective date of these final regulations, we will not
penalize States if they operated their TANF programs in a manner that
is consistent with a reasonable interpretation of the statute. Also, we
decided to delay the effective date of these rules so that States have
a reasonable period of time to implement the new regulatory provisions.
Please refer to Sec. 260.40 for additional discussion of issues related
to the effective date.

Section 262.3--How Will We Determine if a State Is Subject to a
Penalty? (Sec. 272.3 of the NPRM)

    We have concluded that no one method can be used for monitoring
State performance. The following discussion explains the three
methods--the single audit, data collection and reporting, and financial
reporting--that we will use to determine State noncompliance with
requirements that may lead to penalties.
Single Audit
    Under the requirements of the Single Audit Act, as of July 1, 1996,
States operating Federal grant programs meeting a monetary threshold of
$300,000 must conduct an audit under the Act. Most States must audit
annually; a few may audit biennially. Because of the substantial
funding under TANF, all TANF States meet the audit threshold.
    The single audit is an organization-wide audit that reviews State
performance in many program areas. We will implement the Single Audit
Act through use of Office of Management and Budget (OMB) Circular A-
133, ``Audits of States, Local Governments, and Non-Profit
Organizations.'' Because of amendments to the Act in 1996, OMB recently
revised the Circular, merging former Circulars A-128 and A-133. It
published the new Circular in the Federal Register on June 30, 1997, at
62 FR 35277.
    In conducting their audits, auditors use a variety of tools,
including the statute and regulations for each program and a compliance
supplement issued by OMB. This supplement focuses on certain areas of
primary concern to that program. We prepared, and OMB has issued, a
TANF program compliance supplement for those penalties for which the
single audit will be our primary or secondary compliance instrument. We
will update the compliance supplement based on these final regulations.
    The Single Audit Act does not preclude us or other Federal offices
or agencies, such as the Office of the Inspector General (OIG), from
conducting additional audits or reviews. In fact, there is specific
statutory authority to conduct such additional audits or reviews. In
particular, 31 U.S.C. 7503(b) states:

    Notwithstanding subsection (a), a Federal agency may conduct, or
arrange for additional audits that are necessary to carry out its
responsibilities under Federal law or regulation. The provisions of
this chapter do not authorize any non-Federal entity (or sub-
recipient thereof) to constrain, in any manner, such agency from
carrying out or arranging for such additional audits, except that
the Federal agency shall plan such audits to not be duplicative of
audits of Federal awards.

    Additionally, we will conduct quality control reviews of selected
State audits to determine whether States conducted their audits in
accordance with the Single Audit Act, OMB Circular A-133, and the
compliance supplement. Pursuant to OMB Circular A-133, sections
____.400(a)(3) and (5), we will

[[Page 17801]]

take appropriate action when we find any audits to be deficient.
    We will use the single audit, in conjunction with other reviews,
audits, and data sources, as appropriate, to identify noncompliance for
which the State may be liable. We will rely heavily on Single Audit Act
activities for determining a State's liability for some penalties and
will use the single audit to gather and verify information for other
penalties. For example, we will use the single audit, supplemented by
other reviews, audits, and activities under the Single Audit Act, to
identify situations where a State used funds under section 403 in
violation of the Act. (See Sec. 263.10 on Misuse of Funds.) The misuse-
of-funds penalty is the only penalty for which the statute identifies a
specific method (i.e., the Single Audit Act) for determining penalty
liability.
    We will supplement information from the single audit with our own
audits and reviews, and reviews and audits conducted by OIG and its
contractors. We may identify a need to conduct such audits as the
result of complaints from individuals and organizations, requests by
the Congress to review particular areas of interest, information
collected by our reporting systems, or other indications of problems in
State compliance with TANF program requirements.
    When we determine that a State is subject to a penalty for the
misuse of funds, we may apply a second penalty if we determine that the
State intentionally misused Federal TANF funds. (You will find the
criteria for determining ``intentional misuse'' at Sec. 263.12.) The
single audit will be the primary vehicle for this penalty because of
its link to the determination of misuse of funds.
    The single audit will also help us identify noncompliance that
could result in imposition of the following four penalties: (1) Failure
to participate in the Income and Eligibility Verification System (see
Sec. 264.11); (2) Failure to comply with paternity establishment and
child support enforcement requirements under title IV-D of the Act (see
Sec. 264.31); (3) failure to maintain assistance to an adult single
custodial parent who cannot obtain child care for a child under age six
(see Sec. 261.57); and (4) failure to sanction recipients who refuse to
work (see Sec. 261.54). For these process-focused penalties, we
determined that we can make appropriate use of the single audit,
supplemented by other reviews and audits, to monitor State compliance.
    The audit compliance supplement includes guidance to auditors on
how to audit these areas. As in the case of the misuse-of-funds
penalty, we may conduct other reviews and audits, if necessary. For
example, we anticipate that we may receive complaints from individuals
and organizations concerning the penalty for a State's failure to
maintain assistance to an adult single custodial parent who cannot
obtain child care. A number of substantiated complaints might indicate
that we need to conduct an additional review.
    The single audit might identify a lack of State compliance in other
penalty areas, e.g., the five-year limit on Federal assistance. If it
does, we will not ignore those findings. Therefore, we will also impose
a penalty based on single audit findings in other penalty areas.
    For most programs, other than TANF, the Single Audit Act procedures
provide for disallowance in cases of substantiated monetary findings.
However, in accordance with section 409(a), under TANF, we will be
taking penalties, rather than disallowances. When the single audit
determines a specific violation, the penalty amount that we will apply
is the penalty amount associated with the specific penalty provision or
provisions, for example, misuse of funds and failure to end Federal
assistance after 60 months of receipt. Likewise, where we, or OIG,
conduct an audit or review, the penalty amount that will apply is the
penalty amount associated with the specific penalty or penalties
specified under section 409 and these rules.
Data Collection and Reporting
    We will monitor State compliance with the penalties for failure to
satisfy minimum participation rates (see Sec. 261.21) and failure to
comply with the five-year limit on Federal assistance (see Sec. 264.1)
primarily through the information required to be reported by section
411(a) (i.e., State reporting of disaggregated case-record
information). (See part 265 and the Appendices for data collection and
reporting requirements.)
    We believe that Congress intended that the data elements in section
411(a) be used to gather information for these two penalty areas. Thus,
we concluded that the section 411(a) data collection tools would be our
primary means for determining these penalties. We may also need to
conduct reviews in the future to verify the data submitted by States,
particularly in these two areas where a fiscal penalty is applicable.
States should maintain records to adequately support any report in
accordance with 45 CFR 92.42.
    Accurate data are essential if we are to apply penalties fairly. If
the State submits insufficient data to verify its compliance with the
requirements, or if we determine that a State cannot adequately
document the data that it has submitted to show that it has met its
participation rates or the five-year time limit, we will enforce the
participation rate penalty or five-year time-limit penalty.
    In the consultations we held during the development of the NPRM,
some participants recommended that we use the single audit as the means
for determining all the penalties. However, since States must otherwise
report the data that directly speak to their compliance in these two
areas, and timely determination of State compliance is necessary, we
did not accept that recommendation. Instead, we will rely on the
quarterly reports required under part 265 of these regulations.
Financial Reporting
    All States are subject to the basic MOE penalty for failure to
maintain a certain level (i.e., 80 or 75 percent) of historic effort.
Those States that choose to receive contingency funds under section
403(b) are subject to a separate maintenance-of-effort penalty for
failure to maintain 100 percent of historic effort. Also, in a year
that they receive WtW formula grants, States are subject to an
additional penalty for failure to meet the basic MOE requirement.
    We have developed a TANF Financial Report (see Appendix D of part
265). We designed this report to gather information required under
sections 403(b)(4), 405(c)(1), 409(a)(1), 409(a)(7), 409(a)(10),
409(a)(12), 409(a)(13), 411(a)(2), 411(a)(3), 411(a)(5), including data
on administrative costs and types of State expenditures. It will also
gather financial information to enable us to award grant funds, close
out accounts, and manage other financial aspects of the TANF program.
In addition, we will use this report to monitor State compliance with
the basic MOE and Contingency Fund requirements and to aid us in
determining if Federal TANF funds have been used properly.
    Consistent with section 5506(a) of Pub. L. 105-33, the TANF
Financial Report is due 45 days after the end of each quarter. Upon
receipt of the report for the fourth quarter, i.e., by November 14, we
should have State-reported information indicating whether or not the
State met its MOE requirements for the prior fiscal year.
    On the TANF Financial Report, States will inform us of the amount
of expenditures they have made for basic and Contingency Fund MOE
purposes. For the basic MOE, States must inform

[[Page 17802]]

us of the amount of expenditures made in the State TANF program and in
separate State programs. (See part 264, subpart B, for more information
on the Contingency Fund MOE requirement.)
    In addition, to collect the necessary information on all MOE
programs--both those operated within the TANF program and separate
State programs--we require supplemental information in an annual
report. The annual report, which may be provided as a separate report
or as an addendum to the fourth quarter TANF Data Report, requires that
States submit for each program for which the State claims MOE
expenditures, the total annual State expenditures and the total annual
State expenditures claimed as MOE. (See Sec. 265.9(c) for more
information on the contents of the annual report.)
    If we reduce a State's SFAG as the result of a penalty, the State
must expend an equal amount of its own funds in the immediately
succeeding fiscal year. If the State fails to replace the funds as
required, the State is subject to the penalty at Sec. 262.1(a)(12). The
penalty amount is up to two percent of the adjusted SFAG plus the
amount not expended to replace the reduction to the SFAG due to the
penalty.
    We will use the TANF Financial Report (or Territorial Financial
Report) to determine if a State has complied with these replacement
provisions. Instructions to the TANF Financial Report (see Appendix D)
require States to include amounts that they are required to contribute
as a result of any penalties taken against the State. (We will include
a similar requirement in the Territorial Financial Report.)
    As in the case of the penalties for failure to meet the
participation rates or comply with the five-year limit on assistance,
our program management responsibilities may require us to verify the
data submitted by States on the TANF Financial Report and annual
report, particularly data on MOE expenditures and ``replacement
funds.'' States should maintain records in accordance with 45 CFR
92.42. We will also use the annual report to help us to determine
whether a State met its MOE requirements.
    If the State submits insufficient MOE data to verify its compliance
or if we determine that the State cannot adequately document data that
it has submitted showing that it has met its MOE requirements, we will
apply the penalties for failure to meet the basic MOE requirements
(including the penalty related to WtW funding) and the Contingency Fund
MOE requirements. For the basic MOE, we may have to estimate the actual
level of qualifying MOE expenditures. We would then base the amount of
the penalty on the degree to which the State has not adequately
demonstrated that it has met the applicable MOE requirement.
    We will penalize States for failing to repay a loan provided under
section 406 (see Sec. 264.40). A specific vehicle for determining a
State's compliance with these requirements is unnecessary. In our loan
agreements with States, we will specify due dates for the repayment of
the loans, and we will know if States are not making the required
payments.
    We will penalize States for failing to submit a report required
under section 411(a) by the established due dates (see Secs. 265.4 and
265.7). As noted before, we are requiring that the reports must not
only be timely, but they must also be complete and accurate. Thus, we
may take actions to review the accuracy of data reporting if
appropriate. If we determine that the data required under section
411(a) are incomplete or inaccurate, we may apply the penalty for
failing to submit a report. As discussed above, if the data that are
inaccurate or incomplete pertain to other penalties (i.e., the
participation rate, the five-year time limit on assistance, the basic
MOE, the WtW MOE penalty, or the Contingency Fund MOE requirements) and
their unavailability impedes our ability to determine a State's penalty
liability, we will apply the penalties associated with these
requirements in lieu of a reporting penalty.
    Regardless of how we determine that a State is subject to a
penalty, the determination of whether a State has access to a possible
reasonable cause exception or corrective compliance depends on the
specific penalty provision. States cannot avoid all penalties through
the reasonable cause exception or a corrective compliance plan (see
Sec. 262.4).
    We received a few comments on this section and made some changes to
the regulations in response. A discussion of the comments and responses
follows.
    Also, in preparing the final rules, we noticed that we did not
discuss how we will determine that a State is subject to a Welfare-to-
Work formula grant penalty. We have added this discussion to the final
rule.
    Comment: Commenters suggested that we should permit the use of the
single audit for uncovering noncompliance with additional requirements
beyond those we identified in the NPRM.
    Response: Although we discussed the direct and indirect uses of the
single audit in determining compliance with a number of requirements,
we wrote the regulation itself more narrowly. We agree with these
comments and have revised the regulation at Sec. 262.3(a) to indicate
that, in addition to using the single audit as the primary method to
determine if a State is subject to certain penalties, we will use the
single audit, as appropriate, as a secondary method of determining if a
State is subject to other penalties.
    Comment: A commenter noted that, in paragraph (c), our reference to
Sec. 275.6 of the NPRM was incorrect.
    Response: We agree and have corrected Sec. 262.3(c) to refer to
verification of data in accordance with the provisions of Sec. 265.7 of
this chapter.
    Comment: Commenters suggested that our standards for determining
penalties are vague. Other commenters asked what we mean when we say
that information in the data or financial reports is ``insufficient.''
    Response: In some cases, our standards are specific, such as for
determining work participation rates and compliance with time limits.
However, we find that, given that this is a new program, it is
impossible to draw sharp lines that fully define all situations, and it
is appropriate to leave room for discretion in a block grant
environment. Moreover, since States can dispute our determinations and
have appeal rights, they have protection from arbitrary decisions.
    Obviously, we want strong, clear standards for ``complete and
accurate'' because the information reported by States in their data and
financial reports is critical in determining States' compliance with
TANF requirements and their potential penalty liability. However, our
standards have to be fair at the same time.
    In the preamble to part 265, you will find a broader discussion of
the importance of accurate, complete and timely reporting of
information.

Section 262.4--What Happens if We Determine That a State Is Subject to
a Penalty? (Sec. 272.4 of the NPRM)

    If we determine that a State is subject to a penalty, we will send
the State agency a notice that it has failed to meet a requirement
under section 409(a). This notice will: (1) Specify the penalty
provision at issue, including the applicable penalty amount; (2)
specify our source of information and the reasons for our decision; (3)
invite the State to present its arguments if it believes that the
information or method we used were in error or were insufficient, or
that its actions, in the absence of Federal regulations, were based on
a reasonable interpretation of the statute; and (4) explain if, how,
and

[[Page 17803]]

when the State may submit a reasonable cause justification under 409(b)
and/or corrective compliance plan under 409(c). States must postmark
their responses to our notice within 60 days of their receipt of our
notice.
    For penalties where the reasonable cause and the corrective
compliance plan provisions both apply, we encourage States to submit to
us both their justification for reasonable cause and a corrective
compliance plan within 60 days of receipt of our notice of failure to
comply with a requirement. Our objective is to expedite the resolution
of a State's failure to meet a requirement.
    A State may choose to submit a reasonable cause justification
without a corrective compliance plan. In this case, we will notify the
State if we do not accept the State's justification of reasonable
cause. Our notice will also inform the State that it has an opportunity
to submit a corrective compliance plan. The State will then have 60
days from the date it receives this notice to submit a corrective
compliance plan. (Under this scenario, we will send the State two
notices--the first will inform the State that it may be subject to a
penalty, and the second will inform the State that we determined that
it did not have reasonable cause.) We have added a provision to the
regulations to clarify this process. A State may also choose to submit
only a corrective compliance plan if it believes that the reasonable
cause factors do not apply in a particular case.
    The reasonable cause and corrective compliance provisions in the
statute do not apply to five penalties: (1) failure to repay a Federal
loan on a timely basis; (2) failure to maintain the applicable
percentage of historic State expenditures for the basic MOE
requirement; (3) failure to maintain 100 percent of historic State
expenditures for States receiving contingency funds; (4) failure to
expend additional State funds to replace grant reductions due to the
imposition of one or more penalties listed in Sec. 262.1; and (5)
failure to maintain 80, or 75, percent, as appropriate, of historic
State expenditures during a year in which the State receives a Welfare-
to-Work grant.
    If, upon review of the State's submittal, we request additional
information in order to determine reasonable cause, the State must
provide this information within 30 days of the date of our request. We
have established this deadline to make sure the process is not delayed.
However, under unusual circumstances we may give the State an extension
of the time to respond to our request for additional information.
    We received some comments on this section. One expressed the view
that our notification provisions were reasonable; others raised issues
about the proposed rule. Below, we address the comments and resulting
changes we made to the regulations. In addition to these changes, we
reversed the order of sub-paragraphs (e) and (f) so that they follow
the logical sequence of actions in the penalty process.
    Comment: A commenter recommended that we send our notice that a
State is subject to a penalty to the State agency director.
    Response: In the NPRM we said we would notify the State. By State,
we meant the State agency. We assume the commenter thought we would
notify the Governor. We have modified the regulation to say that we
will notify the State agency and added a definition of State agency to
Sec. 260.30.
    Comment: One commenter suggested that we list in the regulation the
four components of the initial penalty notice to the State that we
included in the preamble to the NPRM.
    Response: We agree with this suggestion and have amended the
regulation at Sec. 262.4(a) accordingly.
    Comment: Another commenter asked that we include, in our penalty
notice to a State, a description of the data and method we used to
determine that the State is subject to a penalty.
    Response: We thought that we covered this in the NPRM when we said
we would specify which penalty we would impose and the reasons for the
penalty. However, in the final rule at Sec. 262.4(a), we have revised
the language to list the source of information as one of the four
specific components that we will include in our notice to the State.
    Comment: A commenter asserted that States should be able to raise
any relevant issue in response to a penalty notice and not be limited
to responding on the three grounds of incorrect penalty determination,
reasonable cause, or corrective compliance.
    Response: Unfortunately, the commenter did not include any examples
of issues that would not fit in these three categories. We think that a
State will be able to include all relevant considerations under one of
these three categories.
    Comment: A few commenters noted that the regulation proposed at
Sec. 272.4(d) conflicted with Sec. 271.55(c), which said that
reasonable cause and corrective compliance were not available when a
State was being penalized for failing to impose penalties on
individuals.
    Response: The proposed regulation at Sec. 272.4(d) was correct.
Reasonable cause and corrective compliance are available when a State
is being penalized for failing to impose penalties on individuals. The
final rules at Secs. 262.4(d) and 261.55 reflect this policy.
    Comment: A commenter asked that we make reasonable cause and
corrective compliance available to States that are being penalized for
failing to expend additional State funds to replace penalty amounts.
    Response: We do not have the authority to make this change, since
the statute specifies that the reasonable cause exception and
corrective compliance plan do not apply to this penalty.
    Comment: Several commenters asked that we establish a time frame
for when we will respond to a State's reply to our penalty notice.
    Response: We have added a provision to the regulations at
Sec. 262.4(f) to say that, generally, we will respond within 60 days to
the State's reasonable cause submittal, and that we will either accept
or reject the State's corrective compliance plan within 60 days of our
receipt of the plan.
    Comment: A commenter asked whether a State may request
reconsideration or submit additional information based on our decision,
or whether its only recourse at that point is to file a formal appeal.
    Response: Although there are no further formal steps available to
the State short of a formal appeal, it is our hope that State and
Federal staff will engage in an ongoing dialogue in an effort to
address any penalty-related issue. This dialogue may begin as soon as
the State is working to develop new policies or begins to have trouble
meeting a requirement, and well before we notify the State that we
intend to penalize it. It may continue until the issue is resolved, but
will not extend the time frames States have for responding to our
notices. Therefore, we advise States to make their complete and best
arguments during the time allotted.
    Comment: A number of commenters asserted that two weeks is not long
enough for States to respond to our request for further information.
    Response: We agree that under some circumstances two weeks may not
be long enough, so we are increasing the time States have to respond to
30 days. Also, under unusual circumstances, we may give States an
extension of the time that they have to respond to our request for
information. We have amended the regulation at Sec. 262.4(e)
accordingly.
    Comment: Commenters suggested that notices and requests be sent by
certified mail so that there is evidence of receipt.

[[Page 17804]]

    Response: A State may choose to send its responses by certified
mail, but we are not convinced that we need to include this as a
regulatory requirement.
    Comment: A commenter asked that we specify in the regulations that
we would not assess any penalties pending the resolution of a State's
claim of reasonable cause.
    Response: If a State claims reasonable cause and we find against
the State, the State may then submit a corrective compliance plan or
file an appeal to the HHS Departmental Appeals Board (DAB), as
discussed in Sec. 262.7. If the State does not take either action, we
will assess the penalty in the quarter or fiscal year that immediately
follows our final decision, as appropriate. However, if the State
submits a corrective compliance plan, we will not assess a penalty
until the corrective compliance process is completed. If the State
appeals to the DAB, we will not assess the penalty until the appeals
process is completed. If the DAB upholds our decision, we will take the
penalty and charge interest back to the date of our final response that
formally notifies the Governor of the State of an adverse action.

Section 262.5--Under What General Circumstances Will We Determine That
a State Has Reasonable Cause? (Sec. 272.5 of the NPRM)

    Under the provisions of section 409, we will not impose certain of
the penalties if a State demonstrates that it had reasonable cause.
Also, we will reduce or excuse certain penalties if a State corrects or
discontinues the violations under an accepted corrective compliance
plan.
    After reviewing these statutory provisions, we decided that we
should not consider the reasonable cause exception of the statute in
isolation. Rather, we would view it in conjunction with the provision
for developing corrective compliance plans. In this context, we
acknowledge the new Federal and State roles under TANF and commit to
working with States to minimize adversarial Federal-State issues. Our
primary task is to help each State operate the most effective program
it can to meet the needs of its caseload and the goals and provisions
of the law. Through these rules, we hope to focus States on positive
steps that they should take to correct situations that resulted in a
determination that they are subject to a penalty, rather than to let
them simply avoid the penalty. As such, we consider it appropriate to
emphasize the use of the corrective compliance plan process over the
reasonable cause exception. Consequently, we have decided to limit the
list of reasonable cause criteria.
    In the discussion that follows, we describe: (1) the factors that
we will consider in deciding whether or not to excuse a penalty based
on a State's claim of reasonable cause; (2) the contents of an
acceptable corrective compliance plan; and (3) the process for applying
these provisions. Our goal is to treat the reasonable cause and
corrective compliance plan provisions as part of an integrated process.
    We have included factors that would be applicable to all penalties
for which the reasonable cause provision applies. We will find that a
State has reasonable cause under the following situations: (1) Natural
disasters and other calamities (e.g., hurricanes, tornadoes,
earthquakes, fires, floods, etc.) whose disruptive impact was so
significant as to cause the State's failure to meet a requirement; (2)
formally issued Federal guidance that provided incorrect information
resulting in the State's failure; and (3) isolated problems of minimal
impact that are not indicative of a systemic problem (e.g., although a
State's policies and procedures require that Federal TANF assistance be
time-limited to five years and include computer safeguards to protect
against violations, ten families somehow slip through and receive
assistance for longer than five years).
    We also have included two separate factors that would apply in
cases when the State fails to satisfy the minimum participation rates,
and one specific factor that would apply to cases when the State fails
to meet the five-year limit. We discuss these specific factors in our
preamble discussion of domestic violence and Secs. 261.52 and 264.3.
    As discussed elsewhere in this preamble, we have also added a
factor that will apply if States fail to meet either of the first two
deadlines for FY 2000 for submitting complete and accurate reports
under the new reporting requirements. We added this factor in response
to comments and out of our own concern about the possible concurrent
demands of Y2K and TANF reporting requirements. States must be in a
position to commit the systems resources necessary to become Y2K
compliant in order to ensure that there is no disruption in the
benefits to their neediest citizens.
    We did not have the latitude under the law merely to extend the
reporting deadlines (because they are set in statute). Also, we were
unwilling to extend the ``emergency reporting'' into FY 2000 and
provide a later effective date for the new reporting provisions because
important TANF provisions (e.g., the work participation rates) depend
upon consistent data and policies throughout the entire fiscal year.
Thus, we have addressed the concern as a reasonable cause issue.
    Under the new provision, States that miss the deadlines for
submitting complete and accurate data for the first two quarters of FY
2000 will receive reasonable cause if: (1) they can clearly demonstrate
that their failure was attributable to Y2K compliance activities; and
(2) they submit the required data by July 1, 2000.
    In determining reasonable cause under all of these regulatory
criteria, we will consider the efforts the State made to meet the
requirement. We will also take into consideration the duration and
severity of the circumstances that led to the State's failure to
achieve the requirement. The burden of proof rests with the State to
explain fully the circumstances, events, or occurrences that constitute
reasonable cause for its failure to meet a particular requirement. The
State must provide us with sufficient relevant information and
documentation to substantiate its claim of reasonable cause. We have
added a provision to the regulations to clarify the factors that we
will consider and the State's burden of proof. If we find that the
State has reasonable cause, we will not impose the penalty.
    We received quite a number of comments on this section. We discuss
the comments and the changes we made to the regulations below.
    Comment: Virtually all commenters with comments on this section
argued that our proposed list of reasonable cause factors was too
narrow and that we needed to give ourselves more discretion. Commenters
gave a number of examples of factors that we should consider, including
good faith effort, circumstances beyond the State's control, inadequate
Federal guidance, increases in a State's caseload, characteristics of
the caseload, high unemployment rates or other labor market
characteristics, changing economic conditions, and other adverse
economic factors.
    Response: As we noted in the NPRM, PRWORA did not specify any
definition of reasonable cause or indicate what factors we should use
in deciding whether to grant a reasonable cause exception for a
penalty. In our deliberations on reasonable cause factors, we
considered the diverse opinions expressed during our consultation
process and our NPRM comment period, as well as the need to support the
commitment of Congress, the Administration, and States to the work and
other objectives of the TANF program. In keeping with these objectives,
we are providing reasonable

[[Page 17805]]

cause factors for a limited number of circumstances that are beyond a
State's control and placing a greater emphasis on corrective solutions
for those circumstances a State can control. We strongly believe that
States must correct problems that detract from moving families from
welfare to self-sufficiency.
    At the same time, we agree with the commenters that it would be
difficult to foresee all possible circumstances under which we would
want to grant reasonable cause. Accordingly, while we have included the
same general factors that we included in the NPRM, we no longer limit
ourselves to considering only these factors. While we do not anticipate
routinely determining that a State had reasonable cause based on other
factors, we do not want to preclude a State from presenting other
circumstances. Also, we decided that we were more restrictive than we
intended when we limited the third reasonable cause factor to isolated,
nonrecurring problems. We have amended the regulations to say that we
may grant a State reasonable cause when it has encountered isolated
problems of minimal impact that are not indicative of a systemic
problem.
    Comment: A number of commenters were opposed to our provisions
precluding reasonable cause if a State diverted families to a separate
State program that achieved the effect of avoiding the work
participation rates or diverted the Federal share of child support
collections.
    Response: As we previously discussed in the section of the preamble
entitled ``Separate State Programs,'' we have eliminated the proposed
connection between a State's decisions on separate State programs and
its eligibility for reasonable cause. Therefore, we have deleted the
provisions that were at paragraphs Sec. 272.5(c) and (d) of the
proposed rule.
    Comment: A commenter suggested that the regulations should provide
that, as part of the process of determining when we would impose
penalties and penalty amounts, we should give consideration to factors
such as whether the State has administered its TANF program fairly,
whether it has provided services and supports to families to enable
them to comply with program requirements, and whether State-imposed
requirements on families are reasonable.
    Response: The TANF legislation assumed that States are in the best
position to determine which families will be served and what assistance
they will receive. As we previously discussed, our authority to
regulate and judge State policies and actions are limited, and we have
decided not to stretch our regulatory authority by incorporating such
factors into all our penalty determinations. There are other provisions
in the statute (such as the bonus and ranking provisions, the annual
reports to Congress, and annual reports on State child poverty rates)
that provide an opportunity to look at whether at-risk families are
being helped or hurt by State TANF programs. At the same time, there
are a couple of penalty provisions (e.g., those dealing with the
imposition of sanctions) where the issues of fairness and adequate
recipient protections are more germane and we specifically address some
of these issues. You should look to the preamble discussion entitled
``Worker and Recipient Protections'' and the preamble for part 261 for
other ways we are addressing this concern.

Section 262.6--What Happens if a State Does Not Demonstrate Reasonable
Cause? (Sec. 272.6 of the NPRM)

    Section 409(c), as amended by section 5506 of Pub. L. 105-33,
provides that, prior to imposing a penalty against a State, we will
notify the State of the violation and allow the State the opportunity
to enter into a corrective compliance plan. If a State does not claim
reasonable cause or if it claims reasonable cause simultaneously with
submitting a corrective compliance plan, it will have 60 days from the
date it receives our notice of a violation to submit its corrective
compliance plan. If, in response to our notice of a violation, the
State initially submits only a claim of reasonable cause, and if we
deny this claim, the State has 60 days from the date it receives our
second notice (i.e., denying its reasonable cause claim) to submit its
corrective compliance plan. If a State does not submit an acceptable
corrective compliance plan on time, we will immediately send the State
a formal notice of adverse action and assess the penalty. Outside of
the notice(s), we will not remind the State that the corrective
compliance plan is due.
    The corrective compliance plan must provide a complete analysis of
the situation and factors that prevented the State from meeting the
requirement. It also must identify the time period in which the State
will correct or discontinue the violation, and the milestones,
including interim process and outcome goals, the State will achieve to
assure that it will fully correct or discontinue the violation within
the specified time period. In order to highlight the importance of
corrective compliance, the plan must include a certification by the
Governor that the State is committed to correcting or discontinuing the
violation in accordance with the plan.
    We recognize that each plan must be specific to the violation (or
penalty) since each State operates its TANF program in a unique manner.
Thus, we will review each plan on a case-by-case basis. In determining
whether or not to accept a plan, we will consider the extent to which
the State's plan indicates that it will completely correct or
discontinue, as appropriate, the situation leading to the penalty.
    The steps that a State takes to correct or discontinue a violation
may vary. For example, where we penalize a State for misusing Federal
TANF funds, we would expect it to remove this expenditure from its TANF
accounting records (charging it to State funds, as allowable) and
provide steps to assure that such a problem does not recur. Where a
State has reduced or denied assistance improperly to a single custodial
parent who could not find child care for a child under six, correcting
the violation might require that the State reimburse parents
retroactively for the assistance that it improperly denied them. The
State's corrective compliance plan also would have to describe the
steps to be taken to prevent such problems in the future.
    Section 409(c)(3) requires that a violation be corrected or
discontinued, as appropriate, ``in a timely manner.'' A State's timely
correction of a problem is critical to assuring that the State is not
subject to a subsequent penalty. At the same time, we recognize that
the causes of violations will vary, and we cannot expect States to
rectify all violations in the same time frame. Thus, we do not want to
unduly restrict the duration of corrective compliance plans. At the
same time, we do not want to allow States to prolong the corrective
compliance process indefinitely and leave problems unresolved into
future fiscal years. Accordingly, in our NPRM, we proposed that the
period covered by a corrective compliance plan end no later than six
months after the date we accept a State's corrective compliance plan.
We have amended this provision, as discussed below.
    We will consult the State on any modifications to the corrective
compliance plan that we believe are necessary and seek mutual agreement
on a final plan. Such consultation will occur only during the 60-day
period for acceptance specified in the law. Any modifications to the
State's corrective compliance plan resulting from such consultations
will constitute the State's final corrective compliance plan and will
obligate the State to take the actions

[[Page 17806]]

and meet the time frames specified in the plan.
    We will either accept or reject the State's corrective compliance
plan, in writing, within the 60-day period that begins on the date that
we receive the plan. If a State does not agree to modify its plan as we
recommend, we may reject the plan. If we reject the plan, we will
immediately send a formal notice to the State of the adverse action.
The State may appeal our decision to impose the penalty in accordance
with the provisions of section 410 of the Act and the regulations at
Sec. 262.7.
    If we have not rejected a plan in writing by the end of the 60-day
period, the plan is deemed to be accepted, as required by the statute
at section 409(c)(1)(D).
    If a State corrects or discontinues the violation in accordance
with its corrective compliance plan, we will not impose the penalty.
    The statute permits us to collect some or all of the penalty if the
State has failed to correct or discontinue the violation. Therefore, we
may reduce the amount of the penalty if a State has not fully rectified
the violation in one or more of the following limited situations: (1)
The State made significant progress in correcting or discontinuing the
violation; or (2) a natural disaster or regional recession prevented
the State from coming into full compliance.
    We received a number of comments on these provisions that led us to
make some changes to the regulations. Also, we made some minor edits to
ensure consistency within the parts of this regulation. We discuss the
comments and changes below.
    Comment: In the NPRM, we asked for comments from States and other
interested parties on our proposal to restrict the time period for a
corrective compliance plan. Commenters supported the general concept of
a corrective compliance plan, and one commenter thought the six-month
period was reasonable for most cases. However, most commenters replied
that the period we had proposed was unreasonably short, especially
since the statute does not require a short time frame. Many suggested
that we extend the time period to 9, 12, or 24 months. Others suggested
that the State should determine the time frame, or that it be part of
the negotiation of the plan by the State and ACF and be determined on a
case-by-case basis. Another commenter suggested that the period extend
until 90 days after the close of the State's next legislative session.
Commenters argued the need for more time based on the possible need to
adjust contracts, re-design programs, change policies and procedures,
notify recipients, make data system changes, train staff, and get the
State legislature to take necessary action.
    Response: In responding to these comments, we want to reinforce the
importance of achieving compliance with the statute quickly, but we
also recognize that we need to consider a State's ability to make the
changes necessary to achieve compliance within a fixed time frame. We
are not interested in setting a time frame that States cannot meet, but
we also do not want to give States more time than they absolutely need.
In addition, in the case of the work participation rate and time-limit
penalties, where we measure performance over the course of a fiscal
year, we thought it was important that corrective compliance also be
measured over the course of a fiscal year. Based on this thinking, we
have revised the regulations. In general, the final rules provide more
flexibility in establishing time frames for corrective compliance
plans. For the work participation rate and time-limit penalties, they
incorporate a modified six-month corrective compliance period. More
specifically, they provide that the State achieve compliance for the
first fiscal year that ends at least six months after our receipt of
the corrective compliance plan. For example, if a State failed its work
participation rate in a prior fiscal year and we received its
corrective compliance plan on February 1, the State would have to
achieve the participation rates in effect for the current fiscal year.
If we received the plan after April 1, the State would have to achieve
the participation rates in effect for the following fiscal year.
    We made this adjustment to the rules in large part because we
calculate liability for work participation and time-limit penalties
based on fiscal year data. We also realized that there could be
significant delays in the submittal of corrective compliance plans
(because participation rate and time-limit information is not available
immediately, and we need time both to resolve disputes about the
penalty findings and to decide State claims for reasonable cause).
Thus, we could not necessarily expect a State to achieve compliance
during the first year following a failure.
    Nevertheless, we would hope that a State could achieve compliance
during that time frame. We would not want to see a State's failure
extend into a third fiscal year. If it did, there could be negative
consequences for the State. States especially need to work towards
increasing their work participation rates as quickly as possible
because: (1) the rates increase over time; (2) the base penalty amount
increases when a State incurs consecutive penalties; and (3) a State is
eligible for a smaller reduction based on degree of noncompliance if it
fails to meet the rates in successive fiscal years.
    For both the work participation and time-limit penalties, a State
will normally have indication that a problem exists during the year for
which it is penalty-liable, and it should begin to address the problems
well before it submits its corrective compliance plan. For example, by
July of a fiscal year, a State should have a good idea of whether it is
on track to meet its work participation requirements. If it is not, and
does not begin to make changes soon, not only will it fail to meet the
requirements for the current fiscal year, but it is unlikely that it
will be able to increase its performance enough to meet the required
rates for the next fiscal year. Our notice to a State that it is
subject to a penalty should serve as confirmation of information the
State already has. A corrective compliance plan period does not
necessarily have to be lengthy in order to provide the State sufficient
time for correcting or discontinuing a violation.
    For the remaining penalties that are eligible for corrective
compliance, we would permit a State to propose a time frame in its
corrective compliance plan. We would expect the State to achieve
compliance expeditiously, often in less than six months. States should
correct some failures, for example, for failing to comply with IEVS
requirements or submitting a data report late, within a month or two.
    We expect each State to justify its time frame for each penalty. We
will assess the time frame proposed by the State based on the nature of
the violation, any unusual circumstances, and other factors that affect
the speed with which the State can respond, such as whether it would
need to make systems changes or take legislative action.
    Comment: A commenter asked that we notify States of our acceptance
of a corrective compliance plan and asked us to clarify when the
corrective compliance period begins.
    Response: We did not address these factors in the NPRM, but have
revised the regulations to specify that we will accept or reject the
plan in writing and that the time period for the corrective compliance
plan begins on the date that the State receives our written acceptance
of the plan. If we fail to respond, the time period for the corrective
compliance plan begins on

[[Page 17807]]

the date that is 60 days after the date we received the State's plan.
    Comment: Several commenters stated that 60 days is insufficient for
a State to prepare a corrective compliance plan and recommended that we
give States 90 days.
    Response: We are prevented from making this change by the statute,
which specifies that a State has 60 days from the date that it receives
our notification to submit a corrective compliance plan.
    Comment: A couple of commenters noted that the proposed rules at
Sec. 272.6 contained an incorrect citation.
    Response: While the commenters were correct that the citation in
the NPRM was erroneous, we have made changes to the paragraphs in that
section that corrected that problem.
    Comment: Some commenters expressed the view that the contents we
specified for the corrective compliance plan are reasonable. Other
commenters objected to our requiring certification of the plan by the
Governor, and one commenter suggested that the certification be made by
the director of the State agency.
    Response: The Governor is responsible for submitting the State TANF
plan and for committing State funds to the program. On this basis, we
believe it is also important for the Governor to demonstrate awareness
of and support for the corrective compliance plan.
    Comment: A commenter asked that we consider a State's good faith
effort in determining the amount of a penalty when a State fails to
completely correct or discontinue the violation pursuant to its
corrective compliance plan. Other commenters asked that we broaden the
circumstances under which a penalty is reduced, with some recommending
that we consider other factors such as natural disasters, economic
circumstances, or other unanticipated or extreme events.
    Response: We have said that we will reduce the penalty if the State
can demonstrate that it made significant progress toward correcting or
discontinuing the violation or that its failure was due to a natural
disaster or regional recession. We believe this gives us sufficient
latitude to consider mitigating circumstances and the good-faith effort
a State has made. For a discussion of the specific standards we will
use in deciding to reduce work participation rate penalties, please see
the preamble for Sec. 261.51.
    Comment: A number of commenters opposed our provisions denying a
penalty reduction if a State diverted families to a separate State
program that achieved the effect of avoiding the work participation
rates or diverted the Federal share of child support collections.
    Response: As we previously discussed, we have eliminated the
connection between a State's decisions regarding its separate State
programs and penalty reductions and have removed the provisions that
appeared in Sec. 272.6(i)(2) of the NPRM.

Section 262.7--How Can a State Appeal our Decision To Take a Penalty?
(Sec. 272.7 of the NPRM)

    Once we make a final decision to impose a full or partial penalty,
we will formally notify the State that we will reduce the State's SFAG
payable for the quarter or the fiscal year and inform the State of its
right to appeal to the Departmental Appeals Board (the Board).
    Section 410, which covers any adverse actions with respect to the
State TANF plan or the imposition of a penalty under section 409,
provides that the Secretary will notify the Governor of the State of
the adverse action within five days. To facilitate the appeal, we will
also send a copy of the notice to the State agency.
    Within 60 days after the date a State receives this notice, the
State may file an appeal of the action, in whole or in part, with the
Board. We indicated in the NPRM that the statute allowed only 60 days
for the Board to reach a decision after the appeal is filed. A number
of commenters believed that the 60 days in the statute indicated a
minimum time before a decision could be issued, not a maximum time. The
NPRM interpretation was based on the conference report which indicated
a Board decision was required ``within 60 days'' (H.R. Rep. No. 725,
104th Cong., 2d sess., p. 302). However, in light of the comments, we
have re-examined the language of the statute itself, which states that
a decision will be made in ``not less than 60 days'' after the appeal
is filed. ``Not less than'' is usually interpreted as a minimum
requirement, as the commenters indicated. Therefore, we have revised
the regulation to allow a minimum time of 60 days before a decision is
made. Nevertheless, we believe that penalties procedures should be
handled as expeditiously as possible. We also believe that this is
possible in the TANF penalty situation because the opportunity for
reasonable cause and corrective compliance before most TANF penalties
should have clarified the issues before the penalty decision.
    We are requiring that the State submit its brief and the supporting
documentation for its case when it files its appeal. To further
facilitate this process, we have added a provision to the regulation at
Sec. 262.7(a)(1) that ACF's notice must include sufficient factual and
legal information on the basis for imposition of the penalty to allow
the State to respond in an appeal. In addition, we have allowed the
State the opportunity to respond to ACF's reply brief and to submit any
additional documentation it considers necessary. A State should send a
copy of any appeal documents to the Office of the General Counsel,
Children, Families and Aging Division, Room 411-D, 200 Independence
Avenue, S.W., Washington, D.C. 20201.
    In the final rule, we have slightly increased the time for us to
submit our reply brief and supporting documentation-- to 45 days after
our receipt of the State's submission. This 45 days, plus the 21 days
allowed for the State's reply brief, will ensure that the DAB makes no
determination prior to 60 days after a State has filed its appeal.
Further, briefing and argument will be at the discretion of the Board,
but could include an evidentiary hearing. A State's appeal to the Board
will also be subject to the following regulations at part 16 of title
45: Secs. 16.2, 16.9, 16.10, and 16.13-16.22, to the extent they are
consistent with this section.
    Section 410(b)(2) provides that the Board will consider an appeal
on the basis of the documentation the State submits, along with any
additional information required by the Board to support a final
decision. In deciding whether to uphold an adverse action or any
portion of such action, the Board will conduct a thorough review of the
issues.
    Finally, a State may obtain judicial review of a final decision by
the Board by filing an action within 90 days after the date of the
final decision. States may file either with the district court of the
United States in the judicial district where the State agency is
located or in the United States District Court for the District of
Columbia. The district courts will review the final decision of the
Board on the record established in the administrative proceeding, to
determine if it is arbitrary, capricious, an abuse of discretion or
otherwise not in accordance with law, or unsupported by substantial
evidence. The court's review will be on the basis of the documents and
supporting data submitted to the Board.
    We discuss below the comments on this section and our responses.
    Comment: A number of commenters believed the time period for the
appeal process was too constrained to allow

[[Page 17808]]

adequate consideration of the issues. These commenters noted that the
statute could be interpreted to require a minimum of 60 days before a
determination could be made, rather than the maximum the NPRM proposed.
    Response: For the reasons previously discussed, we agree with the
commenters and have revised the regulation accordingly.
    Comment: One State indicated that the notice should include details
on the reasons for the penalty.
    Response: We agree with the commenter that the notice should
contain sufficient detail on the factual and legal basis for the
penalty to allow the State to respond and have revised the regulation.
However, we believe the agency should have the opportunity to raise new
issues in response to the State's brief and therefore have not
specified that reasons not raised in the notice are waived. Since the
State now has an opportunity to respond to the ACF brief and to submit
additional documentation, we do not believe this policy will
disadvantage the State.
    Comment: One State noted that the practice of notifying the
Governor differed from past practice of notifying the agency and
suggested that we also notify the TANF agency.
    Response: Although the statute requires notice to the Governor, we
agree with the commenter that it would facilitate the process if we
also give the TANF agency a copy of the notice and have amended the
regulation accordingly.
    Comment: A number of commenters suggested the State should be able
to submit a reply brief as a matter of right. They also suggested the
Board's authority to develop the record be clarified.
    Response: The NPRM limited the State's right to submit a reply
brief as a matter of right because of the limited time availability
under the proposed 60-day maximum. Since we have eliminated this 60-day
time issue in the final regulation, we agree with the commenters that
the State should be able to submit an appeal as a matter of right and
have amended the regulation.
    We have also clarified that the Board's discretion to develop the
record included the discretion to hold an evidentiary hearing. We would
note that Secs. 16.9 and 16.10 of this title, which are made applicable
by Sec. 262.7(e), contain additional detail on the Board's discretion
to develop the record.
    Comment: One State expressed concerns about using the Departmental
Appeal Board as the forum for hearing appeals.
    Response: The statute specifies the Departmental Appeals Board as
the entity to hear appeals.
    Comment: One commenter believed that we should include all sections
of 45 C.F.R. part 16 as part of the appeal process. This commenter also
believed that we should not treat failure to file a copy of an appeal
with the Office of the General Counsel as a jurisdictional defect.
    Response: We selected the provisions of part 16 that fill in the
gaps in the TANF statutory framework. We have not added additional
sections because we do not think they are necessary.
    The failure to file the appeal with the Office of the General
Counsel is not a jurisdictional defect. However, we would toll the time
period for filing of our reply brief until OGC receives the brief.
    Comment: One State noted that the specific provision on when the
State's appeal is considered filed, at Sec. 272.7(f)(1) of the NPRM,
varied from the time contained in 45 C.F.R. 16.20, which we adopted in
the NPRM.
    Response: As part of the changes in the timing of an appeal in the
final rule, we have deleted this NPRM provision and thus eliminated the
conflict. However, we have also added a provision to Sec. 262.7(e) to
clarify that the named provisions of part 16 are adopted only to the
extent that they are consistent with the specific provisions of this
section.

VIII. Part 263--Expenditures of State and Federal TANF Funds (Part
273 of the NPRM)

Section 263.0--What Definitions Apply to This Part? (Sec. 273.0 of the
NPRM)

Administrative Costs
(a) Background
    Under the TANF statute, States may not spend more than 15 percent
of either their Federal TANF funds or their State MOE dollars on
administrative costs. At section 404(b), the statute excludes
expenditures for ``information technology and computerization needed
for tracking or monitoring'' from the administrative cost cap that
applies to Federal TANF funds (i.e., the Federal cap).
    The proposed rule addressed the subject of administrative costs in
five separate places: (1) the definition of qualified expenditures at
Sec. 270.30 provided that, for MOE purposes, administrative costs were
subject to a 15-percent cap (i.e., the MOE cap); (2) Sec. 273.0
provided a definition of administrative costs; (3) Sec. 273.2(a)(5)
discussed the 15-percent limit on the amount of MOE expenditures that
could be spent on administrative costs and reflected our decision to
exclude the same information technology and computerization costs from
the MOE cap as the Federal cap; (4) the preamble for Sec. 273.11
explained that we would consider expenditures of more than 15 percent
of a State's Federal TANF funds on administrative funds to be a misuse
of Federal TANF funds; and (5) the preamble and regulation at
Sec. 273.13 provided that, in determining the Federal cap, we would use
the definition of administrative costs at Sec. 273.0(b) and not count
information technology and computerization for tracking and monitoring
as administrative costs. The preamble for Sec. 273.13 also explained
that we would look to see whether a State's cumulative expenditures on
administrative costs from its grant for any fiscal year exceeded 15
percent of the grant amount and that we would consider expenditures
above the limit to be a misuse of funds.
    The proposed definition at Sec. 273.0(b) provided that:
``Administrative costs means costs necessary for the proper
administration of the TANF program or separate State programs. It
includes the costs for general administration and coordination of these
programs, including indirect (or overhead) costs.'' It also provided
examples of eleven types of activities that would be classified as
``administrative costs,'' such as salaries and benefits not associated
with providing program services, plan and budget preparation,
procurement, accounting, and payroll.
    In the preamble, we stated our belief that the proposed definition
would not create a significant new administrative burden on States. We
hoped that it was flexible enough to facilitate effective case
management, accommodate evolving TANF program designs, and support
innovation and diversity among State TANF programs. We also said that
it had the significant advantage of being closely related to the
definition in effect under the Job Training Partnership Act (JTPA).
Thus, it should facilitate the coordination of Welfare-to-Work and TANF
activities and support the transition of hard-to-employ TANF recipients
into the work force.
    More importantly to commenters, the preamble also indicated that we
would consider eligibility determinations to be administrative costs,
but allow case management to be treated as a program cost. It also
required that portions of a worker's time be allocated based on this
distinction. Specifically, the NPRM preamble said:


[[Page 17809]]


    You will note that the definition we have proposed does not
directly address case management or eligibility determination. We
understand that, in many instances, the same individuals may be
performing both activities. In such cases, to the extent that a
worker's activities are essentially administrative in nature (e.g.,
traditional eligibility determinations or verifications), the
portion of the worker's time spent on such activities will be
treated as administrative costs, along with any associated indirect
(or overhead) costs. However, to the extent that a worker's time is
essentially spent on case-management functions or delivering
services to clients, that portion of the worker's time can be
charged as program costs, along with associated indirect (or
overhead) costs.

    In the preamble, we also indicated that we expected administrative
costs incurred by subgrantees, contractors, community service
providers, and third parties to be part of the administrative cost cap
and that we would determine such costs in the same way as agency costs.
Specifically, we said:

    We have not included specific language in the proposed rule
about treatment of costs incurred by subgrantees, contractors,
community service providers, and other third parties. Neither the
statute nor the proposed regulations make any provision for special
treatment of such costs. Thus, the expectation is that
administrative costs incurred by these entities would be part of the
total administrative cost cap. In other words, it is irrelevant
whether costs are incurred by the TANF agency directly or by other
parties.
    We realize this policy may create additional administrative
burdens for the TANF agency and do not want to unnecessarily divert
resources to administrative activities. At the same time, we do not
want to distort agency incentives to contract for administrative or
program services. In seeking possible solutions for this problem, we
looked at the JTPA approach (which allows expenditures on services
that are available ``off-the-shelf'' to be treated entirely as
program costs), but did not think that it provided an adequate
solution. We thought that too few of the service contracts under
TANF would qualify for simplified treatment on that basis.

    We welcomed comments on how to deal with this latter dilemma, as
well as comments on our overall approach. We had discussed this issue
thoroughly during our pre-NPRM consultations, but thought this was a
policy area where no single, clear solution existed.
(b) Overview of Comments
    About one-third of all respondents to the NPRM submitted comments
on our administrative cost provisions. A substantial majority of these
comments came from representatives of State or local governments, but
we also received comments from unions, community organizations,
advocacy groups, national associations, business groups, and Congress.
We received comments from a significant majority of the States.
    Commenters generally opposed both the breadth of the proposed
``administrative cost'' definition and the scope of its application. To
some extent, unions, community organizations, legal aid and advocacy
groups were an exception to this general rule. Comments from these
groups tended to be more supportive of the proposed rule. However, they
expressed concern about the impact of these policies on the amount of
resources that would be available for direct benefits to needy families
and the potential impacts of the proposed rules on a State's decisions
about program administration, staffing, and contracting. One argued for
more specific exclusions from the definition (including costs
associated with the delivery of program services and overhead) out of
concern about the effect of a tight cap on case manager pay.
    To deal with the number and complexity of responses on this issue,
we have decided to cluster the comments into the following five general
categories: (1) the actual definition (including issues about the
appropriateness of a Federal definition, adopting definitions from
other programs, the treatment of eligibility determination and case
management costs, and the treatment of automated data processing
costs); (2) the treatment of costs incurred on contracted services; (3)
general questions about the calculation of the two caps; (4) specific
issues related to how we determine whether a State has exceeded the MOE
cap on expenditures of State MOE funds; and (5) specific issues related
to how we determine whether a State has exceeded the Federal cap
(including whether the appropriate base for computing the Federal cap
is the pre-transfer or post-transfer grant amount).
    As you will notice from the discussion that follows, regardless of
where they appear in the rule, the administrative cost issues are
closely connected to each other. For example, if we have a prescriptive
definition of administrative cost, this policy would exacerbate
concerns about the negative effects of requirements for subcontractors
to track such costs in the same way as TANF agencies.
    Although few commenters directly addressed the combined effects of
the proposed policies, we considered the combined effect of all these
provisions in drafting our responses.
    The subject area that received the most attention from commenters
was the proposed definition. Commenters disagreed about whether there
should be a Federal definition, suggested alternative definitions that
we could adopt, argued for exclusion of case management and eligibility
determination costs, raised some issues about the treatment of
automated data processing costs, and posed a few miscellaneous
questions.
(c) Federal Definition
    Comment: A relatively small number of commenters spoke directly to
the question of whether there should be a Federal definition of
administrative costs. The commenters' views were mixed, although more
argued against a Federal definition than for one.
    Among the arguments put forth in support of a definition were: the
value of having comparable approaches among TANF jurisdictions; the
importance of protecting benefits for needy families especially in
light of the elimination of constraints that had existed under the
former AFDC program; and the importance of having a meaningful and real
Federal limitation on administrative costs.
    Those opposed to a Federal definition argued that: (1) it should be
the State's prerogative to define administrative costs; (2) we had no
authority to define ``administrative costs''; or (3) we could defer to
State definitions and choose to regulate at some subsequent date if we
found that States were not adhering to the statutory limits.
    Response: While we do believe in granting States broad flexibility
to design their programs and have left key definitions up to the
discretion of the States, we also believe that there is a need for
Federal guidance on the definition of ``administrative costs.'' The
approach in this rule is a compromise between a Federal and State
definition. It sets a Federal framework that specifies some items that
must be considered ``administrative costs,'' but does not attempt to
fully define the term.
    We believe this framework is important. First, as the comments we
received demonstrate, there is no common view of the meaning of this
term. If we left this matter entirely to State discretion, we could
expect a diversity of approaches, and States might be subject to widely
different penalty standards. Also, the fear of a penalty might lead
some States to define the term so narrowly as to substantially
undermine the intent of the administrative cost cap provisions.
    We disagree with the comment that we lack the authority to define
``administrative costs.'' We have

[[Page 17810]]

responsibility for four penalty provisions--two on use of Federal funds
and two on MOE requirements--where the level of State expenditures on
administrative costs is a key issue. On many occasions, we have heard
statements about the importance of having clear Federal standards for
any penalty decisions that we make. In that context, we have both the
authority and the responsibility to provide standards in this area.
    As we indicated in the preamble to the proposed rule, we considered
not proposing a Federal definition. While that option had some appeal,
we were not disposed to deferring totally to State definitions. The
philosophy underlying the administrative cost caps is very important;
in order to protect needy families and children, it is critical that
the substantial majority of Federal TANF funds and State MOE funds go
towards helping needy families.
    We also indicated that we thought that, by providing a general
framework to States, we could avoid numerous disputes with individual
States about whether their definitions represented a ``reasonable
interpretation of the statute.''
(d) Applying Other Federal Definitions
    Comment: A substantial number of commenters suggested that the TANF
program adopt the definition proposed for the Child Care and
Development Fund. A much smaller number suggested that we adopt the
definition in effect under the Job Training and Partnership Act (JTPA)
program. Commenters argued that adoption of these other definitions
would improve program consistency and simplify program operations at
the local level. They also endorsed CCDF's exclusion of ``eligibility
determination'' as an administrative cost. One argued that the
different definition could put local agencies in the untenable position
of not being able to hire staff.
    Response: In terms of program coordination, we do not believe that
there is a strong advantage to selecting the CCDF definition over
JTPA's. Where TANF programs work extensively with local providers of
employment and training services, compatibility with JTPA may be more
important; where TANF and child care programs are administered by a
single agency or use a common set of service providers, compatibility
with child care providers may be more important.
    In the NPRM, we noted that our proposed definition was closely
related to the JTPA definition and thus should facilitate the
coordination of WtW and TANF activities and support the transition of
hard-to-employ TANF recipients into the workforce. As caseloads decline
and the proportion of hard-to-serve clients rises, coordination between
these two programs may become even more critical.
    While adopting the CCDF definition might facilitate TANF and CCDF
coordination, we do not believe that this coordination depends upon a
uniform definition. Also, given the differences in the caps of the two
programs (15 percent versus 5 percent) and the different legislative
histories, there is little reason to believe that Congress intended a
uniform definition.
(e) Treatment of Eligibility Determinations
    Comment: Many of those commenting on this issue objected to our
proposed inclusion of eligibility determination within the
administrative costs definition. Some argued that eligibility
determination was not an administrative activity and was not easily or
logically separable from case management. Still others commented on the
burden associated with our proposal, the general need for State
flexibility in this area, and the potential negative effects on a
State's ability to fund critical staff who work directly with clients.
    One State agency indicated that the distinction in our proposal was
not burdensome and would require only a slight change in its Random
Moment Study.
    Many commenters took strong exception to our characterization of
any portion of the eligibility determination process as administrative.
Among other things, they were concerned that: (1) it was inconsistent
with existing State practice; (2) the nature of work with families is
undergoing significant change, and application of the traditional AFDC
approach is no longer appropriate; (3) because eligibility
determination is part of the case management function, it should be
categorized as a program or service function than administration; (4)
the administrative responsibilities of staff performing functions such
as screening and assessment are integral to providing services; (5)
front-line eligibility determination is arguably a direct service,
under the first statutory goal of the TANF program; and (6) as workers
assume new roles, differentiating between eligibility and service
delivery is becoming increasingly difficult and less useful.
    A couple of commenters indicated that our regulations needed to
draw a clearer line between administrative and program costs. One
commenter provided several specific examples of situations where the
line between administrative and program costs that we drew in the
proposed rule was unclear, such as in diversion and sanction activities
and in determining hardship exceptions and compliance with behavioral
requirements.
    A significant number of commenters spoke to the burden of the
proposed requirement on TANF agencies. They argued that State and local
systems are not geared towards allocating expenses this way. They do
not want to divert resources to this activity.
    Commenters also made a general plea for flexibility, saying that
States need flexibility in order ``for the role of front line staff to
continue to evolve to best meet the goals of welfare reform'' and to
enable States to build partnerships with local service providers.
    Finally, several commenters noted that we presented this policy
only in the preamble, not in the regulation itself.
    Response: While we do not want our rules to distort State choices
about how to deliver services or to divert State resources to cost
accounting activities unnecessarily, we have a responsibility to uphold
the intent of the statutory administrative cost cap provisions by
ensuring that States are not spending large amounts of money on
eligibility determinations rather than program benefits or services.
    Also, we do not agree that States must incur a significant
administrative burden in order to identify the costs associated with
eligibility determination activities. We recognize that the nature of
staff responsibilities is changing and the line between case management
and eligibility determination is blurring. Thus, it may be more
difficult to develop rules for allocating the time of workers between
administrative and program activities. However, once a State develops
its allocation rules, the process of allocating staff time is
straightforward and no more difficult than the current cost allocation
process.
    We also recognize that the TANF program offers the possibility for
States to administer programs in new ways. We understand that States
are moving towards blended functions, and we support such efforts.
These final rules do not in any sense require States to have separate
administrative and program staff. They merely require that States
provide a reasonable method for determining and allocating
administrative and program costs.
    Welfare agencies have a long history of identifying the costs of
eligibility determinations and allocating these costs as administrative
activities. A variety of other significant, related programs--such as
Medicaid, the Child

[[Page 17811]]

Health Insurance Program (CHIP), and Food Stamps--continue to follow
this practice. Thus, this kind of cost allocation has been standard
operating procedure in a number of programs and has been accepted as a
normal part of doing business.
    We also believe that a clear policy on eligibility determinations
might produce more consistent penalty determinations and reduce audit
disputes, appeals, and litigation regarding application of the misuse
of funds and MOE penalties.
    Based on these considerations, we have decided to add eligibility
determinations to the list of administrative activities at
Sec. 263.0(b)(2). More specifically, this rule reflects the basic
definition that was in the proposed regulation at Sec. 273.0(b) (with
the same basic examples of administrative cost activities), but adds
the NPRM preamble policy that required eligibility determination to be
treated as an administrative cost. We recognize that this is a
significant policy decision that merits inclusion directly in the
regulatory text; we agree with commenters that it should not be
relegated to the preamble.
    Under the final rule, States may develop their own definitions of
administrative costs and cost allocation plans, consistent with this
regulatory framework.
    Also, as we discuss later, we provide States some flexibility in
the methods they use to determine the administrative costs associated
with contracts. However, we want to reiterate a point we made in the
preamble for Sec. 273.13 of the proposed rule: States must properly
allocate costs. They must attribute administrative, program, and
systems costs to benefiting programs and appropriate cost categories,
in accordance with an approved cost allocation plan and the cost
principles in part 92.
(f) Other Miscellaneous Suggestions for Inclusions and Exclusions
    Comment: A couple of commenters suggested that our definition make
a distinction between general overhead or indirect costs (which would
be considered administrative costs) and overhead and indirect costs
related to the provision of program services (which would be excluded).
A couple of commenters made the broader suggestion that our definition
should indicate that administrative costs do not include items such as
diversion activities, assessments, development of employability plans,
work activities, post-employment services and supports, and case
management.
    Response: The comments identified an area where the language in the
proposed rule was unclear. To address this problem, we have revised the
regulatory text. The revised language at Sec. 263.0(b)(1) excludes
costs of providing services and the associated direct administrative
costs from the definition of administrative costs. The revised language
at Sec. 263.0(b)(2) clearly treats indirect (or overhead) costs as
administrative costs. We included these costs as administrative costs
in the final rule because we believe this approach is most consistent
with the intent of the administrative cost caps and is the simplest and
most straightforward approach for States to implement.
    Comment: One commenter suggested that we specify that diversion
assessments are program costs and not administrative costs.
    Response: We believe the changes that we have made to
Sec. 263.0(b), and just discussed, adequately address this concern. The
rule at Sec. 263.0(b) now indicates that diversion and assessment
activities are both program service costs and not considered
administrative costs. (Note: Here, we would make a distinction between
assessment activities designed to identify needs and develop
appropriate service strategies versus assessing income, resources, and
documentation for eligibility determination purposes; the latter are
administrative costs.)
    Comment: One commenter said we should specifically define case
management.
    Response: We do not believe there is a need for a Federal
definition of this term.
    Comment: One commenter asked that we clarify that ``public
relations'' activities would not include State expenditures on
providing information to clients.
    Response: While we believe that the common meaning of ``public
relations'' would not include providing client information, at the new
Sec. 263.0(b)(1), we have added ``providing program information to
clients'' as one example of providing program service. Thus, this
activity would not be classified as an administrative cost under our
rules.
    Comment: One commenter asked that we clarify that domestic violence
and substance abuse services are not considered administrative costs.
    Response: We believe the new language at Sec. 263.0(b)(1)
adequately addresses this concern. It more directly states that costs
of providing services are outside the definition of administrative
costs, and it explicitly provides that screening and assessments are
examples of program services.
(g) Computer-Related Costs
    Comment: Several commenters had concerns that the definition of the
exclusion for computer-related costs was not sufficiently clear in the
NPRM. Their reasons were mixed. A couple of commenters wanted to make
sure that States did not have ``unfettered discretion'' in this area;
they saw this provision as a major loophole and did not want to see
money diverted from meeting the needs of poor families. Other
commenters felt that the regulations did not adequately address the
information technology exclusion.
    Response: We received a variety of comments on the exclusions of
information technology and computerization costs from the 15-percent
caps. Based on these comments, we have made some clarifying changes to
the regulatory language (which appear in Secs. 263.2(a)(5) and 263.13)
and are providing some guidance in the preamble. However, we do not
believe it is necessary or appropriate to develop detailed Federal
regulations on this issue. While the new regulatory language makes the
regulation more consistent with the statutory language and makes the
language for the Federal and MOE caps more consistent, it also reflects
our willingness to defer to State policies, as long as those policies
reflect a reasonable interpretation of the statutory language.
    We believe that the revised regulatory language represents the best
reading of the statutory language at section 404(b). The statute
provides for exclusion of certain systems costs in determining whether
a State has exceeded the Federal cap on administrative expenditures. It
does not exclude such systems costs from the definition of
administrative costs. Thus, in this rule, you will note that the
systems exclusion is not part of the definition of administrative costs
at Sec. 263.0(b). Rather, it appears in the sections where we explain
how we determine if a State has excess expenditures on administrative
costs.
    Comment: We received several comments asking us to clarify that
personnel costs necessary to comply with reporting requirements and for
tracking and monitoring computer systems are covered by the exclusion.
Likewise, we received a few comments asking us to clarify that the
following items would be excluded: (1) data collection and reporting
activities (such as hardware, personnel and supply costs they incur in
meeting the TANF disaggregated data reporting requirements); (2)
activities such as rental and purchase of computer

[[Page 17812]]

equipment and systems procurement; and (3) preparation of reports
required under the Act.
    Response: Under the final rules, we exclude from the 15-percent cap
all costs associated with the portions of information technology and
computer systems that are used for tracking or monitoring required by
or under part IV-A of the Act. The excludable costs are the full range
of costs directly associated with the development, maintenance, and
support of the relevant systems or the relevant portions of larger
systems. Nonsystems costs related to monitoring and tracking (e.g., for
the salaries and benefits of data entry clerks, evaluation staffs,
statisticians, and report writers) are not covered by this exclusion.
    Based on the comments, we have made some modest changes to the
definition of administrative cost at Sec. 263.0(b) and the descriptions
of the administrative cost caps at Secs. 263.2(a)(5) and 263.13. Under
the language in the proposed rule, we had not generally recognized that
some activities that would otherwise be ``administrative'' in nature
could be part of the systems exclusion. The one exception we mentioned
was ``management information systems,'' proposed at Sec. 273.0(b)(10).
    To provide the clarification commenters requested, we have revised
the language at Secs. 263.2(a)(5) and 263.13 to specify that the
systems exclusion covers items that ``would fall within the definition
of administrative costs at Sec. 263.0(b).'' In other words, items that
would normally be administrative costs, but are systems-related and
needed for monitoring or tracking purposes under TANF, fall under the
systems exclusion. Thus, we would not consider them in determining
whether a State has exceeded either of the 15-percent caps.
    We also added language at Secs. 263.2(a)(5) and 263.13 to specify
that the systems exclusion covers the salaries and benefits costs of
personnel who develop, maintain, support, or operate information
technology or computer systems used for tracking and monitoring. Under
the revised language, it is clearer that States may exclude personnel
and other costs associated with the automation activities needed for
TANF monitoring and tracking purposes. For example, they may exclude
expenditures related to computerization of both the fiscal and program
data collection and reporting requirements in part 265 and computer
charges related to generating required data and reports. However, they
do not exclude nonsystems costs related to monitoring and tracking
(such as personnel costs for data entry clerks, statisticians, and
report writers).
    Also, we made a minor change to the last example in our list of
examples of administrative costs. The revised language refers
generically to ``preparing reports and other documents'' rather than
``reports and documents related to program requirements.'' We revised
the language to avoid confusion; the NPRM language was too similar to
the statutory exclusion at section 404(b).
    Comment: One commenter said the regulation should address the
permissibility, within the exclusion, of electronic benefit transfer
(EBT), Fingerprint Imaging Projects, or other automated fraud
prevention activities.
    Response: While all these activities might be commendable, the
statutory exclusion is only for expenditures ``needed for tracking or
monitoring required by or under this part.'' EBT would not fit within
the exclusion because it is neither a tracking nor a monitoring
activity; as the statute at section 404(g) indicates, EBT systems are
``for providing assistance.''
    Fingerprint imaging and other anti-fraud activities might fall
under the systems exclusion. For example, expenditures to develop a
computerized fingerprint imaging system to identify fugitive felons or
individuals who have fraudulently misrepresented their residence would
clearly qualify as monitoring under the exclusion.
    Since we are not regulating the definition of this exclusion, we
are not attempting to draw fine lines between what systems costs should
be included versus excluded. We expect States to implement policies
that are consistent with a reasonable interpretation of the statute and
these regulations.
(h) Costs Incurred by Contractors
    Comment: Another area receiving a significant number of comments
was our proposal to apply the definition of administrative costs to
contractors and other agencies. The vast majority of commenters opposed
this proposal.
    One State indirectly argued that the policy was unnecessary,
pointing to the State's own cost consciousness and cognizance of the
need to limit administrative expenditures in contracts.
    A few commenters noted that we had included this policy proposal
only in the preamble, but not in the proposed regulatory text. At least
one asked that we add the preamble language to the regulation.
    One TANF agency requested that we provide more guidance on how
States should segregate the administrative costs associated with
subcontracted services.
    We organized most of the comments on this issue into four broad
categories: (1) suggestions that the 15-percent administrative cost cap
apply solely to costs incurred by the TANF agency; (2) the potential
effects of applying the administrative cost cap limitation to
contractor agencies; (3) the possible negation of existing performance-
based contracts; and (4) functionality considerations.
    A few commenters recommended that the administrative cost cap apply
solely to the expenditures of the TANF agency or that we should treat
State and local agencies alike, but not contractors.
    A much larger number of commenters expressed general concerns about
requiring the tracking of administrative costs to contractors. They
objected to: (1) the increased administrative burden on the TANF agency
and difficulties associated with tracking administrative costs of
contractors; (2) diversion of resources away from needy families to
tracking; and (3) inconsistencies between our policy and the policies
of other programs (e.g., JOBS and JTPA). Commenters also claimed that
our proposed policy would increase the administrative costs of the
program, hamper State and local efforts to improve program
administration and services, discourage collaborations with community-
based organizations and other service providers, violate Congressional
intent in limiting our regulatory authority, and impede State
procurement activities. For example, contractors might choose not to
compete because they would be reluctant to provide detailed
itemizations of their expenses, and States might refrain from
contracting for fear that unknown contractor costs might cause them to
exceed the cap on administrative expenditures. Several commenters
expressed concerns that our proposed policy would discourage the
development of performance-based contracts and similar funding
arrangements.
    A subset of commenters said we should base the treatment of
subcontractor costs on functionality considerations, looking at the
function performed by the contractor or subcontractor, not whether
contractors incur administrative costs. A few argued that direct
program services provided by contractors were not administrative in
nature. Commenters did not want the treatment of contract costs to be
based on ``an extremely difficult differentiation between
administrative and programmatic costs.''

[[Page 17813]]

    Response: We have decided that States should be able to determine
the administrative costs associated with contracts and subcontracts
based on the function or nature of the contract. For example, if a
State contracts for case management or job placement services, which
meet our definition of program services, the cost of the contracts
would be treated as program costs, not as administrative costs.
Further, as we discuss later, the entire costs of a contract for
payroll services would be treated as an administrative cost subject to
the 15-percent cap. If the State had a contract that included a mix of
administrative and programmatic activities, it would need to develop a
method for attributing an appropriate share of the contract costs to
administrative costs. We have revised the regulatory language to
reflect that decision.
    The approach in the proposed rule reflected some genuine concerns
about weakening the administrative cost caps and distorting State
decisions about whether to contract. Some commenters expressed similar
concerns. However, after reflecting on the totality of comments
received, we are convinced that the costs of our proposed approach
would have outweighed the benefits. The approach also might have
significantly undermined one of our regulatory objectives, i.e., to
give States the flexibility they need to serve low-income families.
    In administering and operating its TANF Program, each State should
make a determination of the most cost-effective and efficient method of
performing each of the necessary administrative and programmatic
functions. It may use in-house staff and resources, engage other State
or local government agencies, or solicit services from outside
contractors. Presumably, with each State's procurement procedures
requiring free and open competition, and oversight by auditors and
State legislative and regulatory bodies, the result of any solicitation
will be a high-quality service delivered at a reasonable and acceptable
cost.
    We believe that, once a particular function is determined to be
either administrative or programmatic, that characterization does not
vary based on the nature or identity of the service provider.
Therefore, if a contract is for a singular administrative or
programmatic service, the final rules would treat the entire contract
price as an administrative or programmatic cost, respectively. A State
would not need to further itemize the contract costs or consider the
individual cost components used to support the contract price.
    For example, payroll services is a traditional administrative
function. If a State opts to contract out the payroll responsibilities
for its TANF program, a State would treat the entire cost of that
contract as an administrative cost within the 15-percent cap. It would
be unnecessary to further define the contractor's own administrative
costs.
    On the other hand, if the State contracted with a third party to
perform a variety of functions that included a mix of administrative
and programmatic activities, the State would need to develop a method
for attributing an appropriate share of the contract costs for
administrative activities as administrative costs. Likewise, if another
agency (State, local, or private) were administering a piece of the
TANF program, the State would need to have a method for attributing an
appropriate share of the other agency's costs to administrative
activities.
    Presumably, in developing its individual cost proposals, each
contractor includes an allocated portion of their own administrative
costs or overhead. However, the matter of interest here is the extent
to which Federal and State expenditures are going to administrative
activities, not the individual cost components of contracted services.
    Our approach is consistent with the regulations at 45 CFR part 92
and should maintain the integrity of the 15-percent administrative cap
provisions.
    We do not believe this policy will necessarily bias State decisions
about how to deliver TANF services, e.g., towards contracting out, or
privatization, of program operations. First, the initial expenditure
reports we have received from States suggest that their administrative
costs are running well within the 15-percent caps; thus, they do not
appear to have a strong incentive to change any of their administrative
practices. Second, many other very important considerations go into
State contracting decisions--including the State agency's internal
capacity and expertise and larger political and budgetary
considerations. Third, we would expect the State agency, State
legislature, and other interested parties to consider the impact on
public employees as part of their deliberations. Lastly, because there
is a limited difference in the treatment of administrative costs
incurred by TANF agencies and third parties, the potential incentive
effects of this policy (towards privatization) are limited.
(i) Consolidated Caps
    Comment: A couple of commenters suggested that we should have a
single administrative cost cap that covers both Federal and MOE
expenditures.
    Response: The statute clearly requires a separate cap for each.
Also, it would not be feasible to apply the 15-percent limitation
across the total Federal TANF and State MOE dollars. The MOE cap
applies to the total amount of qualified State expenditures for the
fiscal year, i.e., per fiscal year. The Federal cap applies to the
adjusted SFAG. If a State reserves amounts from its fiscal year grant,
then the Federal cap could reflect expenditures over a number of fiscal
years.
(j) Compliance Periods
    Comment: One commenter questioned the requirement for quarterly
compliance with both the Federal and MOE caps. The commenter suggested
annual evaluation as an alternative.
    Response: We assume this comment reflects a reaction both to the
information required on the quarterly TANF Financial Report and some
unclear regulatory language in the proposed rule. First, while we do
require quarterly reporting of Federal and State administrative and
systems costs, we never intended to make quarterly determinations
whether the expenditure of State funds violated the MOE cap. The
statute at section 409(a)(7) clearly provides that this would be an
annual determination. Also, in reflecting on this comment we realized
that our regulatory text did not clearly state that compliance with the
MOE cap would be determined on an annual basis. Therefore, we have
added the phrase ``for the fiscal year'' to Sec. 263.2(a)(5)(i) to
clarify that this is an annual determination.
    The Federal administrative cost cap works somewhat differently. For
the purpose of the Federal cap, we would look at the total cumulative
amount spent on administrative activities from each annual Federal TANF
grant. Unless and until the total amount expended as administrative
expenditures (exclusive of appropriate systems costs) exceeded 15
percent of the Federal TANF grants (except WtW grants) for any fiscal
year, we would not identify a violation of the Federal administrative
cost cap. The Department of Labor administers the WtW administrative
cost limit. This policy is consistent with the discussion in the
preamble to the proposed rule for Sec. 273.13.
(k) Base for Computing the Cap
    Comment: A significant number of commenters (particularly those

[[Page 17814]]

representing States or State interests) argued that we should calculate
the 15-percent administrative cost cap based on the SFAG amount before
any State transfers to title XX or CCDBG (i.e., the adjusted SFAG).
    Several of these commenters maintained that, in defining the
Federal cap, section 404(b)(1) refers specifically to 15 percent of the
``grant.'' They interpret this language to mean that total SFAG amount
would be the appropriate number to use in determining the maximum
amount for administrative costs. A few made the additional comment that
we did not have the authority to reduce the amount of 15-percent
administrative funds available to the State under the statute by
applying the 15-percent limitation to a smaller base amount than the
adjusted SFAG.
    Commenters also expressed concerns that our proposed policy would
result in disincentives to the States to transfer funds to CCDBG or
title XX.
    Finally, a few commenters noted that our proposed rules used a
different base amount for computing the administrative cost cap and for
computing penalties. More specifically, we proposed to determine
penalties based on the adjusted SFAG (i.e., the SFAG amount minus
Tribal adjustments, but prior to any transfer), but we computed the
administrative cap for TANF based on the adjusted SFAG minus transfers.
This inconsistency seemed unjustified.
    Response: As we noted briefly in the discussion for Sec. 260.30, we
made a change to the definition of ``adjusted SFAG'' that addresses the
consistency concerns of commenters. The revised definition, which is
used for determining both the Federal administrative cost cap and
penalty amounts, excludes monies transferred to either the SSBG or
CCDBG programs. Like the proposed definition, it also excludes funds
removed from the State's grant because Tribes in the State elected to
operate their own TANF programs.
    Although the language of the administrative cost limit refers to
``the grant,'' we do not believe what is ``the grant'' is clear in this
context. We did not base the Federal administrative cap on the pre-
transfer amount because we believe that proposal would produce a
peculiar and undesirable policy result. In effect, it would allow
States to double-dip on their administrative expenditures. The
transferred funds would be part of the base that we would use to
determine how much Federal TANF money was available for administrative
costs within the TANF program. It would also be part of the base for
determining how much money was available in CCDBG or SSBG for capped
administrative expenditures within these programs, since the statute
provides that transferred funds are subject to the requirements of
these programs.
    We understand the concern that our policy in this area might create
modest disincentives for States to transfer Federal TANF funds to CCDF
and SSBG. However, we would point out a few factors that should
mitigate those concerns: (1) the initial TANF expenditure reports
suggest that administrative costs are generally running substantially
below the 15-percent caps; thus, States that transfer funds should be
able to live within the post-transfer cap amount; (2) this policy
affects the Federal cap only, not the MOE cap; (3) States that elect to
transfer funds might enjoy some reductions in their administrative
costs because they can operate more streamlined child care and social
services programs; (4) some of the costs associated with the new TANF
data rules are excludable from the cost caps under the information
technology and computerization exclusion; and (5) in several places,
these final rules reduce the data reporting and administrative burdens
to which States would have been subject under the proposed rules.
    You will find the discussion of the issues related specifically to
the MOE cap in the preamble for Sec. 263.2 and the discussion of issues
related specifically to the Federal cap in the preamble for
Sec. 263.13.

Subpart A--What Rules Apply to a State's Maintenance of Effort?

Section 263.1--How Much State Money Must a State Expend Annually to
Meet the Basic MOE Requirement? (Sec. 273.1 of the NPRM)

Overview
    To ensure that States would continue to contribute their own money
towards meeting the needs of low-income families, section 409(a)(7)
requires States to maintain a certain level of spending on programs on
behalf of eligible families. If a State does not meet the ``basic MOE''
requirements in any fiscal year, then it faces a penalty for the
following fiscal year. The penalty consists of a dollar-for-dollar
reduction in a State's adjusted SFAG.
    In the NPRM and in the discussion that follows, we address each of
the terms used in the basic MOE requirement.
(a) Historic State Expenditures
    Each State's basic MOE requirement reflects its historic spending
on welfare programs. We calculated the historic State expenditures
based on the State's FY 1994 share of expenditures for the AFDC, EA,
AFDC-related child care, transitional child care, At-Risk Child Care
and JOBS programs (including expenditures for administration and
systems operations).
(b) Adjusting a State's Basic MOE Level
    The statute authorizes an adjustment to a State's basic MOE level
when a Tribe or a consortium of Tribes residing in the State submits a
plan to operate its own TANF program, and we approve this plan. We will
reduce the State's basic MOE requirement beginning with the effective
date of the approved Tribal plan.
    Section 409(a)(7)(B)(iii) excludes from the basic MOE calculation
any IV-A expenditures made by the State for FY 1994 on behalf of
individuals covered by an approved Tribal TANF plan. Because TANF
funding for Tribes may also reflect a State's IV-F (JOBS) expenditures,
we also concluded that it was appropriate to reduce a State's basic MOE
levels for IV-A and IV-F expenditures. In summary, we proposed to
determine the percentage reduction in the SFAG due to Tribal programs
and apply the same percentage reduction to the State's basic MOE
requirement. The State's revised basic MOE level would apply for each
fiscal year covered by the approved Tribal TANF plan(s).
    For example, if the amount of the Tribal Family Assistance Grant
represents ten percent of the State's SFAG, then we would reduce the
State's basic MOE requirement by ten percent. This approach provides a
consistent method for determining both the reduction in the State's
SFAG and its required basic MOE level.
(c) Applicable Percentage
    Under section 409(a)(7)(B)(ii), if any State fails to meet the
minimum work program participation rate requirements in the fiscal
year, then it must spend at least 80 percent of its FY 1994 spending
level. If a State meets the minimum work participation rate
requirements, then the ``applicable percentage'' is 75 percent of its
FY 1994 spending level for the year. We refer to the dollar amount
representing 75 percent or 80 percent of the FY 1994 State expenditures
as the basic MOE level.
    We calculated each State's total FY 1994 expenditures and basic MOE
levels by using data on the State share of expenditures for AFDC
benefits and

[[Page 17815]]

administration, EA, FAMIS, AFDC/JOBS Child Care, and Transitional and
At-Risk Child Care programs reported by States on form ACF-231 as of
April 28, 1995, as well as the State share of JOBS expenditures
reported by each State on form ACF-331 as of April 28, 1995.
    We transmitted tables showing FY 1994 spending amounts and basic
MOE levels to the States via Program Instruction Number TANF-ACF-PI-96-
2, dated December 6, 1996. On October 31, 1997, we issued TANF Program
Instruction Number TANF-ACF-PI-97-9 informing States of revised basic
MOE levels. The revised basic MOE levels reflected a correction in the
calculation of the State share of FY 1994 At-Risk Child Care (ARCC)
expenditures. Although the data sources remained the same for all
States, some of the reported ARCC expenditure amounts used were revised
after the original calculation. As a result, the basic MOE levels for
some States increased. As TANF-ACF-PI-97-9 was issued so close to the
NPRM publication date, we were not able to include information on it in
the NPRM.
    We also determined FY 1994 spending and basic MOE levels for each
of the Territories. For IV-A expenditures for Puerto Rico, we used the
Financial Report Form ACF-231 as of April 28, 1995. For Guam and the
Virgin Islands, we used the share of expenditures that corresponded to
the amount on the Federal grant awards for FY 1994, i.e., the
Territories' share of AFDC benefit payments (25 percent), EA (50
percent), administration (50 percent), and Child Care (25 percent). For
JOBS, the Territories' basic MOE levels reflect expenditures reported
on the ACF-331 as of April 28, 1995.
    In addition, for both IV-A (AFDC, EA, and child care) and JOBS,
Guam and the Virgin Islands (but not Puerto Rico) benefit from Pub. L.
96-205, as amended (48 U.S.C. 1469a). This law permits waiver of the
first $200,000 of the Territories' share of expenditures. Therefore,
for Guam and the Virgin Islands, we reduced the share that they were
required to contribute, and thus their basic MOE amount, by $200,000.
(d) FY 1997 Basic MOE Level
    Under the proposed rules, we indicated that the State could prorate
its basic MOE level for FY 1997 by taking the total FY 1994 State
expenditures provided to the State in Program Instruction Number TANF-
ACF-PI-96-2, multiplying that number by the number of days during FY
1997 that the State operated a TANF program and dividing by 365. The
State's TANF implementation date is the date given in the Department's
completion letter to the State. The State had to meet 80 (or 75)
percent of the resulting amount.
Comments and Responses
    We received a few comments on this section. Two commenters
commended our proposal to reduce a State's basic MOE proportionately
when the State's TANF grant is reduced once a Tribe or a consortium of
Tribes residing in a State has received approval to operate its own
TANF program. Most of the other comments focused on the applicable
basic MOE level relative to a State's work participation rates. We have
made no substantive changes to the provisions in this section as a
result of the comments we received. However, as the result of some of
the comments we received, we have clarified the regulation. A
discussion of the comments follows.
(a) Applicable Percentage
    Comment: A few commenters requested that we amend the regulations
to provide that a State's failure to meet the two-parent minimum work
participation rate for a year does not automatically require the State
to meet 80 percent of its historic State expenditures. Instead, the
commenters recommended that, where the State fails only the two-parent
rate, the State must increase its spending level between 75 percent and
80 percent based on the ratio of the State's two-parent caseload to the
State's entire caseload. Associations representing States pointed out
that such an adjustment would be consistent with the proposed
regulation under Sec. 271.51 to reduce the maximum penalty amount for
failure to meet the work participation rate if the State fails only the
two-parent rate.
    Response: We recognize that the size of a State's two-parent
caseload may be small in comparison to the State's total caseload.
However, we do not have any discretion under the statute to adjust a
State's basic MOE in this way. Section 409(a)(7)(B)(ii) explicitly
provides that a State must meet 80 percent of its FY 1994 spending
level unless it meets the ``requirements'' of section 407(a) of the Act
for the fiscal year. Section 407(a) includes the minimum participation
rate requirements for both all families and two-parent families.
    In contrast, section 409(a)(3) requiring a penalty for failing to
satisfy minimum participation rates expressly provides for a reduction
in the penalty with respect to a fiscal year based on the degree of
noncompliance.
    Comment: Two commenters thought we should modify the final rule to
provide that the 75-percent spending level applies for every fiscal
year in which a State meets only the all-family participation rate.
They contended that the two-parent participation rate should not affect
the required spending level, particularly for a State that has a very
low two-parent caseload relative to its total caseload. One of the
commenters also believed that the 75-percent spending level should
apply immediately unless it can be shown after the fact that a State
has not met the work participation rate requirements.
    Response: We found no statutory basis for excluding the two-parent
rate from a State's applicable spending requirement. The 75-percent
spending standard, in the parenthetical at section 409(a)(7)(B)(ii),
requires that States meet both rates.
    We also disagree with the commenter's assertion that the 75-percent
spending level should apply immediately. To the contrary, the statute
requires that all States maintain an 80-percent spending level for each
fiscal year. The reduction is a parenthetical addition if the State
meets both participation rates for the fiscal year. Thus, a State would
need to demonstrate that it actually meets both rates for the fiscal
year for the 75-percent spending level to apply. This language suggests
that, to avoid the chance of penalty, it would be most prudent for a
State to plan to spend at the 80-percent level every year.
    Comment: One commenter indicated that we must be clear in the
regulations that a State qualifies for the 75-percent MOE standard if
it meets the Federal requirement for the year following application of
the caseload reduction credit.
    Response: We have revised the final rule to clarify that a State's
basic MOE will be reduced to 75 percent of FY 1994 expenditures if it
meets both the all-family and the two-parent participation rate that
applies following application of the caseload reduction credit.
    Comment: Two commenters suggested that we clarify the basic MOE
requirement for FY 1997. The commenters noted that the work
participation rate requirements apply no earlier than the fourth
quarter of FY 1997 for any State. As a result, the basic MOE
requirement for FY 1997 should only be based on whether a State met the
work participation rate requirements for the fourth quarter of FY 1997.
If a State achieves the required work participation rate for the July-
September 1997 quarter, the State's basic MOE requirement should be 75
percent of its historic expenditures.
    Response: We agree that the earliest period States must report
information

[[Page 17816]]

necessary to calculate participation rates under section 407 is the
fourth quarter of FY 1997. The penalty for failure to submit a required
quarterly report in a timely manner is one of several penalties that
has a delayed effective date. Section 116(a)(2) of PRWORA provides that
certain penalty provisions do not take effect until July 1, 1997, or
six months after we receive the State's complete TANF plan. We consider
the State's TANF implementation date to be the date that we received
its complete plan. Most States had to submit a report for all or part
of the fourth quarter of FY 1997.
    However, there is no delay in the penalty for failure to meet the
basic MOE requirement. This is one of several penalty provisions that
apply immediately, i.e., from the date a State implements its TANF
program. Thus, each State must maintain 80 percent of its historic
expenditures for FY 1997 unless it meets the work participation rate
requirements both for all families and two-parent families. (The
penalty for failure to satisfy the minimum participation rates
requirement also has a delayed effective date; however, that penalty is
separate from the basic MOE requirement.)
    The participation rates for a fiscal year are an average monthly
rate. For the States that had to submit a report for all or part of the
last quarter of FY 1997, we will calculate the average monthly rate for
all families and for two-parent families based on the number of months
that must be covered in the required quarterly report.
    The remaining States must meet 80 percent of their historic
spending levels unless they choose to submit data demonstrating that
they actually met both participation rates either for the period during
which they were operating their TANF program or for the last quarter of
FY 1997, whichever they choose. We decided to give these remaining
States this option because we did not think it would be fair to judge
their performance over a longer period of time than States that
implemented TANF at an earlier date. Also, we have more flexibility
with respect to these States since the statute does not specify a
precise time frame for measuring their performance.
    Comment: One commenter wrote that a State should not have to meet
the 80-percent level of effort if we waive the State's penalty for
failing to achieve either of the required work participation rates due
to reasonable cause. Another commenter requested clarification in this
area.
    Response: Under section 409(a)(7)(B)(ii) of the Act, the 75-percent
standard only applies if a State meets the minimum work participation
rates, not when a State has reasonable cause for failing to satisfy the
rates. Granting reasonable cause does not mean that the State met the
rates. States that fail to satisfy the minimum work participation
rates, but receive partial or full penalty relief, must still meet the
80-percent MOE requirement.
(b) FY 1997 Basic MOE Level
    Comment: One commenter asked that the final rule clarify how we
will calculate the FY 1997 basic MOE level.
    Response: We calculated the prorated basic MOE levels by first
determining the number of days in FY 1997 that a State operated the
TANF program. We then multiplied the resulting number of days by the
State's basic MOE level for the year, then divided by 365 (the number
of days in FY 1997).
    We originally published the States' basic MOE levels in Program
Instruction Number TANF-ACF-PI-96-2, dated December 6, 1996. We also
sent letters dated January 7, 1997, to TANF program directors
explaining that we would prorate basic MOE levels for FY 1997 only. As
explained earlier, we have since recalculated basic MOE levels for
States to correct for the revised State ARCC expenditure figures for FY
1994. As a result, the basic MOE levels for some States increased. We
transmitted the revised basic MOE levels for States via TANF Program
Instruction Number TANF-ACF-PI-97-9 dated October 31, 1997. This
instruction also included each State's prorated FY 1997 basic MOE
levels.
    However, for States whose revised basic MOE level increased, we did
not apply the revised rate retroactively. Rather, we are determining
State compliance with the FY 1997 basic MOE level requirements based on
the original numbers published in TANF-ACF-PI-96-2, prorated as
applicable. All revised State basic MOE levels published in TANF-ACF-
PI-97-9 apply beginning FY 1998.

Section 263.2--What Kinds of State Expenditures Count Toward Meeting a
State's Basic MOE Expenditure Requirement? (Sec. 273.2 of the NPRM)

Overview
(a) Qualified State Expenditures
    Section 409(a)(7)(B)(i) establishes the criteria for the
expenditure of State funds to count toward a State's basic MOE level.
Congress wanted States to be active partners in the welfare reform
process. Thus, States must spend a substantial amount of their own
money on aid to needy families. While Congress gave States significant
flexibility in this area, it did establish a number of important
statutory restrictions on which State expenditures qualify towards the
basic MOE requirements.
    Section 409(a)(7)(B)(i) defines ``qualified State expenditures'' to
include certain expenditures by the State under all State programs. We
interpret ``all State programs'' to mean the State's family assistance
(TANF) program plus any other separate State program that assists
``eligible families'' and provides appropriate services or benefits.
Thus, States could expend State funds for MOE purposes in three ways.
    In addition to expending State funds in separate State programs,
States may expend funds within the TANF program in two different ways.
They may commingle their State funds with Federal grant funds, or they
may use State funds that have been segregated from their Federal grant
funds.
    We remind States that there are specific statutory requirements
that affect the use of State funds under a State's TANF program. States
need to be mindful of the TANF requirements to help avert penalties
under section 409 of the Act. The specific TANF requirements that apply
depend upon which of various programmatic terms is used in the language
describing the requirement.
    States may also expend State funds in a State program separate from
TANF to provide the benefits and services listed under section
409(a)(7)(B)(i)(I) of the Act, e.g., cash assistance, child care
assistance, and education activities. None of the TANF program
requirements directly apply to eligible families served in separate
State programs.
    Requirements in the statute that use the terms ``under the
program,'' ``under the program funded under this part,'' and ``under
the State program funded under this part'' apply to the State's TANF
program, regardless of the funding source. That is, they apply to
segregated Federal programs, commingled State/Federal programs, and
segregated State programs. Thus, all families receiving TANF assistance
(whether funded with State or Federal TANF funds) must meet work
participation and child support requirements.
    Conversely, some Federal requirements derive from the provisions in
the statute that use the term ``grant,'' or ``amounts attributable to
funds provided by the Federal government.'' These terms refer to the
Federal TANF funds provided to the State under

[[Page 17817]]

section 403. Therefore, they only affect the use of Federal TANF funds,
unless the State commingles its money with Federal TANF funds. If a
State commingles its funds, the Federal and State funds become subject
to the same rules. Thus, commingling of State and Federal TANF funds
can reduce the total amount of flexibility available to the State in
its use of Federal and State funds.
    Requirements pertaining solely to the use of Federal TANF funds do
not apply to families assisted under TANF with State-only funds.
Consequently, if a State segregates its TANF State funds from its
Federal TANF funds, State expenditures on assistance must comply only
with all of the rules that generally pertain to the TANF program, e.g.,
work and child support requirements. They are not subject to
requirements that pertain only to the use of Federal TANF funds.
    A State might choose to operate a ``segregated'' TANF program
because certain limitations, e.g., time limitations and certain alien
restrictions, apply to the program funded with Federal TANF funds that
would not apply to a TANF program funded wholly with State funds.
    Whether the expenditure of State funds is within the TANF program
or separate from the TANF program, to count toward meeting the State's
basic MOE, all expenditures must: (1) be made to or on behalf of an
eligible family; (2) provide ``assistance'' to eligible families in one
or more of the forms listed in the statute under section
409(a)(7)(B)(i)(I); and (3) comply with all other requirements and
limitations set forth in this part of the regulations, including those
set forth in Secs. 263.5 and 263.6.
(b) Eligible Families
    Section 409(a)(7)(B)(i)(I) provides that State funds under all
State programs must be spent with respect to eligible families to count
toward the State's basic MOE. Section 409(a)(7)(B)(i)(IV) further
clarifies that an eligible family means a family eligible for
assistance ``under the State program funded under this part.'' The
``State program funded under this part'' is the State's TANF program.
    Thus, we proposed that, in order to be considered an ``eligible
family'' for MOE purposes, a family must have a child living with a
custodial parent or other adult caretaker relative (or consist of a
pregnant individual) and be financially needy under the TANF income and
resource standards established by the State under its TANF plan. This
definition includes two categories of families. It includes all
families funded with MOE funds under TANF, including certain alien
families or time-limited families who cannot be served with Federal
TANF funds, but who are being served in a segregated State TANF
program. (We discuss this alien limitation in detail further on in this
section.) It also includes a family that meets these criteria, but is
not receiving TANF, and instead is receiving benefits and services from
a separate State program. The expenditures to provide these benefits
and services under all State programs may count toward the MOE
requirement, provided the expenditures also meet all other requirements
and limitations set forth in part 263.
    A State is free to define who is a member of the family for Federal
TANF purposes and may use this same definition for MOE purposes. For
example, it could choose to assist other family members, such as
noncustodial parents, who might significantly enhance the family's
ability to achieve economic self-support and self-sufficiency. By
including such individuals within its definition of family, a State
could provide them with services through TANF or a separate State
program. Noncustodial parents could then engage in State-funded
activities such as work or educational activities, counseling, or
parenting and money management classes.
    The NPRM stated that we expect States to define ``child''
consistent either with the ``minor child'' definition given in section
419 or some other definition applicable under State law. The State must
be able to articulate a rational basis for the age they choose.
    The definition of ``eligible family'' expressly includes families
that ``would be eligible for such assistance but for the application of
section 408(a)(7) of this Act.''
    Under section 408(a)(7), States may not use Federal TANF funds to
provide TANF assistance to a family that includes an adult who has
received federally funded assistance for a total of 60 months.
Therefore, if a family becomes ineligible for Federal assistance under
the TANF program due to this time limit, but still meets the definition
of eligible family, then this family may be considered an eligible
family for MOE purposes. (Note: In the NPRM, in Sec. 273.2(c), we did
not accurately cite the applicable criteria. The final rule at
Sec. 263.2(c) corrects this error; in referencing paragraph (b) of this
section, it captures all three criteria for ``eligible families.'')
    Section 5506(d) of Pub. L. 105-33 (the Balanced Budget Act of 1997)
clarified that the definition of an eligible family also includes
lawfully present aliens who would be eligible for TANF assistance, but
for the application of title IV of PRWORA.
    Thus, the definition of eligible family allows States to claim MOE
expenditures with respect to three types of family members: (1) those
who are eligible for TANF assistance; (2) those who would be eligible
for TANF assistance, but for the time-limit on the receipt of federally
funded assistance; and (3) those lawfully present who would be
eligible, but for the application of title IV of PRWORA. An alien
family who meets any one of these three criteria may be considered an
eligible family provided they also meet the family composition
requirement (i.e., have a child living with a custodial parent or other
caretaker relative or be a pregnant individual) and financial
eligibility criteria established by the State. These last two
requirements are based on the statutory language stating that eligible
families ``means families eligible for assistance under the State
program funded under this part (TANF) * * * and that would be eligible
for such assistance.''
    While this three-part definition of eligible families may appear to
allow States to claim qualified expenditures with respect to all
lawfully present alien eligible family members, i.e., both qualified
and nonqualified aliens, as discussed further below, this is not
necessarily the case. Nor is it the case that the amendment to the
definition under the Balanced Budget Act precludes States from claiming
MOE for illegal aliens under certain circumstances.
    While we mentioned the 1997 amendment in the NPRM, at that time, we
had not fully analyzed the significance of the statutory language
defining ``eligible families'' for MOE claiming purposes, relative to
the extant eligibility provisions in title IV of PRWORA. Title IV of
PRWORA sets forth the aliens who are eligible for Federal public
benefits and for State and local public benefits; whereas, the
definition of eligible families limits the expenditures that may be
claimed for MOE. While there is obvious overlap between these two
concepts, they are distinct and must be analyzed separately.
    To understand eligibility for Federal TANF benefits, readers must
be familiar with the definition of qualified alien, Federal public
benefit, and Federal means-tested public benefits. Section 401 in title
IV of PRWORA provides that, in general, only qualified aliens, as
defined in section 431 of PRWORA, are

[[Page 17818]]

eligible for Federal public benefits. (At the end of this discussion,
we explain two very limited circumstances under which it may be
possible for a State to provide certain benefits to all aliens.) The
definition of ``qualified aliens'' at Sec. 260.30 refers to section 431
of PRWORA, as amended (e.g., by the Illegal Immigration Reform and
Immigrant Responsibility Act of 1996 (Pub. L. 104-208) and the Balanced
Budget Act of 1997 (Pub. L. 105-33)). The revised definition of
``qualified aliens'' includes: legal permanent residents; asylees;
refugees; aliens paroled into the U.S. for at least one year; aliens
whose deportations are being withheld; aliens granted conditional
entry; battered alien spouses, battered alien children, the alien
parents of battered children, and alien children of battered parents
who fit certain criteria; and Cuban/Haitian entrants.
    The Department has interpreted the term ``Federal public benefit''
(see TANF-ACF-IM-98-5, dated August 24, 1998, transmitting the notice
with comment period interpreting ``Federal public benefit,'' published
in the Federal Register dated August 4, 1998, vol. 63, No. 14, and
available on line at /programs/ofa/im98-5.htm).
It has determined that the TANF program (when using Federal TANF funds)
generally provides a Federal public benefit. The Department also issued
an interpretation of the term ``Federal means-tested public benefit''
and designated the TANF program as a Federal means-tested public
benefit. (See the Federal Register for August 26, 1997, 62 FR 45256.)
    Even qualified aliens may be ineligible for means-tested Federal
public benefits. Section 403 of PRWORA prohibits qualified aliens, with
exceptions, who arrive on or after August 22, 1996, i.e., newly arrived
aliens, from receiving Federal means-tested public benefits for five
years.
    Exceptions to the five-year bar include qualified aliens who are
refugees, asylees, or aliens whose deportation is being withheld,
Amerasians, and Cuban/Haitian entrants, as well as veterans, members of
the military on active duty, and their spouses and unmarried dependent
children.
    However, as discussed below, States may elect to help any newly
arrived aliens who are eligible family members by providing either
State or local public benefits or benefits that are not a State or
local benefit. If the State uses segregated State funds in TANF or
funds in separate State programs to provide such benefits, the
expenditures may qualify as MOE.
    Further, in regard to whether qualified aliens may be eligible for
Federal TANF benefits, section 402 of PRWORA provides that States have
the authority in TANF to decide whether to help qualified aliens who
arrived in this country prior to August 22, 1996, and qualified aliens
who arrived on or after August 22, 1996, but for whom the five-year bar
had expired. In other words, States are authorized to decide whether
qualified aliens are eligible for the State's TANF program. However, a
State may not deny certain qualified aliens eligibility even if it
decides that as a general matter qualified aliens are not eligible to
receive Federal TANF benefits. States may not deny eligibility to
refugees and asylees, aliens whose deportation has been withheld,
Amerasians, and Cuban/Haitian entrants. These groups are eligible for
Federal TANF benefits for five years after the date of their entry into
the country or the date asylum or withholding of deportation was
granted. Also, States may never deny eligibility to legal permanent
residents who have worked forty qualifying quarters or to aliens who
are veterans, members of the military on active duty, and their spouses
and unmarried dependent children.
    As with other parts of the TANF program, the way the State
structures the delivery of TANF and MOE benefits determines which
eligibility requirements apply. If a State commingles Federal TANF
funds with State funds, then the benefits provided must follow the
rules at section 401(c) for Federal public benefits. A State providing
Federal public benefits to aliens and using commingled TANF funds to
help aliens may claim, for MOE purposes, only the expenditures that it
makes with respect to eligible qualified alien family members. Eligible
qualified aliens include those who are eligible for TANF assistance;
would be eligible for TANF assistance, but for the time limit on
receiving federally funded TANF assistance; or are lawfully present in
this country and would be eligible for TANF assistance, but for the
application of title IV of PRWORA. If the State decides to restrict the
eligibility of noncitizens to receive TANF benefits and commingles its
MOE funds, then it will only be able to claim toward MOE the
expenditures that it must make on behalf of the excepted qualified
aliens mentioned above.
    If a State does not commingle Federal and State funds, but instead
uses segregated State funds in its TANF program or separate State
program funds to provide benefits that meet the definition of a State
or local public benefit, then it must follow the rules of section 411
of PRWORA. State or local public benefits have the meaning prescribed
under section 411(c) of PRWORA. It is generally up to the State to
determine if the benefits it offers are State or local public benefits
within the meaning of the Act. However, because we interpreted that the
TANF program, using Federal TANF or State commingled funds, generally
provides a Federal public benefit, we also would interpret that the
TANF program, using State TANF funds that have been segregated from
Federal TANF funds, generally provides a State or local public benefit
(subject to the limited circumstances explained at the end of this
discussion). We make this interpretation because the statutory language
in section 401(c) is identical to the language in 411(c). Within the
meaning prescribed under section 411(c), States would also determine
whether various separate State or local programs or activities are
State or local public benefits.
    Section 411(a) of PRWORA provides that only qualified aliens and
certain nonqualified aliens are eligible for State or local public
benefits. The nonqualified aliens consist of nonimmigrants under the
Immigration and Nationality Act or aliens paroled into this country
under section 212(d)(5) of such Act for less than one year. There are a
handful of legal nonqualified aliens, e.g., temporary residents under
the Immigration Reform and Control Act (IRCA), aliens with protected
status, and aliens in deferred action status who are prohibited from
receiving State or local public benefits under this provision. (At the
end of this discussion, we explain two very limited circumstances under
which it may be possible to provide certain benefits to all aliens.)
    Section 411(d) of PRWORA permits States to expand alien eligibility
by providing State or local public benefits to illegal aliens. But this
may occur only if the State enacts a law after August 22, 1996, that
affirmatively provides that illegal aliens are eligible to receive (all
or particular) State or local public benefits.
    Section 412 of PRWORA also allows States, at their option, to
further limit alien eligibility for State public benefits. There are
time-limited exceptions for refugees, asylees, or aliens whose
deportation has been withheld. Like Federal TANF benefits, these groups
are eligible to receive State public benefits under TANF during their
first five years following entry, the grant of asylum, or the
withholding of deportation. The