Temporary Assistance for Needy Families is Title IV-A of the Social Security Act, as rewritten by Public Law 104-193 in 1996, codified at 42 U.S.C. 601 and following, and the TANF block grant created by that part places its demands first on states rather than on individual families. A state that accepts the grant must submit a plan built on outline and certification, must spend federal and counted state dollars only in a manner reasonably calculated to accomplish a statutory purpose, must engage a measured share of families receiving assistance in countable work activities, must spend state funds at a required maintenance of effort level, and must observe a federal limit on the use of federal funds for assistance after sixty months. An individual family, by contrast, receives no federal entitlement from the part itself, because section 601(b) states that the part is not to be interpreted to entitle any individual or family to assistance under a state program funded under the part. The statute therefore requires design, spending discipline, work measurement, and reporting from governments, while leaving the existence and level of a benefit to state law within the federal boundaries that the following sections describe.

After reading, a reader can name the four statutory purposes, notice that only two are means-tested, and describe the caseload reduction credit that decides whether the work requirement binds. Nearly every operative rule in Part A depends on distinctions stated early and applied later. The central claim has a short form that a reader can carry through every later provision: two of four purposes have no income test. Section 601(a) conditions the first purpose on need, since assistance is for needy families so that children may be cared for in their own homes or in the homes of relatives, and conditions the second purpose on need, since the end of dependence is framed for needy parents through job preparation, work, and marriage. The third purpose, prevention and reduction of out-of-wedlock pregnancies with annual numerical goals, contains no reference to need, and the fourth purpose, encouragement of the formation and maintenance of two-parent families, contains no reference to need either. A grant that appears at first sight to be a cash assistance program is therefore, on the face of the purposes clause, a broader family policy fund for a defined share of lawful spending.

The money provisions explain why that breadth had lasting fiscal consequences between 1996 and 2019. Section 603 set the state family assistance grant from federal payments under the predecessor programs in the middle of the 1990s, using the greatest of several measures drawn from fiscal 1992 to fiscal 1995, and then repeated the resulting state amounts without an inflation adjustment, a caseload adjustment, or a population adjustment. The Congressional Research Service, in tables updated July 1, 2019, recorded the combined grant at 16.567 billion dollars in fiscal 1997 and at 16.512 billion dollars in fiscal 2018 after a reduction of 0.33 percent beginning in fiscal 2017 to finance research and technical assistance, and placed the fiscal 2018 grant 36.2 percent below the fiscal 1997 grant in real terms. A fixed nominal grant combined with a purposes clause broader than cash assistance gave states a pool of dollars that stayed constant as caseloads fell, and section 404(e) allowed reservation of unused dollars without a fiscal year limit, subject to use only within the program. The work provisions supply the second major distinction, between a headline standard and an effective standard. Section 607(a) set a participation standard of 50 percent for all families and 90 percent for two-parent families, defined engagement through hours and through twelve countable activities divided between core and supplemental tiers, and then allowed the caseload reduction credit in section 607(b)(3) to subtract the percentage decline in the caseload from the base year to the comparison year, after exclusion of declines caused by eligibility changes. A decline of 40 percent therefore produced an effective all families standard of 10 percent, and a decline of 50 percent or more produced an effective standard of zero. Department of Health and Human Services annual reports recorded a weighted average effective requirement of 3.9 percent for all families in fiscal 2003 and 5.0 percent in fiscal 2006, against a nominal rate of 50 percent in each year. The Deficit Reduction Act of 2005 recalibrated the credit from a fiscal 1995 base year to a fiscal 2005 base year for standards applicable in fiscal 2007, brought families in separate state programs into the calculation, and directed federal definition and verification of countable activities. Block grant structure and the caseload reduction credit are standard examinable material in social policy and public budgeting coursework, and candidates practice and revise US government and civics material on ReportMedic as part of preparation on federal grant design, precisely because a statute of that kind cannot be understood from a headline rate alone.

A third distinction decides which federal rules attach to a given dollar. Congress used the term assistance throughout Part A without defining that term in the statute, and the Department of Health and Human Services defined assistance in the regulation issued April 12, 1999, at 45 C.F.R. 260.31, as benefits designed to meet ongoing basic needs. Non-assistance, under the same regulation, covered nonrecurrent short-term benefits that addressed a specific crisis, did not meet recurrent needs, and did not extend beyond four months, together with work subsidies, child care and transportation for employed families, refundable tax credits, contributions to Individual Development Accounts, and services that provided no basic income support. The sixty month limit, the participation calculation, child support assignment and cooperation rules, work sanctions, and detailed case reporting in section 611 all followed assistance, so a shift from ongoing cash aid toward services and short-term benefits changed both the obligations on a family and the federal record of that family. Maintenance of effort and penalties completed the design. Section 609(a)(7) required state spending at 80 percent of the fiscal 1994 level, reduced to 75 percent for a state that met both participation rates, with a dollar for dollar grant reduction for a shortfall, while the Administration for Children and Families reported for fiscal 2018 that basic assistance accounted for 21.4 percent of combined federal and state funds, against 71 percent in fiscal 1997.

The Purposes Clause and the End of Entitlement

The operative text for Temporary Assistance for Needy Families sits at the opening of Part A of Title IV of the Social Security Act, in sections 601 to 619 of Title 42 of the United States Code. Congress placed a statement of purpose and a denial of individual entitlement at the head of the part, before the grant provisions, the state plan provisions, and the performance and penalty provisions that occupy most of the remaining sections. The design begins with purpose, because purpose controls the lawful uses of the money. Under section 604, a state may use a federal grant in any manner reasonably calculated to accomplish the purpose stated in section 601, and that cross reference converts a short introductory paragraph into the spending constitution for the entire block grant. The thesis of the scheme is therefore precise: the real policy sits in the purposes clause, not in the spending rules that follow, because the spending rules take scope from the purposes clause.

The text arrived through the 1996 welfare reform statute that created the block grant, the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, enacted on August 22, 1996 as Public Law 104-193. The opening title of that act struck the prior Part A and inserted a new Part A headed as block grants to states for temporary assistance for needy families. The Congressional Research Service, in a legislative history published in 2016, recorded that states were required to implement the new program by July 1, 1997, with an option to implement earlier upon submission of a state plan. Between August 1996 and July 1997, states drafted plans under a purposes clause that had replaced a more detailed federal framework. The 1996 act repealed that earlier architecture and substituted a grant whose federal definition consisted, in large part, of four purposes and a set of state obligations stated elsewhere in the part.

Which of TANF’s four purposes is tied to annual numerical goals?

The statute lists four purposes in sequence: assistance for needy families so children can live at home or with relatives; an end to dependence of needy parents through job preparation, work, and marriage; prevention and reduction of out-of-wedlock pregnancies, with annual numerical goals; and encouragement of two-parent families.

Section 601(a) opens by declaring that the purpose of the part is to increase the flexibility of states in operating a program designed to achieve the listed ends. The opening clause names flexibility for states as the overarching purpose, and the four numbered purposes then define the ends toward which that flexibility may be exercised. Flexibility is not stated as an end in itself. The Congressional Research Service, in a block grant primer issued in 2018, described the same structure, presenting the statutory purpose as an increase in state flexibility directed toward four statutory goals and explaining that federal funds could be used in any manner reasonably calculated to accomplish that purpose. A purposes clause that begins with state flexibility reads as a grant statute: the federal contribution is money plus purposes, and the state contribution is program design.

The first purpose addresses assistance to needy families so that children may be cared for in their own homes or in the homes of relatives. The object of the assistance is the family, framed through the needs of the child. The stated result is care in a home setting, either the parental home or the home of a relative, rather than care in an institution or in a placement outside the family network. The qualifying word is needy, and need is left for state definition in eligibility rules adopted under a state plan. The first purpose is therefore the closest of the four to the predecessor program, since Aid to Families with Dependent Children had also directed assistance toward needy children and toward care in a home. Continuity should not be overstated. Under the predecessor structure, federal law specified categories of eligibility in considerable detail, and a state plan had to conform to federal requirements as a condition of federal payment. Under section 601(a), the first purpose states an end and a population, and then stops. The clause sets no benefit level, no income threshold, no asset limit, and no duration, and each element was assigned to state decision, subject to provisions on time limits, work, and reporting elsewhere in Part A.

The second purpose addresses the dependence of needy parents on government benefits and states that such dependence is to end through promotion of job preparation, work, and marriage. The clause shifts the subject from families to parents, and shifts the policy object from the condition of the child to the economic conduct of the adult. Job preparation and work connect to the work participation provisions and to services, such as child care and transportation support, that the Congressional Research Service in 2018 listed among the uses of block grant funds directed toward employment. Marriage, placed third in the same series, belongs to a different category. Marriage is a family status rather than a labor market activity, and inclusion in the second purpose signals that the statute treats family formation as an instrument of economic policy. The second purpose, like the first, contains a need condition: the parents described are needy parents. Services justified under the second purpose therefore carry a means related character when justification rests on the text itself, even though a state possesses wide latitude in setting the income level that counts as need. That latitude became a central feature of administration between 1996 and 2019, as states defined need for child care subsidies, refundable state tax credits, and employment services at levels well above the limits applied to monthly cash assistance.

The third purpose moves from assistance and employment services toward prevention. Section 601(a) states the end as prevention and reduction of the incidence of out-of-wedlock pregnancies, coupled with establishment of annual numerical goals for preventing and reducing that incidence. The goal setting element distinguishes the third purpose from the other three. No numerical goal is required for home based care, for ending dependence, or for two-parent family formation. For pregnancy prevention, the statute asks for annual numerical goals, a formulation that turns a social objective into a measurable target. The activities contemplated are different in kind from cash assistance. Pregnancy prevention operates through education, counseling, health services, and public communication, and such activities address a population defined by risk or by age rather than by receipt of a benefit. The clause contains no reference to needy families, needy parents, or need at all. The omission is conspicuous against the first and second purposes, where needy appears in each clause, and against the name of the program itself. An express condition in two adjacent clauses and an absence of that condition in the next clause point toward a deliberate distinction rather than toward an implied condition carried across the sentence.

The fourth purpose states that the program is designed to encourage the formation and maintenance of two-parent families. Formation concerns families that do not yet exist in that form, and maintenance concerns families that exist and face pressures toward separation. The clause therefore authorizes attention both to marriage promotion at family formation and to supports intended to keep two parents together after children arrive. As with the third purpose, the clause contains no need condition. No income limit, no reference to needy families, and no restriction to families receiving assistance appears in the text. The Congressional Research Service, in analysis published in 2018, stated the resulting rule in paraphrase: activities funded to achieve the third and fourth purposes were not limited to needy families, could be provided to the population as a whole without a test of economic need, and could reach families without children. That reading follows the text, because a pregnancy prevention campaign addressed to adolescents or a marriage education program addressed to engaged couples cannot sensibly be confined to families that have passed a means test for cash assistance.

Why the absence of a need test in two purposes matters

The drafting distinction between the first two purposes and the last two purposes is the legal mechanism by which a welfare block grant became a general purpose state fund. A block grant governed only by the first purpose would resemble a flexible cash assistance grant, with a population limited to needy families with children. A block grant governed only by the first two purposes would add employment services for needy parents, a broader instrument than cash alone but one anchored to economic need. The addition of the third and fourth purposes, without a need condition, breaks that anchor for a substantial share of lawful spending. A state may fund a media campaign on pregnancy prevention for a general audience, a school based program open to all students, a fatherhood program serving fathers across income levels, or a marriage education program serving couples who have never applied for assistance, and each expenditure can be defended as reasonably calculated to accomplish a stated purpose of Part A. The Administration for Children and Families, in guidance issued in 1999, treated an expenditure as permissible when the expenditure served one of the four statutory purposes.

The consequences appear in the composition of spending documented by the Congressional Research Service for fiscal years from 2010 to 2019. During that decade, the Congressional Research Service recorded that basic assistance accounted for only a portion of combined federal and state spending, with substantial shares directed toward child care, work related services, refundable tax credits, child welfare services, prekindergarten, and family formation and pregnancy prevention activities. Shares varied by state and by year, and no single share is necessary to the legal point. The legal point is structural: a purposes clause that reaches beyond needy families permits a spending portfolio that reaches beyond needy families. Flexibility advocates in the 1996 debate, including Republican governors who had sought waiver based experimentation under the predecessor program, presented breadth as a virtue on the ground that states possessed better knowledge of local needs than a single federal eligibility code. Accountability concerns, stated by the Congressional Research Service in descriptive analyses issued between 2010 and 2019, held that breadth without a need test weakened the connection between federal appropriations and aid to the poorest families. The statute resolves that tension in favor of breadth, and then relies on reporting, on maintenance of effort rules in section 609, and on state political accountability to discipline choices that the purposes clause permits.

How the state plan requirement works in practice

Section 602 supplies the second element of the opening design, and section 602 should be read as a certification mechanism rather than as a benefit schedule. Under section 602(a), a state qualifies as an eligible state for a fiscal year by submitting a plan to the Secretary of Health and Human Services during the period specified in the statute, a period framed by reference to a twenty seven month window ending with the close of the first quarter of the fiscal year. The plan must include an outline of the family assistance program that the state intends to conduct. The outline must describe a program, designed to serve all political subdivisions of the state, though not necessarily in a uniform manner, that provides assistance to needy families with children or with children expected to be born, and that provides parents with job preparation, work, and support services intended to enable the parents to leave the program and become self sufficient. The outline must also address work engagement for parents and caretakers receiving assistance and coordination with other resources. The special provisions that follow require the state document to indicate treatment of families moving into the state, to indicate whether assistance will be provided to noncitizens, and to set forth objective criteria for the delivery of benefits and the determination of eligibility and for fair and equitable treatment, including an explanation of hearing or appeal opportunities for recipients adversely affected.

The certifications in section 602(a) give the plan legal character. A state plan includes certifications by the chief executive officer of the state on subjects that include designation of the state agency or agencies that will administer and supervise the program, provision of equitable access to assistance for members of Indian tribes domiciled in the state who are not served under a tribal plan, and establishment and enforcement of standards against program fraud and abuse, including standards on nepotism, conflicts of interest, kickbacks, and political patronage. An optional certification on domestic violence standards, covering screening, referral to supportive services, and good cause waivers of specified requirements, appears in the same subsection. The Administration for Children and Families, in guidance consolidated from 1997 to 1999, treated the plan as the document through which a state described program operation and supplied the required certifications.

Two features of section 602 explain why the plan does not function as a federally approved benefit schedule. First, the statute requires an outline and certifications, not a table of benefit amounts approved line by line by the Secretary. A plan states that a state will conduct a program serving needy families, designates an administering agency, and sets forth objective criteria for eligibility and fair treatment, but amounts, income disregards, and resource limits live in state law and state manuals rather than in the federal plan document. Second, the statute contemplates variation within a state as well as among states, since the program must serve all political subdivisions but not necessarily in a uniform manner. Two states could therefore set maximum benefits at sharply different levels while each satisfied section 602. Comparisons published by the Congressional Research Service and by the Administration for Children and Families for selected months in 2017 and in 2018 showed wide interstate variation in maximum benefits for a family of three, a pattern consistent with a certification model rather than with a national benefit schedule. The federal role is real but bounded: the Secretary receives the plan and determines whether the required outline and certifications are present, while benefit design remains with the state within provisions elsewhere in Part A, including the five year limit on federally funded assistance in section 608 and the work requirements in section 607.

What the no individual entitlement clause does

Section 601(b) states, in a single sentence paraphrased here, that Part A is not to be interpreted to entitle any individual or family to assistance under any state program funded under the part. The sentence is an interpretive direction addressed to courts and to agencies, and the breadth of the direction matters. The clause does not say that a state may choose to deny entitlement as a matter of state law. The clause says that the federal part itself supplies no entitlement, whatever a state chooses to provide under state authority. A benefit paid under a state program funded by the block grant therefore rests on state law for existence and for enforceability, subject to constitutional limits and to any independent federal requirement that applies on another ground. The Congressional Research Service, in primers published in 2017 and in 2018, presented the point as a defining contrast: Aid to Families with Dependent Children was an entitlement to individuals who met state determined eligibility standards within the federal framework, and the statute for Temporary Assistance for Needy Families states expressly that the program is not an entitlement for individuals.

The predecessor entitlement had practical content that the 1996 act removed. Under Aid to Families with Dependent Children, a family that met the applicable eligibility rules held a claim to assistance, federal funding shared in the cost for eligible families without a fixed national cap of the block grant kind, and a state could not simply close intake to qualified families because a fixed grant had been exhausted. Eligibility disputes under the predecessor program therefore proceeded on the premise that wrongful denial of a benefit owed under the rules was a remediable wrong. Section 601(b) removes that premise for programs funded under Part A. After July 1, 1997, the general implementation date recorded by the Congressional Research Service in 2016, an applicant denied assistance could invoke the fair treatment criteria and hearing opportunity described in the state plan under section 602, state administrative procedure, and constitutional protections against arbitrary administration, but could not invoke Part A itself as the source of a right to receive assistance.

Practical consequences follow directly from that legal change. A state may close intake when funds allocated to cash assistance in a state budget are committed, because no federal entitlement obliges the state to serve every qualified applicant. A state may impose a waiting list for a service funded under the block grant, or may limit enrollment in a subsidized employment or education component, without violating a federal promise of aid to each eligible family. A state may set eligibility thresholds at a low level, may adopt diversion practices that steer applicants toward a single short term payment or toward services instead of ongoing assistance, and may reduce benefit amounts through state law, without confronting a federal entitlement that would override the state choice. Conversely, a state may choose generosity: state law may extend assistance to categories excluded from federal funding, and a state may create enforceable rights under state law if a state legislature elects that course. Section 601(b) neither compels generosity nor forbids generosity. The clause removes a federal floor of individual right and leaves protection to state decision, bounded by the purposes clause, by express requirements in sections 602 to 619, and by fiscal reality under a fixed grant structure described by the Congressional Research Service in 2018.

The interaction between section 601(b) and section 601(a) completes the opening design. The purposes clause broadens what a state may fund, reaching beyond needy families for pregnancy prevention and two-parent family formation, while the entitlement clause narrows what an individual may demand, reaching even the core case of a needy family with children seeking cash assistance at home. Broad authority for the state and no federal right for the applicant are two sides of the same 1996 settlement. Support for that settlement, voiced in 1996 by President William J. Clinton in signing Public Law 104-193 on August 22, 1996 and by congressional sponsors of the block grant design, rested on accountability through state choice and on the claim that a fixed grant would force prioritization. Opposition, voiced in 1996 by the Children’s Defense Fund under Marian Wright Edelman, rested on loss of a guaranteed floor for poor children and on the risk that broad purposes would divert resources away from the poorest families during economic downturns. Section 601 enacts the mechanism about which the argument revolved: four purposes, two conditioned on need and two free of any need condition, a state plan built on outline and certification, and a single sentence denying individual entitlement.

The Four-Purpose Spending Map

Statutory purpose	Requires need	Spending it authorises	Reported share of combined funds
Provide assistance to needy families so children may be cared for in their own homes or in the homes of relatives	Yes	Basic assistance in cash, vouchers, and benefits for food, clothing, shelter, utilities, and household needs	21.4 percent as basic assistance in fiscal 2018
End the dependence of needy parents on government benefits by promoting job preparation, work, and marriage	Yes	Child care, work activities, education and training, work supports, and refundable tax credits authorized for needy parents	Reported within child care at 17.0 percent, work activities at 10.7 percent, and refundable credits at 9.0 percent in fiscal 2018
Prevent and reduce the incidence of out-of-wedlock pregnancies and establish annual numerical goals	No	Pregnancy prevention education, counseling, health services, and public communication authorized for a general population	1.5 percent as prevention of out-of-wedlock pregnancies in fiscal 2018
Encourage the formation and maintenance of two-parent families	No	Marriage education, fatherhood programs, and family formation services authorized without a need test	0.7 percent as fatherhood and two-parent family formation and maintenance in fiscal 2018

The Money: How the Grant Is Calculated, and What the Fixed Amount Did Over Time

How much is the TANF block grant?

The basic TANF block grant for the fifty states, the District of Columbia, and tribes totals 16.489 billion dollars per year in nominal terms, a figure set in the 1996 statute and described by the Congressional Research Service in 2019 as essentially unchanged in nominal dollars since enactment.

That total was never calculated from a measure of need in 1996, from a count of poor children in that year, or from a projection of what cash assistance, work programs, and child care would cost in later years. The total was the sum of state by state amounts, and each state amount was a backward looking measure of what the federal government had paid that state under the programs that the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 replaced. The Congressional Research Service, in its primer on Temporary Assistance for Needy Families financing updated through 2019, identified the predecessor programs as Aid to Families with Dependent Children, Emergency Assistance, and the Job Opportunities and Basic Skills Training program. Under those programs the federal government had reimbursed a share of state spending without a national dollar ceiling for Aid to Families with Dependent Children and Emergency Assistance, so payments rose and fell with claimed spending. Section 603 of title 42 ended that open ended reimbursement and substituted a fixed grant determined once, by reference to spending in the middle of the 1990s.

A Grant Built From Spending Before the Block Grant Existed

Section 603(a)(1) names the payment the family assistance grant, and the definition that controls the amount uses the phrase state family assistance grant. The enacted definition in the 1996 law, at section 403(a)(1) of the Social Security Act as added by that law and codified at 42 U.S.C. 603(a)(1), set the grant as the greatest of several measures drawn from fiscal 1992 to fiscal 1995. One measure was based on fiscal 1994, taking the total amount required to be paid to the state under the former section 403 for fiscal 1994, with an added element tied to Emergency Assistance in fiscal 1995 where that spending exceeded the fiscal 1994 level and where a state plan amendment for Emergency Assistance had been approved. A second measure took one third of the total required to be paid for fiscal years 1992, 1993, and 1994 combined, which had the arithmetic effect of an average over those three years. A third measure was built from the first three quarters of fiscal 1995, combined with the amount for the corresponding period, so that a state whose spending had risen into fiscal 1995 was not confined to the lower totals of an earlier year. Child care spending under specified subsections of the former law and spending under the former part F program were excluded from parts of that calculation.

The choice of the greatest of those measures mattered for distribution. A state that had spent heavily in fiscal 1994 received a grant anchored to that peak rather than to a lower three year average, and a state whose federal payments had grown by fiscal 1995 could have that later level reflected through the fiscal 1995 measure. The Congressional Research Service legislative history, published in 2016, traced the formula to a reconciliation of House and Senate versions that emphasized different base years. The result was a distribution map drawn from history. States that had operated larger programs, paid higher benefits, or enrolled more families in the base years entered the block grant era with larger grants per poor child, and states that had operated smaller programs or paid lower benefits entered with smaller grants per poor child, and the statute preserved that relationship by repeating the same state amounts. No provision in section 603(a)(1) directed the Secretary of Health and Human Services to recalculate the state amounts from updated poverty, caseload, or spending data after the base period closed. The Center on Budget and Policy Priorities, in analyses published in 2011 and in 2018, found that states disadvantaged by the historical formula received less than half the federal funding per poor child received in other states, even after the supplemental grants described below were counted for the years when those grants operated.

A Fixed Nominal Amount With No Adjustment Mechanism

The second feature of section 603 is the absence of an adjustment mechanism. The grant amounts were written in nominal dollars, meaning dollars of the year in which payment was made without conversion to constant dollars, and the statute contained no inflation adjustment, no caseload adjustment, and no population adjustment. The Congressional Research Service stated the point directly in its frequently asked questions report, in the version updated July 1, 2019: a basic block grant amount, nationally and for each state, was established in the 1996 law, that amount for the fifty states, the District of Columbia, the territories, and tribes combined was listed in the Congressional Research Service tables as 16.567 billion dollars, and from fiscal 1997 through fiscal 2016 that amount remained the same and was not adjusted for inflation, for the size of the assistance caseload, or for changes in the poverty population. The combined figure of 16.567 billion dollars included the small territorial grants described below, which accounted for the difference from the 16.489 billion dollar figure for the state family assistance grants themselves. From fiscal 2017, the Congressional Research Service reported, the grant was reduced by 0.33 percent to finance research and technical assistance related to the program, producing a reduced combined amount of 16.512 billion dollars in fiscal 2017 and fiscal 2018. That reduction was a small legislated deduction rather than an indexation.

The absence of adjustment operated in both directions. A state whose need rose, whether because unemployment rose, because poverty rose, because population grew, or because the cost of providing a given benefit rose with prices, received the same nominal dollars from the basic grant, apart from the separate and conditional streams described below. Conversely, a state whose caseload fell kept the same nominal dollars. The statute did not reclaim the difference between the fixed grant and the lower cost of a smaller caseload, and it did not require a state to return funds that falling caseloads had freed from cash assistance spending, provided the funds were used for purposes authorized under the program. Supporters of the 1996 structure, as summarized in the Congressional Research Service legislative history, treated federal budget predictability and state freedom to redirect freed funds toward work supports and child care as advantages. Analysts at the Center on Budget and Policy Priorities, in analyses in 2011, 2012, and on April 2, 2018, treated the same symmetry as a weakness during downturns, because need could rise while the basic resource remained flat. That fiscal fact also sits underneath the caseload reduction credit addressed in another part of the article on the statute. A declining caseload had two consequences under the statute as a whole, in that the credit made a percentage based participation standard easier to meet and the fixed grant meant that the dollars associated with the departed cases remained with the state for other authorized uses rather than returning to the federal Treasury.

What a Fixed Nominal Amount Did as Prices Rose

Because the nominal amount did not move, the real value of the grant fell every year in which prices rose. Two dated and attributed measurements, both available before October 1, 2019, quantified that fall, and the figures differ for stated methodological reasons rather than because one measurement refuted the other. The Congressional Research Service, in tables updated July 1, 2019, calculated the grant in nominal dollars and in constant 1997 dollars, using the Consumer Price Index for All Urban Consumers from the Bureau of Labor Statistics as the price series. In that table the combined grant stood at 16.567 billion dollars in fiscal 1997 and at 16.512 billion dollars in fiscal 2018, after the 0.33 percent reduction described above, while the fiscal 2018 amount expressed in constant 1997 dollars was 10.564 billion dollars. The cumulative change was recorded as 36.2 percent, so the Congressional Research Service concluded that in real terms the fiscal 2018 block grant stood 36.2 percent below the fiscal 1997 grant. The table also showed gradual erosion, reaching 20.4 percent by fiscal 2006 and 33.1 percent by fiscal 2016, from the repeated combination of a flat nominal line and a rising price level.

The Center on Budget and Policy Priorities, in an analysis dated April 2, 2018, projected a larger figure for a later year. That analysis stated that the block grant had lost more than one third of its inflation adjusted value since 1996 and that, without cuts proposed in the budget then under discussion, the block grant was projected to lose 39 percent of its inflation adjusted value by 2019. The footnote to that projection identified the method as an analysis adjusted for inflation using the Consumer Price Index for All Urban Consumers research series, a revised historical price series, rather than the standard Consumer Price Index for All Urban Consumers series used in the Congressional Research Service table. Three differences therefore separate the 36.2 percent figure from the 39 percent figure. The end years differ, because one figure measured an outcome through fiscal 2018 while the other projected through 2019, and an additional year of price increase adds to the measured real loss. The character of the figures differs, because one was a table of recorded values through a completed fiscal year and the other was a projection made in April 2018 for a fiscal year that had not ended. The price series differ, because the research series revised parts of the historical inflation record and therefore produced a somewhat different conversion from nominal dollars to constant dollars. Neither source presented the other as in error, and the two results are consistent with a single underlying fact, namely that the statute provided no inflation adjustment and that more than one third of the real value present in fiscal 1997 had disappeared by the end of the second decade of the program. Population growth between 1996 and 2019 further reduced the fixed dollars available per poor child, a pressure that a price index does not capture, as the Center on Budget and Policy Priorities noted in its 2011 analysis.

The Smaller Streams Written Alongside the Basic Grant

Section 603 and the sections near it in the code contained several smaller funding streams, each with a different logic from the basic grant. The largest qualification in the original 1996 design was the contingency fund at 42 U.S.C. 603(b). The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 established that fund at 2 billion dollars to provide extra funding for states that were economically needy, in recognition, as the Congressional Research Service explained in its primer, that fixing the basic grant without adjustment for economic circumstances raised a concern that funding might prove inadequate during economic downturns. A state had to meet a test of economic need, expressed through unemployment and through participation in the food stamp program that was renamed the Supplemental Nutrition Assistance Program in 2008, and a state also had to meet a spending condition, separate from the general maintenance of effort requirement, under which state spending counted for the contingency fund had to exceed the level of state spending in the predecessor programs in fiscal 1994, a threshold that was likewise not adjusted for inflation. Federal payments then matched qualifying state expenditure above that threshold, subject to the amount available in the fund. The original 2 billion dollars was a capped fund rather than an open ended appropriation, and the Congressional Research Service reported that the original amount was depleted in early fiscal 2010, after which Congress provided appropriations in later laws that continued the fund, including an annual appropriation of 608 million dollars beginning in fiscal 2017. Tribes and the territories were not eligible for contingency fund grants under the rules summarized by the Congressional Research Service, so the qualification applied within the fifty states and the District of Columbia.

A second stream, the supplemental grants, addressed distribution rather than economic downturns. The grants were provided under the authority in the area of 42 U.S.C. 603(a)(2) for states with high population growth and for states with low historic grants per poor child, the two groups that the historical formula had placed at a disadvantage. The Congressional Research Service reported that seventeen states qualified, and full funding was 319 million dollars per year in the account given by the Center for Law and Social Policy in its summary of the Claims Resolution Act of 2010. The seventeen states included Alabama, Arizona, Florida, Georgia, Nevada, North Carolina, Texas, and Utah. The Center on Budget and Policy Priorities reported in 2011 that in most recipient states the grants had accounted for nearly ten percent of the state annual federal allocation. Funding was discontinued after June 30, 2011, as the Congressional Research Service stated in its primer. The Claims Resolution Act of 2010 had funded the grants through June 30, 2011 at a reduced level, and no funding followed for later years within the period ending October 1, 2019.

A third stream arrived a decade after the original design and reflected the fourth statutory goal, the promotion of the formation and maintenance of two parent families. The Deficit Reduction Act of 2005, enacted on February 8, 2006 as Public Law 109-171, established funding of 150 million dollars per year for healthy marriage promotion and responsible fatherhood initiatives, administered through grants and contracts awarded by the Department of Health and Human Services. The Congressional Research Service, in its legislative history of the block grant and in a separate 2006 report on the Deficit Reduction Act provisions, divided that 150 million dollars into 100 million dollars per year for healthy marriage initiatives and 50 million dollars per year for responsible fatherhood initiatives, with a further detail in the 2006 report that up to 2 million dollars per year of the total could be set aside for demonstration projects coordinating services for tribal families. The funded activities, as the Congressional Research Service summarized the 2005 law, included education on the value of marriage, relationship skills education, mentoring for fathers, and employment support connected to fatherhood programs. The money was awarded for specified initiatives rather than paid to each state by the historical formula.

A fourth stream divided the basic grant itself along a different line. Under 42 U.S.C. 612, a federally recognized Indian tribe, or a consortium of tribes, could submit a tribal family assistance plan and, on approval by the Secretary of Health and Human Services, receive a tribal family assistance grant directly. The Administration for Children and Families, in guidance on tribal funding, and the Congressional Research Service, in its primer, described the amount in parallel terms. The tribal grant equaled the federal share of expenditures in fiscal 1994 under the predecessor programs that was attributable to Indian families residing in the service area identified in the tribal plan, including administrative expenditures attributable to those families. The grant was financed by a reduction in the state family assistance grant of the state or states in which that service area lay, so tribal grants carved a share out of the existing state grant rather than adding a new national amount on top of the 16.489 billion dollars. A tribe specified both a service area and a service population in the tribal plan, and the amount reflected expenditures for all Indian families in the designated service area, as the Administration for Children and Families explained in its published answers to tribal funding questions.

Grants to the territories formed the remaining part of the family assistance structure. Puerto Rico, Guam, and the United States Virgin Islands received family assistance grants to operate programs under the same general framework of plan, work, and maintenance of effort requirements that applied to the fifty states and the District of Columbia, as the Administration for Children and Families summarized in budget material published in 2005. A territory allocation was based on historic funding levels, with a total of 77.9 million dollars made available annually for the territorial grants in that account, an amount consistent with the difference between the 16.489 billion dollar state figure and the 16.567 billion dollar combined figure in the Congressional Research Service tables. Territories were also eligible for a separate matching grant, subject to a ceiling under section 1108 of the Social Security Act, at a 75 percent federal matching rate and up to 15 million dollars annually, according to the same summary. American Samoa was designated as a state for the program in section 1101(a)(1) of the Social Security Act but had opted not to receive funding, as the Congressional Research Service noted.

Money That Did Not Expire at the End of the Year

Section 404(e) of the Social Security Act, the authority to reserve amounts for assistance, allowed a state to reserve unused dollars without a fiscal year limit. The Congressional Research Service stated the rule in its primer: a state could reserve unused dollars without a fiscal year limit, and that permission allowed states to accumulate balances of unused grants. Reserved balances could be used only within the program and could not be transferred to the Child Care and Development Fund or to the Social Services Block Grant. The regulation history for the carryover rule, published by the Administration for Children and Families in the Federal Register on April 6, 2010, traced an amendment made by the American Recovery and Reinvestment Act of 2009, which broadened the permissible use of carried over funds from assistance alone to any benefit or service that could be provided under the state or tribal program. Before that 2009 amendment, carried over funds had been confined to assistance as defined in regulation, together with associated administrative expenses.

The interaction between the carryover rule, the fixed grant, and falling caseloads in the first years produced an accumulation that drew congressional attention between 1998 and 2002, because caseloads fell faster than spending plans adjusted while the grant continued at the full fixed amount. The Congressional Research Service reported that on September 30, 2000 states held an unobligated balance of 3.2 billion dollars, meaning funds neither spent nor committed under an obligation, in addition to amounts that states had obligated but not yet liquidated. The design rationale recorded by the Congressional Research Service was that the ability to reserve funds was viewed at the creation of the program as a way of saving for unanticipated expenditures, such as an increase in benefit costs during an economic downturn, when the fixed basic grant would not rise on its own. The statute imposed no deadline by which a reserved dollar had to be spent or surrendered.

The Work Participation Requirement and the Caseload Reduction Credit

Section 607 of title 42 of the United States Code, codifying section 407 of the Social Security Act, contains the work participation machinery of the Temporary Assistance for Needy Families block grant. The provision performs two functions in tension with each other. Section 607(a) fixes the share of families that a state must engage in work, section 607(c) fixes the hours that engagement requires, and section 607(d) fixes the activities that count toward those hours. Section 607(b)(3) then supplies the caseload reduction credit, a formula that lowers the prescriptive target in proportion to a decline in the assistance caseload. The rate a state faced was therefore never the rate printed in section 607(a) alone. The operative figure, described by the Congressional Research Service in its 2019 primer on the block grant as the effective standard, was the statutory rate minus the credit, with a further adjustment for a state that spent state funds above the required maintenance of effort level. A state could meet the standard by engaging families in countable activities for the required hours, by shrinking the caseload from which the rate was calculated, or by a combination of the two paths.

The contrast with the predecessor program gives the machinery its context. Aid to Families with Dependent Children, the open ended entitlement in place before the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, matched state spending without a fixed federal ceiling and imposed no comparable national participation standard of the same design. The 1996 act substituted a fixed block grant, a five year federal time limit in section 608(a)(7), and the standards in section 607, so that federal accountability ran through measured engagement and caseload size rather than through matching payments for each eligible family served, a distinction examined in the AFDC and TANF models compared directly. Section 609(a)(3) supplied enforcement, authorizing a reduction in the block grant for a state that failed to meet the applicable rate, subject to the reasonable cause and corrective compliance provisions elsewhere in section 609.

The Statutory Targets and the Phase In Schedule

Section 607(a) set two targets. For all families containing a work eligible individual, the target was 50 percent. For two parent families, the target was 90 percent. The denominator consisted of families receiving assistance that included a work eligible individual, subject to the disregards in section 607(b) and in regulation, and the numerator consisted of families in that denominator that were engaged in work under the hours and activity rules in sections 607(c) and 607(d). A family that participated for fewer than the required hours, or only in an activity outside the statutory list, did not enter the numerator even though the family remained in the denominator, a distinction that the Congressional Research Service emphasized in its 2019 primer.

Neither target applied at full strength from the start. Under the original schedule in section 607(a), the all families rate began at 25 percent in fiscal 1997 and rose by five percentage points in each succeeding year, reaching 30 percent in fiscal 1998, 35 percent in fiscal 1999, 40 percent in fiscal 2000, 45 percent in fiscal 2001, and 50 percent in fiscal 2002, where the rate remained in the statute through October 1, 2019. The Department of Health and Human Services restated that schedule in the preamble to the interim final rule published on June 29, 2006. The same 2006 preamble recorded that the two parent rate stood at 75 percent in fiscal 1997 and fiscal 1998 and rose to 90 percent for fiscal 1999 and thereafter. The Administration for Children and Families applied those full rates in information memorandum TANF-ACF-IM-2003-02, which reported that the minimum standards for fiscal 2002 were 50 percent for all families and 90 percent for two parent families.

Hours, Core Activities, and the Twelve Countable Activities

Compliance depended on hours as much as on headcounts. Section 607(c) defined engaged in work through average weekly hours measured across the month, and the Congressional Research Service set out the structure in a table in its 2019 primer drawn from section 407 of the Social Security Act. A single parent with a child under age six counted as engaged at an average of 20 hours per week during the month, and all 20 hours had to fall within core activities. A single parent whose youngest child had reached age six counted at an average of 30 hours per week, of which at least 20 hours had to fall within core activities, leaving an average of up to 10 hours per week that could fall within supplemental activities. A two parent family that did not receive federally funded child care counted at an average of 35 hours per week combined between the parents, of which at least 30 hours had to fall within core activities, leaving an average of up to five hours per week for supplemental activities. A two parent family that received federally funded child care counted at an average of 55 hours per week combined between the parents, of which at least 50 hours had to fall within core activities, again leaving an average of up to five hours per week for supplemental activities. The Administration for Children and Families stated the two parent figures in the same terms in its work participation guidance, specifying 35 hours with at least 30 core hours where federally funded child care was absent, and 55 hours with at least 50 core hours where federally funded child care was present.

The core and supplemental distinction divided the statutory list into two tiers. Section 607(d) listed twelve activities, and the Congressional Research Service in its 2019 primer grouped nine as core and three as supplemental. The core tier consisted of unsubsidized employment, subsidized private sector employment, subsidized public sector employment, work experience, on the job training, job search and job readiness assistance, community service programs, vocational educational training, and providing child care services to an individual participating in a community service program. The supplemental tier, countable only after the core threshold had been met, consisted of job skills training directly related to employment, education directly related to employment in the case of a recipient who had not received a high school diploma or a certificate of high school equivalency, and satisfactory school attendance at a secondary school or in a course of study leading to a certificate of general equivalence in the case of a recipient who had not completed secondary school. A single parent of a child under age six therefore had access in practice to only the nine core activities, because the 20 core hours required of that parent exhausted the total hours requirement, a consequence that the Congressional Research Service noted expressly in its 2019 primer.

Limits qualified the bare list. The Congressional Research Service, in its 2006 guide to the new definitions of work participation, recorded that job search and job readiness assistance counted for a limited number of weeks in a fiscal year, that vocational educational training counted for 12 months in the lifetime of the recipient, and that a cap limited the share of a caseload that could count through vocational educational training and teen school attendance combined.

How could a large caseload decline reduce a state’s effective work standard to zero?

The caseload reduction credit is the reduction in a state target rate under section 607 equal to the percentage decline in average monthly caseload between the base year and the comparison year, after excluding declines caused by federal or state eligibility changes.

Section 607(b)(3) supplied the formula behind that sentence. The credit for a fiscal year equaled the percentage point reduction in a state average monthly caseload between the base year and the comparison year, where the comparison year was the fiscal year immediately preceding the year for which the rate was being determined. If the average monthly caseload in the comparison year stood 40 percent below the base year caseload, after the exclusions described below, the credit was 40 percentage points. Subtraction produced the effective standard. A state with a 40 percent decline faced an effective all families target of 10 percent, because 50 minus 40 equals 10, and an effective two parent target of 50 percent, because 90 minus 40 equals 50, where the state applied the overall decline to the two parent rate. A state whose caseload fell by half faced an effective all families target of zero, because 50 minus 50 equals zero, and a state whose caseload fell by more than half produced a credit larger than the statutory rate, which in administration meant that the effective standard remained at zero rather than falling below zero. The Administration for Children and Families described the same arithmetic in information memorandum TANF-ACF-IM-2003-02 for fiscal 2002, explaining that the all families standard was reduced by the number of percentage points by which the overall caseload had declined from fiscal 1995 to fiscal 2001. For the two parent rate, that memorandum recorded a state option either to use the decline in the two parent caseload or to use the decline in the overall caseload.

The exclusion for eligibility changes bounded the formula. Section 607(b)(3) required that caseload reductions resulting from changes in federal or state eligibility rules be excluded. A state that narrowed eligibility or adopted a categorical exclusion that ended assistance for a class of families could not convert the resulting decline into a lower work target.

The scale of the credit during the first decade overwhelmed the nominal standard in most states. The Department of Health and Human Services reported in the Seventh Annual Report to Congress, covering fiscal 2003, that the weighted average effective requirement was 3.9 percent for all families, against a nominal rate of 50 percent, and 20.2 percent for two parent families, against a nominal rate of 90 percent. The Eighth Annual Report to Congress, covering fiscal 2006, reported a weighted average effective requirement of 5.0 percent for all families and 18.7 percent for two parent families. The Congressional Research Service, in its 2006 overview of welfare reauthorization in the 109th Congress, recorded that 19 states in fiscal 2004 had caseload declines of 50 percent or more from fiscal 1995, reducing the effective all families standard in those states to zero. The national participation rate in the same period fluctuated around 30 percent, as the Congressional Research Service documented in its 2013 report on welfare waivers using Department of Health and Human Services data for fiscal 2002 through fiscal 2009, so that most states met the effective standard even though the national rate remained well below 50 percent. Where the credit equaled or exceeded the statutory rate, the work requirement did not bind through the participation penalty in section 609(a)(3), because the standard against which the state rate was compared had already fallen to zero. Where the credit was small, the nominal rate retained force.

Recalibration, Broader Counting, and Verification

As enacted in 1996, the base year was fiscal 1995, the last full fiscal year before enactment and a year in which the predecessor caseload stood near a historical peak. The Deficit Reduction Act of 2005, Public Law 109-171, signed on February 8, 2006, recalibrated the credit by substituting fiscal 2005 as the base year, effective for the standards applicable in fiscal 2007. The Administration for Children and Families summarized the change in information memorandum TANF-ACF-IM-2015-01, listing recalibration to a base year of 2005 rather than fiscal 1995 as the first of the major work participation changes made by that act and its implementing regulations. From fiscal 2007 onward, a state earned a credit only for declines after fiscal 2005.

The Deficit Reduction Act made two companion changes. The first brought families receiving assistance in separate state programs into the participation calculation. Before the act, a state could fund assistance through a separate state program financed with maintenance of effort dollars, outside the federal calculation, and thereby keep families subject to the 90 percent two parent standard, or families with substantial barriers to employment, outside the rates. The Congressional Research Service, in report RS22369 dated March 1, 2007, recorded that states had used separate state programs to assist two parent families, to operate programs for recipients attending college, and to assist special populations including families with a disabled member, and that the act required states to count families in separate state programs in determining work participation rates. The second change concerned definition and proof. The act directed the Department of Health and Human Services to define each listed activity by regulation to ensure consistent measurement across states, to establish uniform methods for counting and reporting hours, and to determine the circumstances in which a parent living with a child receiving assistance counted as a work eligible individual. The department issued the regulations as an interim final rule on June 29, 2006, published at volume 71 of the Federal Register beginning at page 37454. The act and the rule also required each state to maintain a work verification plan describing how the state would count, verify, and document hours in each activity, and the final rule published on February 5, 2008, at volume 73 of the Federal Register beginning at page 6772, retained that framework with modifications after public comment, as summarized by Liz Schott of the Center on Budget and Policy Priorities in a February 20, 2008 analysis.

The recession that began in December 2007, as dated by the National Bureau of Economic Research and as recounted by the Congressional Research Service in its legislative history of the block grant, placed the recalibrated system under opposite pressure. Rising need increased caseloads in many states after fiscal 2007, and an increase in the comparison year caseload relative to fiscal 2005 produced a smaller credit, or no credit, and therefore a higher effective standard at the moment when employment opportunities contracted. The American Recovery and Reinvestment Act of 2009, Public Law 111-5, signed on February 17, 2009, addressed that interaction through a temporary modification in section 2101 of that act. The Administration for Children and Families implemented the modification in program instruction TANF-ACF-PI-2009-02, which explained that for fiscal 2009, fiscal 2010, and fiscal 2011 a state could use either the ordinary comparison year or the credit qualified for when the comparison year was fiscal 2007 or fiscal 2008, whichever produced the lower caseload. The instruction stated the purpose as holding a state harmless where the state served more families in the ordinary comparison year than in fiscal 2007 or fiscal 2008, so that the required rate would not increase solely because the state assisted more families during a period of increased need. The Congressional Research Service, in its 2009 report on the potential role of the block grant in the recession, described the provision as a freeze of the credit at the fiscal 2008 value or at the largest value available for fiscal 2009 through fiscal 2011.

The Excess Maintenance of Effort Adjustment

A further adjustment operated through state spending above the required maintenance of effort level. Under sections 608(g) and 609(a)(7), a state had to spend state funds at 80 percent of a historical spending level, reduced to 75 percent for a state that met the participation standards, in order to avoid a reduction in the block grant. Spending above the required level was described as excess maintenance of effort. The regulatory provision at section 261.43 of title 45 allowed a state to obtain an additional benefit from that excess spending within the credit calculation. The Congressional Research Service stated the mechanism in its 2019 primer in substance as follows, that a state could count as caseload reduction a calculated number of families assisted with nonfederal dollars spent in excess of the maintenance of effort requirement, and that the resulting credit could further reduce the share of families that had to be engaged for the standard to be met. The Administration for Children and Families expressed the same mechanism from the opposite direction in information memorandum TANF-ACF-IM-2015-01, explaining that a state investing funds in excess of the required 80 percent or 75 percent basic amount needed to include only the pro rata share of caseloads required to meet the basic requirement. In operational terms, the department estimated the number of cases that excess spending could support by dividing excess spending by average spending per case, factored the cases attributable to excess spending out of the comparison year caseload, and calculated the credit against the reduced comparison year figure, which enlarged the measured decline from the base year.

The origin of the adjustment predated the Deficit Reduction Act. The Department of Health and Human Services explained in a Federal Register notice published on August 8, 2008, that the original rule published on April 12, 1999, at volume 64 of the Federal Register beginning at page 17720, had included the provision allowing a state to exclude from the credit calculation cases on which the state had spent excess maintenance of effort, that the provision had been adopted in response to a comment on the proposed rule published in 1997, and that the intent at adoption had been to encourage states to spend above the required level. The Congressional Research Service noted in its 2013 report on welfare waivers that concern had arisen that the component had led states to search state budgets for expenditures that could be counted as maintenance of effort in order to reduce the effective standard.

Reward for Departure or Erosion of the Standard

The published record between 1996 and 2019 contained two competing characterizations of the credit. Defenders described the mechanism as a reward for the central outcome that the 1996 act sought, namely a smaller caseload of families receiving assistance. The Department of Health and Human Services, in the August 8, 2008 Federal Register notice, described recalibration under the Deficit Reduction Act as a measure that effectively raised target rates and encouraged states to help families become independent, language that treated caseload decline and movement toward independence as associated outcomes worthy of credit. Under that characterization, a state that had already moved a large share of the fiscal 1995 caseload into employment or off assistance for other reasons consistent with the statute had performed the work that the standard existed to induce, and a full engagement requirement applied to the smaller caseload that remained would penalize earlier success.

Detractors described the same arithmetic as a mechanism that emptied the standard of content. The Government Accountability Office, in report GAO-05-821 dated August 2005, documented variation in state definitions and measurement before the Deficit Reduction Act definitions took effect and treated inconsistent measurement as a threat to the credibility of the reported rates. The Center for Law and Social Policy, in comments filed in 2006 on the interim final rule, argued that the regulatory definitions restricted state flexibility and directed state effort toward monitoring participation in a limited set of countable activities rather than toward employment and poverty reduction outcomes. The factual premise shared across those positions was not in dispute in the sources cited here. Between fiscal 2003 and fiscal 2006 the weighted average effective all families standard stood between 3.9 percent and 5.0 percent according to the Department of Health and Human Services annual reports, and in fiscal 2004 a substantial group of states faced an effective standard of zero according to the Congressional Research Service. Whether a standard satisfied at zero represented earned recognition for caseload decline or absence of a binding work requirement depended on a judgment about the causes of the decline, a judgment that section 607 itself, which counted decline without inquiring into the destination of each family that left assistance beyond the eligibility exclusion in section 607(b)(3), did not supply.

Assistance and Non-Assistance: The Regulatory Distinction That Decides Which Rules Attach

The strictest federal rules in the Temporary Assistance for Needy Families program do not attach to every dollar that a state spends under the block grant. The five year time limit, the work participation rates, the assignment of child support rights, and the detailed case reporting required of states all turn on a single classification. Each of those rules, as written in the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, applies because a family receives assistance, and none applies in the same way when a state provides a benefit or service outside that term. The distinction therefore operates as a hinge. On one side sits ongoing basic support, surrounded by time limits, work tests, sanctions, and reporting. On the other side sits help ranging from an eviction payment to child care for a working parent, carrying block grant funds without the same conditions.

Congress used the term assistance repeatedly in the 1996 Act, in sections 601 to 619 of title 42, United States Code, without supplying a full definition in the statutory text. The Department of Health and Human Services supplied the definition in the final regulation published on April 12, 1999, and codified at section 260.31 of title 45, Code of Federal Regulations. A regulatory definition recorded an administrative judgment about which benefits resembled the cash aid paid under the former Aid to Families with Dependent Children program closely enough to carry the new rules, and which differed enough in purpose, duration, or form to sit outside those rules. The Congressional Research Service, in its primer, described the same point: the term is used throughout the statute, the statute does not define the term, and the Department defined the term in regulation.

Which rules attach to a benefit only if it counts as assistance?

Assistance means benefits designed to meet ongoing basic needs, while non-assistance means excluded help such as short term crisis benefits, work supports for employed families, employer subsidies, refundable tax credits, and services without basic income support, under the regulation issued by the Department of Health and Human Services in 1999.

The difference sounds narrow, because a single feature often decides the classification. Duration decides part of the boundary, since a benefit that ends within four months under the conditions in the regulation falls outside assistance, while a benefit designed for ongoing needs falls inside the term. Employment status decides another part, since child care and transportation provided to a family without employment count as assistance under section 260.31(a)(3) of title 45, Code of Federal Regulations, while the same services provided to an employed family are excluded under paragraph (b)(3). Form decides still another part, since a payment to an employer as a wage subsidy is excluded, while a payment to a family for ongoing basic needs is included even when receipt is conditioned on participation in work experience or community service. Purpose decides the remainder, since counseling and case management that provide no basic income support are excluded.

Two families can therefore receive help financed from the same block grant, in the same month, and yet stand in entirely different legal positions. One family receives a monthly grant for rent, food, and clothing, and each month adds a month to the federal time count, places the family inside the work participation calculation, and brings the child support assignment and cooperation rules into operation. Another family receives help with a car repair, rent arrears, and child care connected with employment, and none of those consequences follows, provided that the benefits fit the exclusions. The legal treatment differs because Congress attached the strictest rules to one category rather than to block grant spending as a whole.

What the Regulation Counts as Assistance

Section 260.31(a) of title 45, Code of Federal Regulations, as adopted in 1999, provides that assistance includes cash, payments, vouchers, and other forms of benefits designed to meet a family ongoing basic needs, illustrated by food, clothing, shelter, utilities, household goods, personal care items, and general incidental expenses. A benefit designed to meet those recurring household costs on an ongoing basis sits at the center of the definition, regardless of whether the state delivers the benefit as cash, as a payment, or as a voucher. The regulation also includes such benefits even when provided by another agency on behalf of a Temporary Assistance for Needy Families agency, and even when receipt is conditioned on participation in work experience, community service, or another work activity. A state accordingly cannot remove a basic needs benefit from the definition merely by routing the payment through a contractor or by attaching a work condition.

Paragraph (a)(3) extends the definition beyond cash like benefits. Except where paragraph (b) excludes a benefit, assistance also includes supportive services such as transportation and child care provided to families that are not employed. Child care for a parent without a job functions as part of the support package that sustains the family while the parent seeks work, while the same services for a parent with a job function as supports for employment already obtained. The regulation draws the line at employment, so child care for a nonemployed family can count as assistance, while child care after employment begins falls into non-assistance. A family can cross the boundary without any change in the service received, solely because the employment status of the parent changed. The deciding factors are therefore ongoing basic need, duration, employment status, recipient of the payment, and whether the benefit provides basic income support.

Paragraph (c) of the same section states that the definition does not preclude a state from providing other types of benefits and services in support of the statutory goals of the program. The definition therefore exists to decide where specified federal rules attach, not to confine the block grant to assistance alone. The purposes clause of the 1996 Act supplied the first legal mechanism through which spending could move toward activities other than basic cash aid. The assistance definition supplied the second mechanism, because spending that served those purposes through excluded benefits and services escaped the rules attached to assistance.

What the Regulation Excludes from Assistance

Section 260.31(b) of title 45, Code of Federal Regulations, as adopted in 1999, lists the benefits and services that the term assistance excludes, and each paragraph removes a distinct kind of spending from the strictest federal rules. The first exclusion, at paragraph (b)(1), covers nonrecurrent, short term benefits that meet three conditions stated together. Such a benefit must be designed to deal with a specific crisis situation or episode of need. Such a benefit must not be intended to meet recurrent or ongoing needs. Such a benefit must not extend beyond four months. The three conditions work as a unit, so a benefit that continues for a longer period, pays a recurring cost as part of an ongoing plan, or lacks a specific episode of need fails the test. The Administration for Children and Families, in examples published for state applications involving short term benefits, restated the same three criteria and treated housing help connected with arrears, utility help, emergency housing, diversion help, and vehicle repair help as illustrations when the criteria were met. In each example, a discrete event, such as a lost job, a car breakdown, or a threatened eviction, creates a need that a limited payment can resolve.

The remaining exclusions proceed by kind rather than by duration. Paragraph (b)(2) excludes work subsidies, described in the regulation as payments to employers or third parties to help cover the costs of employee wages, benefits, supervision, and training. Paragraph (b)(3) excludes supportive services such as child care and transportation provided to families that are employed, completing the contrast with paragraph (a)(3). Paragraph (b)(4) excludes refundable earned income tax credits. Paragraph (b)(5) excludes contributions to, and distributions from, Individual Development Accounts. Paragraph (b)(6) excludes services such as counseling, case management, peer support, child care information and referral, transitional services, job retention, job advancement, and other employment related services that do not provide basic income support. Paragraph (b)(7) excludes transportation benefits provided under a Job Access or Reverse Commute project, pursuant to section 404(k) of the Act as codified in title 42, United States Code, to an individual who is not otherwise receiving assistance.

The breadth of paragraph (b)(6) deserves particular attention, because services that do not provide basic income support form the most open ended category in the list. Such services can continue for substantial periods and consume substantial funds, yet none becomes assistance, because none provides basic income support. The Congressional Research Service, in its primer, observed that a state may design a benefit in a way that does not meet the definition of assistance, giving temporary help with rent for a family facing homelessness, when limited to four months or less, as an illustration. Design, in that account, is the activity that the regulation invites, since states place benefits within or outside assistance by choosing duration, recipient, and form.

The Time Limit Runs Only on Assistance

Section 608(a)(7) of title 42, United States Code, enacted as section 408(a)(7) of the Social Security Act by the 1996 Act, contains the federal time limit, and the text is addressed to states rather than to individual families. The provision bars a state from using federal funds under the block grant to provide assistance to a family that includes an adult who has received assistance for a total of sixty months, whether or not the months were consecutive. The object of the prohibition is assistance, so months in which a family received only excluded benefits or services add nothing to the count, and the bar operates as a restriction on use of the federal grant rather than as a direct termination of every form of state support. The Congressional Research Service, in its primer, described the federal five year limit as a prohibition on state use of federal funds, rather than as a direct limitation on how long a family may receive help.

The statute then builds exceptions and disregards into the count. Paragraph (a)(7)(C) permits a hardship exemption under which a state may extend assistance paid with federal funds beyond the limit for up to twenty percent of the average monthly number of families receiving assistance during the fiscal year or the immediately preceding fiscal year, a cap that the regulation at section 264.1 of title 45, Code of Federal Regulations, adopted in 1999, restates as twenty percent of the average monthly caseload. Paragraph (a)(7)(D) requires disregard of a month in which an adult received assistance while living in Indian country, as defined in section 1151 of title 18, United States Code, or in an Alaska Native village, where at least fifty percent of the adults were not employed, with the percentage determined from the most reliable data available. Section 264.1 also provides that only months of assistance paid for with federal funds, in whole or in part, count toward the limit. The count itself runs from the commencement of the state program under the 1996 Act, since the original text of section 608, as enacted in 1996 and amended in 1997, measured the sixty months after the date on which the state program commenced. Months of aid received before the program began therefore do not enter the federal count.

Because the clock runs only on assistance paid with federal funds, a state possesses several lawful routes for supporting a family after sixty counted months. A state may continue assistance with federal funds under the hardship exemption, within the twenty percent cap in paragraph (a)(7)(C). A state may also provide benefits and services that fall outside assistance, using federal funds, without adding months or consuming a hardship exemption, since excluded child care, transportation, counseling, case management, and a qualifying nonrecurrent short term benefit never entered the count. A state may also pay for ongoing basic support entirely with state funds outside the federal grant structure, spending that the bar in section 608(a)(7) does not reach. The Congressional Research Service, in its primer, likewise noted that federal law expressly allows a state to use state maintenance of effort funds to aid a family beyond the time limit.

Work, Child Support, and Sanction Rules Follow the Same Line

The work participation calculation follows the same boundary, because the denominator consists of families receiving assistance. Section 607 of title 42, United States Code, enacted as section 407 of the Social Security Act in 1996, sets numerical participation standards for states. The Congressional Research Service, in its primer, described the included group as families receiving assistance from federal funds or from state funds counted toward maintenance of effort that include a work eligible individual, subject to specified statutory exclusions. Families that receive only non-assistance therefore do not enter the denominator, and the hours and documentation rules that determine engagement in work have no occasion to operate on those families through the rate, so the population measured shrinks when spending shifts toward services.

Family rules follow the boundary as well. Section 608(a)(3) of title 42, United States Code, as enacted in 1996, requires a state, as a condition of paying assistance to a family, to require a family member to assign to the state rights to support from another person, up to the total amount of assistance paid to the family, which accrue during the period in which the family receives assistance. Assignment converts collected child support into reimbursement for assistance paid. A family that receives only counseling, child care for an employed parent, a wage subsidy paid to an employer, or a short term crisis benefit receives no assistance within the meaning of section 260.31, and the assignment condition has no assistance payment to which the condition can attach. The Center for Law and Social Policy, in an analysis of assignment for families receiving assistance financed under the block grant, framed the requirement in the same way, as a condition of receipt of assistance. Section 608(a)(2) of title 42, United States Code, addresses cooperation in establishing paternity and in establishing, modifying, or enforcing a support order, and prescribes a reduction of at least twenty five percent of assistance, and possible denial of assistance, where an individual without good cause does not cooperate, a sanction that operates on the grant rather than on receipt of services alone.

Sanction rules for refusal to engage in work complete the pattern. The 1996 Act requires states to reduce or terminate assistance, subject to protections such as the protection for a single parent caring for a child under age six who cannot obtain affordable child care, when an individual in an assistance family refuses to engage in required work. Such a sanction presupposes a grant that can be reduced or terminated, and the federal sanction architecture does not supply a mechanism for services. Recategorization of spending as services therefore escapes three federal structures at once: the sixty month clock in section 608(a)(7), the participation rate in section 607, and the assignment and cooperation rules in section 608(a)(2) and section 608(a)(3), together with the work sanction structure that depends on an assistance grant. That combined escape is the reason that the assistance distinction ranks as the second legal mechanism, after the purposes clause, by which spending moved away from basic cash aid after 1996. The purposes clause widened the ends that funds could serve, and the assistance distinction lightened the conditions that attached when funds served those ends through services, work supports, tax credits, and short term benefits.

Reporting Follows Assistance, and the Record Narrows With It

Section 611 of title 42, United States Code, enacted as section 411 of the Social Security Act in 1996, imposes the detailed reporting duty, framed around families receiving assistance. The section requires each eligible state to collect information monthly and to report information to the Secretary of Health and Human Services quarterly, as disaggregated case record information on families receiving assistance under the state program, or under another state program funded with qualified state expenditures. The items specified include county of residence, ages of family members, family size and relation to the head of the family, employment status and earnings of an employed adult, marital status, race and educational level of adults and children, receipt of subsidized housing, medical assistance, nutrition assistance, and subsidized child care, and the number of months that the family received each type of assistance. The Administration for Children and Families used the collection, through the TANF Data Report administered by the Office of Family Assistance, to assess whether state programs met the participation standards set in section 607.

No comparable case record duty in section 611 attaches to a family that receives only services or other non-assistance. Aggregate financial reporting continued to capture expenditures by category, and the Administration for Children and Families published financial data of that kind for fiscal 2016 in February 2018. Expenditure totals, however, do not identify the families served, the duration of service receipt, or the earnings or education of the adults served, and section 611 does not require case records that would fill those gaps. The Congressional Research Service, in its primer, stated that reporting on the number of families served by funds used for benefits that are not assistance was limited, that states had to provide a total number served for programs financed with maintenance of effort funds but were not required to report the number served with federal funds where the benefit was not assistance, and that no federally prescribed performance measures evaluated the use of funds for benefits that are not assistance. Recategorization changed the federal record as well as the rules.

The two readings of the distinction, stated with attribution, reflect that reporting gap and the rule gap that accompanies it. Officials at the Department of Health and Human Services, in adopting the final regulation of April 12, 1999, structured the categories so that crisis help could be delivered without starting a time count, work supports could follow a parent into employment, and services could be offered without the assignment, participation, and sanction consequences attached to ongoing basic aid. Analysts at the Center on Budget and Policy Priorities, in a chart book on the block grant issued in 2017, described a different consequence of state flexibility, arguing that states had used broad discretion to spend a shrinking share of funds on basic assistance for families in poverty and a growing share elsewhere. The Congressional Research Service, in its primer, noted the combination that connects the critical reading to the text: few federal requirements, limited reporting, and no performance metrics for activities that are not assistance. No verdict between those readings follows from the text alone, which establishes that the classification decides which rules attach, which months count, which families enter a rate, which support rights must be assigned, and which families appear in the federal record.

Maintenance of Effort, Penalties, and Where the Combined Money Went

The federal grant was only half of the financing bargain. The other half was a condition placed on state budgets, enforced through penalties, and visible only when federal and state dollars were counted together. A state could design benefit levels and service programs with wide discretion, but a state could not withdraw its own money and keep the federal grant intact. The maintenance of effort requirement fixed a floor under state spending, the penalty provisions at 42 U.S.C. 609 supplied the sanctions for missing the floor or the work participation rates, and the financial reports filed with the Administration for Children and Families recorded what the combined pool purchased.

Why does the maintenance-of-effort floor not guarantee any level of cash benefits?

Maintenance of effort is the minimum amount of state money a state must spend each year on qualified activities for needy families to receive a full federal TANF grant, set at 80 percent of state spending in fiscal 1994, or 75 percent when both work rates are met.

The requirement appears at 42 U.S.C. 609(a)(7), and the Congressional Research Service primer on TANF financing, prepared by Gene Falk, describes the threshold in the same terms in editions issued before October 1, 2019. Each state measured the obligation against the state share of spending in fiscal 1994 on the programs TANF replaced, namely Aid to Families with Dependent Children, Emergency Assistance, the Job Opportunities and Basic Skills program, and AFDC related child care. A state that met both the all families work participation rate and the separate two parent rate qualified for the lower threshold of 75 percent. A state that missed either rate remained at 80 percent for that year.

The floor was a floor on spending for defined purposes, not a floor on cash benefits, and that distinction governed the spending record. Qualified state expenditures, as the Congressional Research Service summarized the rules, were expenditures for eligible families on benefits, services, and activities reasonably calculated to accomplish a purpose of TANF, together with administrative costs subject to the separate administration rules. An eligible family was in general a family with a child living with a parent or other caretaker relative, or a pregnant woman, that met the financial need standard the state had set in its own plan. Within that frame, counted spending reached well beyond monthly cash payments. Child care for working families counted, as did transportation and other work supports, education and training, family support and preservation services, refundable state tax credits, and programs addressed to pregnancy prevention and two parent family formation, when the statutory conditions for each category were met. The breadth was deliberate, because the purposes stated in section 601 reached beyond income support to ending dependence through work and marriage, reducing out of wedlock pregnancies, and encouraging two parent families, and spending addressed to the third and fourth purposes was not confined to families meeting a test of financial need.

A state could also meet part or all of the requirement through separate state programs financed entirely with state dollars and operated outside the TANF structure, provided the spending served eligible families and a TANF purpose and was reported as the rules directed. The Administration for Children and Families financial tables for fiscal 2017 and fiscal 2018 therefore presented state spending inside TANF and in separate state programs before combining the two into a single total. The mechanism mattered for interpretation, because programs that predated the block grant, including state earned income tax credits, prekindergarten programs, and child welfare services, could be brought inside the ledger when the statutory tests were satisfied. The ledger measured whether qualifying dollars had been spent, not whether the dollars were new dollars the block grant had caused to be spent, and that gap returned in the debate over substitution addressed later in this section.

How the Penalty Schedule Was Built

The penalty structure at 42 U.S.C. 609 covered a catalogue of failures, from misuse of funds to failure to enforce the five year time limit, but two penalties did the main work: the penalty for failure to meet a work participation rate and the penalty for failure to satisfy maintenance of effort. The Congressional Research Service primer table stated the participation penalty as a reduction of the state family assistance grant by a maximum of 5 percent for the first year of failure, with the maximum increased by 2 percentage points for each subsequent consecutive year of failure, subject to a ceiling of 21 percent of the adjusted grant.

The maintenance of effort penalty operated on different arithmetic. Under 42 U.S.C. 609(a)(7), as the Congressional Research Service primer stated the rule, a state that failed to meet the requirement had its grant reduced dollar for dollar by the shortfall in qualified expenditures below the required level. No percentage cap and no escalation schedule applied, because the sanction was restitutionary in form: combined resources fell by twice the shortfall in practical effect, once through the missing state dollars and once through the matching federal reduction. Administration for Children and Families penalty guidance summarized the sanction in the same dollar for dollar terms. A state could not economize by spending less of its own money, since the grant fell by the amount saved, and a state could not economize by accepting a participation penalty, since the withheld amount had to be replaced with state funds.

Between exposure and imposition stood two safety valves, described by the Congressional Research Service frequently asked questions report on TANF in the version updated in 2019. A penalty could be forgiven where a state claimed reasonable cause and the Secretary of Health and Human Services found that reasonable cause existed. A penalty could also be forgiven where a state entered into a corrective compliance plan and subsequently met the standard or cured the failure within the period the plan allowed. The Administration for Children and Families table of penalty provisions recorded reasonable cause and corrective compliance as available for the participation penalty, while the maintenance of effort failure stood apart, with the dollar for dollar reduction not listed as subject to either form of relief, a distinction consistent with the restitutionary character of that sanction. Congressional Research Service accounts of penalty determinations through 2018 documented repeated use of corrective compliance plans and negotiated reductions after the Deficit Reduction Act of 2005 recalibrated the standards, so assessed penalties were frequently lower than the statutory maxima.

Where the Combined Money Went in the Reported Data

The spending evidence came from reports states filed on Form ACF 196 and successor forms, compiled by the Office of Family Assistance. The fiscal 2018 compilation, released on August 7, 2019, was the most complete federal reporting available at the October 1, 2019 vantage of this article, and fiscal 2017 supplied the comparison year. Combined federal TANF and state maintenance of effort expenditures and transfers totaled 31.1 billion dollars in fiscal 2017, according to the Congressional Research Service summary, and 31.3 billion dollars in fiscal 2018, according to the Administration for Children and Families release of August 7, 2019. Basic assistance, defined as cash payments, vouchers, and other benefits meeting a family need for food, clothing, shelter, utilities, household goods, personal care items, and incidental expenses, stood at roughly a fifth to just under a quarter of the pool. Administration for Children and Families Table A.2 placed basic assistance at 22.7 percent of total funds used in fiscal 2017, at 7.0 billion dollars, and at 21.4 percent in fiscal 2018, at 6.7 billion dollars, a decline of 337 million dollars between the two years. The Congressional Research Service, rounding the fiscal 2017 figure, stated that basic assistance represented 23 percent of fiscal 2017 dollars, and Center on Budget and Policy Priorities analysis of the same federal data likewise put basic assistance at 23 percent in fiscal 2017.

The contrast with the opening years was sharp. In fiscal 1997, the first year of TANF operation, states spent 71 percent of TANF funds on basic assistance, a share the Center on Budget and Policy Priorities reported in analysis of the federal spending data. Between fiscal 1997 and fiscal 2018 the basic assistance share fell by approximately fifty percentage points while the combined pool remained broadly stable in nominal dollars, so the movement represented a genuine reallocation of a roughly constant pool rather than dilution by growth in total spending. An account of where the money went is therefore an account of the categories that absorbed that reallocation, and an assessment of the reform as a whole also depends on what the outcomes evidence shows about the reform, which the companion article in this series addresses on its own evidence.

Child care was the largest single destination for the reallocated dollars. The Administration for Children and Families fiscal 2018 release reported that 17.0 percent of combined funds was used for child care in fiscal 2018, including funds transferred to the Child Care and Development Fund, while Table A.1 separated direct child care spending, at 12.2 percent and 3.8 billion dollars, from transfers to the Child Care and Development Fund, at 4.8 percent. For fiscal 2017, the Congressional Research Service reported that 5 billion dollars, or 16 percent of combined funds, was either expended on child care or transferred to the Child Care and Development Fund. The transfer authority came from 42 U.S.C. 604(d), under which, as the Congressional Research Service primer stated the rule, overall transfers to the Child Care and Development Fund and the Social Services Block Grant combined were limited to 30 percent of the federal grant, with a separate limit that no more than 10 percent could be transferred to the Social Services Block Grant. Transferred funds became subject to the rules of the receiving grant, with the condition that funds transferred to the Social Services Block Grant be used for families with incomes below 200 percent of the poverty line. In fiscal 2018, transfers to the Social Services Block Grant accounted for 3.6 percent of combined funds in Table A.1, and total transfers accounted for 8.4 percent. Early care and education as a whole reached 20.5 percent in fiscal 2018, with prekindergarten and Head Start alone at 8.3 percent and 2.6 billion dollars, financed overwhelmingly with state maintenance of effort dollars in that year.

Child welfare services formed a second major destination, although the reporting divided the category across several lines. Table A.1 for fiscal 2018 reported child welfare services, covering family support, family preservation and reunification, adoption services, additional child welfare services, and home visiting, at 5.3 percent and 1.7 billion dollars. Adjacent lines authorized solely under prior law added foster care payments at 0.8 percent and child welfare or foster care services at 1.4 percent, while relative foster care maintenance payments and adoption and guardianship subsidies carried inside basic assistance added 1.4 percent. The Congressional Research Service, using a consolidated definition, reported that TANF contributed about 2.2 billion dollars to the child welfare system in fiscal 2017, for foster care, adoption assistance, and services to families whose children had experienced or were at risk of experiencing abuse or neglect.

Refundable state tax credits formed a third destination of comparable size, consisting almost entirely of state dollars. Table A.1 for fiscal 2018 reported refundable earned income tax credits at 7.3 percent and 2.3 billion dollars, and non EITC refundable credits at a further 1.7 percent, so refundable credits together accounted for 9.0 percent in fiscal 2018. The Congressional Research Service reported the fiscal 2017 figure as 2.8 billion dollars, or 9 percent. A state operating a refundable credit for low income working families, most commonly a state supplement to the federal earned income tax credit, could count the refundable portion paid to families toward maintenance of effort and could apply federal dollars to a portion of the cost. Work activities were smaller than any of the destinations already described. The fiscal 2018 release reported work, education, and training activities at 10.7 percent, divided in Table A.1 among subsidized employment at 0.4 percent, education and training at 5.5 percent, and additional work activities, including work experience, community service, job search, and job readiness, at 4.7 percent. Work supports, covering transportation and related goods and services, added 1.4 percent in fiscal 2018, and the Congressional Research Service placed work related activities at 11 percent in fiscal 2017. Nonrecurrent short term benefits, including diversion payments and emergency help with a specific crisis, accounted for 3.3 percent in fiscal 2018, while prevention of out of wedlock pregnancies accounted for 1.5 percent and fatherhood and two parent family formation and maintenance programs accounted for 0.7 percent.

Program management completed the picture, and the reporting definition was broader than the statutory cap on administrative costs. Table A.1 for fiscal 2018 reported program management at 9.9 percent and 3.1 billion dollars, divided among administrative costs at 6.3 percent, assessment and service provision at 2.9 percent, and systems costs at 0.7 percent. The statutory limit operated on the narrower concept: under 42 U.S.C. 604(b), as summarized in federal grant terms and in 45 C.F.R. section 263.13, a state could spend no more than 15 percent of its state family assistance grant on administrative costs, a definition that excluded direct assessment and service provision costs grouped inside program management in the reporting form. Reported administrative costs of 6.3 percent therefore sat inside the 15 percent ceiling, and the gap was a gap of definition rather than a sign of noncompliance.

Substitution, Flexibility, and the Accountability Argument

The reported categories settled where the money was recorded as going, not what the money purchased that would not otherwise have been purchased, and the difference was the substitution problem. Because money is fungible, a state could satisfy maintenance of effort, and could spend federal dollars, on activities the state had financed from its own budget before 1996 or would have financed in the absence of the block grant. Where a state counted preexisting child welfare spending toward maintenance of effort, the federal ledger showed child welfare spending inside the TANF structure, but no new child welfare effort had necessarily occurred; state dollars released by the substitution became available for purposes outside the TANF ledger that no federal report captured. The same logic applied to prekindergarten spending and to state tax credits that predated inclusion in the TANF accounting. The Government Accountability Office and the Congressional Research Service treated countable preexisting spending as a structural feature of a floor with no general incrementality requirement. The shares reported for child welfare, prekindergarten, and tax credits therefore measured the composition of the counted pool, and the counted pool overstated, to an extent the reports could not quantify, the new effort the block grant and the floor together had induced.

The flexibility defense answered that the overstatement framing misunderstood the design. The statute stated its purpose in 42 U.S.C. 601 as increasing the flexibility of states in operating a program designed to meet the four listed goals, and state officials and the associations that represented them defended the pattern as the design working as Congress had written it. The National Governors Association, whose February 1996 welfare reform proposal had helped shape the block grant, had sought state discretion over the use of funds as a central objective, and governors and state administrators defended broad countable spending on the ground that a family need for income support, child care, transportation, child welfare services, and early education presented itself as a single set of connected needs rather than as separate federal categories. The American Public Human Services Association, representing state and local human service agencies, advanced the related argument in reauthorization debates documented by the Congressional Research Service and the Government Accountability Office in the period before 2019: a structure allowing a state to braid TANF dollars with child welfare, child care, and early education dollars allowed the state to build a broader family support system, to serve families before a crisis produced a cash assistance case, and to continue services for working families whose earnings had ended cash eligibility but whose need for child care and work supports persisted. On that account, the decline in the basic assistance share was partly a consequence of caseload decline itself, since fewer recipient families required fewer cash dollars, and redirection toward child care, tax credits that raised the return to work, and services that kept children safely in their own homes represented the reinvestment the architects of the block grant had predicted.

The accountability criticism, stated at full strength by named analysts at the Center on Budget and Policy Priorities, answered that the reinvestment account could not be squared with the reported totals. Liz Schott, LaDonna Pavetti, and Ife Floyd of the Center on Budget and Policy Priorities, in Center analysis published in 2018 on state TANF spending and family access, documented that states spent less in each core area of basic assistance, child care, and work programs in the later reporting years than in the early years of TANF, and that only about half of combined funds in fiscal 2014 went to core areas combined, with eight states spending one quarter or less there. The Center analysis of June 2016, Why TANF Is Not a Model for Other Safety Net Programs, stated the substitution charge directly: states had used diverted funds to fill budget holes, to fund services for families with incomes above the poverty line, to bolster child welfare systems, and to support early education and higher education, and in some cases TANF dollars had supplanted existing state spending, which enabled states to shift even more funds to other uses. Schott and Pavetti argued, in Center work on TANF performance during and after the recession of 2007 to 2009, that broad flexibility had left the cash safety net unable to respond when need rose, because dollars committed to other categories could not readily be reclaimed and because the fixed grant gave states an incentive to hold caseloads down. On that account, the fiscal 2018 figure of 21.4 percent was the measurable residue of a process in which the cash function at the center of the predecessor program had been allowed to contract while the accounting categories continued to report a full combined pool. The Center placed a human measure alongside the fiscal measure, reporting that in 2017 TANF reached 23 families for every 100 families with children living in poverty, against 68 for every 100 in 1996.

The two accounts disagreed about evaluation and about remedy, but both rested on a common body of reported fact. The flexibility account drew on the text of 42 U.S.C. 601, on the breadth of the qualified expenditure definition, and on the transfer authority in 42 U.S.C. 604(d), each enacted in express terms. The accountability account drew on the Administration for Children and Families tables for fiscal 2017 and fiscal 2018, on Center tabulations of the same federal data reaching back to fiscal 1997, and on the proposition that a floor defined by countable purposes rather than by cash outcomes would predictably produce the recorded pattern. The penalty structure stood between the accounts as the limiting case of the design: Congress had specified exact percentages, escalation steps, and a dollar for dollar sanction for the two obligations chosen for numerical enforcement, participation and aggregate state effort, and had specified no corresponding numerical floor for the share of the combined pool that reached families as cash. The record through fiscal 2018 showed states complying, in aggregate, with the obligations that carried numerical sanctions, and exercising, in aggregate, the discretion that carried none.

The Provisions Most Often Misread, and How the Text Has Been Kept Alive

Five readings of the Temporary Assistance for Needy Families block grant circulated so widely after enactment that each acquired the force of a summary of the law itself. Each fastened on a real phrase or a real number, and each stopped before the qualifying provisions that gave that phrase or number its legal effect. The corrections therefore come from the text: the purposes clause in section 401 of the Social Security Act, codified at 42 U.S.C. 601, the work standard and caseload reduction credit in section 407, the regulatory definition of assistance, the time limit and hardship exemption in section 408, the fixed grant amount in section 403, and the transfer, administrative cost, and penalty provisions. After the extension in the Deficit Reduction Act of 2005 expired at the end of fiscal 2010, Congress funded the program through short measures rather than through a comprehensive reauthorization, and the operative text remained substantially the text written in 1996 and adjusted in 2005. That distinction runs through the myths and contested claims about welfare reform examined one by one and through the five corrections that follow.

Whether the Grant Must Be Spent on Cash Assistance

The first reading held that the block grant had to be spent on cash assistance to poor families, so that every dollar not paid as a monthly benefit to a family meeting a needs test represented a diversion from the statutory purpose. The reading drew support from the first purpose in section 401, which the Congressional Research Service primer on financing and federal requirements summarized as providing assistance to needy families so that children could be cared for in their own homes or in the homes of relatives. The second purpose used the same population, describing an end to the dependence of needy parents on government benefits through job preparation, work, and marriage. On a reading confined to those two purposes, a needs test appeared to govern the whole grant.

The remaining text removed that appearance. The third purpose, prevention and reduction of out of wedlock pregnancies, contained no reference to needy families, and the fourth purpose, encouragement of the formation and maintenance of two parent families, contained no such reference either. The Congressional Research Service primer drew the consequence directly: activities directed toward the third and fourth purposes could be offered without a determination that each participant family was needy, while assistance under the first purpose and activities under the second remained tied to needy families as states defined need within federal limits. Section 404 reinforced the point, authorizing use of the grant in any manner reasonably calculated to accomplish the purpose of the part, an authority the Congressional Research Service described as extending to benefits and services far broader than monthly cash assistance, including child care, work supports, and refundable tax credits administered by a state. Cash assistance occupied the first purpose; the statute did not confine the grant to cash assistance, and no single needs test ran across all four purposes.

What the Federal Work Standard Actually Requires

The second reading held that the federal work requirement meant that half of every state caseload had to be working, so that a state in which fewer than half of recipient families were engaged in work had failed the federal standard. The number at the center of the reading was real. Section 407 set a participation standard of 50 percent for all families receiving assistance and a separate standard of 90 percent for the two parent portion of the caseload, measured by participation in specified activities for a minimum number of hours. The Congressional Research Service, in the report updated on July 1, 2019, stated those figures as the statutory standards and added the first qualification that the reading omitted: the main federal work requirement operated as a performance measure applied to states, and states determined the work rules that applied to individual recipients. A state was measured, in the aggregate, against a standard that the statute then reduced.

The reduction mechanism was the caseload reduction credit in section 407. Under the credit as described by the Congressional Research Service in the July 1, 2019 report, the standard fell by one percentage point for each percentage point decline in a state caseload, so that a state whose caseload had fallen by 30 percent faced an all families standard of 20 percent rather than 50 percent. Regulation added extra credit where a state spent more than the amount required under the maintenance of effort requirement, further lowering the effective standard. The Deficit Reduction Act of 2005 changed the measurement period rather than the arithmetic: Congressional Research Service analysis of the work requirements recorded that the 2005 act rebased the credit effective in fiscal 2007 so that reductions were measured from fiscal 2005 instead of from fiscal 1995, the base year used under the 1996 law. The denominator also mattered, because families without a work eligible individual were excluded from the calculation, including child only families cared for by a nonparent relative who received no assistance on behalf of that relative and families in which the parent was a nonrecipient, such as a parent receiving Supplemental Security Income. A state could therefore satisfy the federal standard with an observed participation rate well below 50 percent. The headline figure described the starting point of a calculation, not the obligation itself.

What Happens When the Clock Reaches Sixty Months

The third reading held that the five year limit ended all help to a family at sixty months. Section 408 supplied the number: a state could not use federal funds to provide assistance to a family that included an adult or a minor child head of household who had received sixty cumulative months of federally funded assistance. The Congressional Research Service primer stated the character of that prohibition with precision. The federal five year limit was a prohibition on the use of federal funds, not a direct limitation on how long a particular family could receive assistance, and the consequences ran through three provisions that the circulating reading collapsed into one.

The first provision concerned the definition of assistance, a term defined in regulation at 45 C.F.R. 260.31 around payments directed at ongoing basic needs. Benefits classified as non assistance did not start the same clock. The clearest category, described in Congressional Research Service analysis of the program during the recession following the financial crisis of 2008, was the nonrecurrent short term benefit: aid designed to deal with a specific crisis situation or episode of need, not intended to meet recurrent or ongoing needs, and not extending beyond four months. A family that received only such a benefit was not, by that receipt alone, subject to the time limit, was not included in the work participation calculation, and was not placed under the child support assignment rules that attached to assistance. The second provision was the hardship exemption in section 408. Federal law permitted a state to continue federally funded assistance beyond sixty months on the basis of hardship, as defined by the state, or where a family included an individual who had been battered or subjected to extreme cruelty, for up to 20 percent of the average monthly number of families receiving assistance, an allowance implemented at 45 C.F.R. 264.1 against the caseload of the fiscal year or of the preceding fiscal year, at state election. The third provision concerned the source of funds. Federal law explicitly allowed a state to use state funds counted toward the maintenance of effort requirement to aid a family beyond the time limit, and the Congressional Research Service primer recorded practice built on that allowance, including assistance continued with state funds in New York and assistance continued on behalf of children alone after aid on behalf of adults ended in California. The sixty month figure was therefore a federal funding rule with a defined term, a statutory exemption for a fifth of the caseload, and a state funding route around it.

How far had the grant’s real value fallen by fiscal 2018?

The basic grant amount set in the 1996 law remained fixed in nominal terms through fiscal 2018, apart from a small reduction beginning in fiscal 2017 to finance research. Congressional Research Service tables updated on July 1, 2019, placed the fiscal 2018 grant 36.2 percent below fiscal 1997 in real terms.

The fourth reading, stated in its strongest form, held that the grant had been raised as costs rose, or that Congress had adjusted the amount to preserve its value, so that the program of fiscal 2018 rested on funding broadly equivalent to the funding of fiscal 1997. The text and the tables allowed no such equivalence. The state family assistance grant established in section 403 totaled 16.567 billion dollars for the fifty states, the District of Columbia, territories, and tribes in fiscal 1997, and the Congressional Research Service table in the report updated on July 1, 2019, carried that same nominal figure through fiscal 2016. Beginning with fiscal 2017, the amount was reduced by 0.33 percent to finance research, evaluation, and technical assistance, producing a figure of 16.512 billion dollars in fiscal 2017 and again in fiscal 2018. No inflation adjustment, caseload adjustment, or poverty population adjustment was written into the basic grant during that period, and the Congressional Research Service stated that the amount was not adjusted for inflation, for the size of the assistance caseload, or for changes in the poverty population. Converted into constant fiscal 1997 dollars with the Consumer Price Index for All Urban Consumers, the fiscal 2018 grant was valued at 10.564 billion dollars in the July 1, 2019 tables, a cumulative decline of 36.2 percent from fiscal 1997, slightly more than one third. Separate appropriations for defined purposes, including supplemental grants, contingency funds, and the temporary emergency fund of fiscal 2009 and fiscal 2010, rested on distinct authority and did not rewrite the fixed figure in section 403.

Whether States May Spend the Grant on Anything at All

The fifth reading held that a block grant of the breadth described in section 404 could be spent on anything at all, so that the purposes clause operated as a formality. The breadth was real, and the limits were equally textual. The spending authority contained the first limit: a state could use the grant in any manner reasonably calculated to accomplish the purpose of the part, the governing test for federal funds and for state spending counted toward maintenance of effort in the Congressional Research Service primer. An expenditure with no reasonable calculation toward one of the four purposes in section 401 fell outside the authority that created the grant. The second limit was quantitative. The statute capped administrative costs at 15 percent of the federal grant funds used and at 15 percent of nonfederal dollars counted toward maintenance of effort, with an exclusion for information technology costs associated with tracking and monitoring recipients. The third limit governed movement of money out of the program. A state could transfer a portion of its grant to the Child Care and Development Block Grant and to the Social Services Block Grant, subject to a combined ceiling of 30 percent of the grant and a subceiling of 10 percent for the Social Services Block Grant, as summarized in Department of Health and Human Services reporting on state spending in fiscal 2015 and in Congressional Research Service accounts of section 404.

The fourth limit was the penalty structure in section 409. Under the penalty table in the Congressional Research Service primer, misuse of grant funds established on audit drew a penalty equal to the amount misused, with an additional penalty of 5 percent of the adjusted grant where the misuse was intentional. Failure to satisfy the work participation standard drew a reduction of up to 5 percent for the first year of failure, increasing by two percentage points for each subsequent year to a maximum of 21 percent. Failure to meet the maintenance of effort requirement drew a dollar for dollar reduction of the grant payable, and failure to enforce the five year time limit drew a penalty of 5 percent of the adjusted grant. Reasonable cause exceptions and corrective compliance plans softened the application of several penalties without removing the penalties from the statute. The grant was therefore a program of broad discretion inside enforceable boundaries.

How Short Extensions Kept an Older Text in Force

The authorization history after fiscal 2010 explained why the operative text retained the shape given to the text in 1996 and 2005. The Deficit Reduction Act of 2005, Public Law 109 171, had extended funding through the end of fiscal 2010, had made changes to the work standards, and had established competitive grants for healthy marriage and responsible fatherhood initiatives, as the Congressional Research Service legislative history recorded. No comparable enactment followed before the fiscal 2010 expiration. From the end of fiscal 2010, that history described the program as funded again by a series of short term extensions, a pattern recorded for an earlier period in different numbers: between fiscal 2002 and fiscal 2006, authority had been extended on a temporary basis 13 times before the 2005 act intervened. The post 2010 series was distinct in its vehicles and its dates.

The first measures of the later series were continuing resolutions: Public Law 111 242, enacted on September 30, 2010, extended funding through December 3, 2010, and Public Law 111 290, enacted on December 4, 2010, extended authority through December 18, 2010. The Claims Resolution Act of 2010, Public Law 111 291, enacted on December 8, 2010, then extended basic funding through the end of fiscal 2011, September 30, 2011. That act, a measure directed in principal part to settlement of claims against the federal government, also divided the combined 150 million dollar appropriation for healthy marriage and responsible fatherhood programs into 75 million dollars for each program, required one time reports from states, and funded supplemental grants only through June 30, 2011, with no supplemental grant funding thereafter. Brief extensions carried forward: the Short Term TANF Extension Act, Public Law 112 35, enacted on September 30, 2011, extended basic funding for three months; the Temporary Payroll Tax Cut Continuation Act of 2011, Public Law 112 78, enacted on December 23, 2011, extended basic funding for two months; and the Middle Class Tax Relief and Job Creation Act of 2012, Public Law 112 96, extended funding through the end of fiscal 2012 while also amending the statute permanently to require state action against withdrawal of cash assistance at automated teller machines in casinos, liquor stores, and establishments providing adult entertainment. Extensions for fiscal 2013 and fiscal 2014 arrived inside continuing resolutions and an omnibus appropriations act, and the fiscal 2017 Consolidated Appropriations Act, Public Law 115 31, extended funding for the remainder of fiscal 2017 and for fiscal 2018. In 2019, the TANF Extension Act of 2019, Public Law 116 4, enacted on January 24, 2019, extended funding through June 30, 2019, and Public Law 116 27, enacted on July 5, 2019, extended funding through September 30, 2019. The Further Continuing Appropriations Act, 2020, Public Law 116 59, enacted on September 27, 2019, then extended funding through November 21, 2019, continuing the same pattern of short term measures.

Most extension measures changed funding dates and little else. Because no comprehensive reauthorization replaced the underlying sections, the work standards as adjusted in 2005, the purposes as written in 1996, the fixed grant amount, the time limit, and the penalty structure remained the operative law that a state plan had to satisfy, while policy debate in Congress addressed subjects that the frozen text treated only partially or not at all. The exceptions proved the pattern by their dates and their narrowness. The American Recovery and Reinvestment Act of 2009, Public Law 111 5, enacted on February 17, 2009, created a temporary Emergency Contingency Fund of 5 billion dollars for fiscal 2009 and fiscal 2010, reimbursing states, territories, and tribes for 80 percent of increased costs of basic assistance, nonrecurrent short term benefits, and subsidized employment associated with the recession that followed the financial crisis of 2008, as the Congressional Research Service legislative history recorded. The fund expired by its own terms at the end of fiscal 2010. Dated amendments at the margins, attached to settlement legislation, payroll tax legislation, continuing resolutions, and omnibus appropriations acts, kept the statute funded without reopening the statute as a whole.

Tribal Programs Under a Separate Plan

One part of the program operated from enactment under a distinct plan structure rather than under a state plan. Section 412 of the Social Security Act, codified at 42 U.S.C. 612, authorized direct funding and administration by Indian tribes. Under that section, a federally recognized Indian tribe with an approved tribal family assistance plan received a tribal family assistance grant, and the grant payable to a state under section 403 was reduced by the portion attributable to the tribal service area, so that the tribal grant was carved from the state grant rather than appropriated as an addition to the national total. The amount was determined from federal payments for fiscal 1994 attributable to expenditures, other than child care expenditures, for Indian families residing in the service area identified in the tribal plan.

The federal frame for tribal programs differed from the state frame in the two subjects that generated persistent misreadings for states. The Congressional Research Service primer stated that tribes submitted plans to the Secretary of Health and Human Services for approval, that the Secretary, with the participation of the tribes, established work requirements and time limits for each tribe, and that tribal programs were not subject to the same work participation standards and five year time limit as state programs. Tribal programs therefore set work rules and time limits through negotiation and plan approval inside the federal statute, rather than through the 50 percent and 90 percent standards and the sixty month prohibition that governed state programs.

Reading the Block Grant as a Whole

A sound reading of Title IV-A begins where the part begins, with section 601, and resists the habit of starting with a benefit amount or a headline work rate. The purposes clause supplies the spending constitution for the grant, because section 604 allows use of the grant in any manner reasonably calculated to accomplish a purpose stated in section 601, and the reading note that governs the whole part follows from the placement of a single word. Needy appears in the first purpose and in the second purpose, and needy does not appear in the third purpose or in the fourth purpose. That placement explains why basic assistance for a needy family and a pregnancy prevention program for a general audience can both rest on the same grant, why the share of combined funds recorded as basic assistance could fall to 21.4 percent in fiscal 2018, as the Administration for Children and Families reported on August 7, 2019, without a change to the purposes clause itself. The no individual entitlement sentence in section 601(b) completes the opening settlement: broad authority for a state, bounded by purposes and by express provisions elsewhere in the part, and no federal right to assistance for an applicant under a program funded under the part.

From that opening, the text maps in a deliberate order. Section 602 turns purpose into a state plan built on outline and certification rather than on a federally approved benefit schedule. Section 603 turns history into money, fixing a grant from predecessor spending in fiscal 1992 to fiscal 1995, repeating that amount without adjustment, and adding smaller streams for economic need, for distribution, for marriage and fatherhood initiatives, for tribes, and for the territories. Section 607 turns accountability into measurement, setting standards of 50 percent and 90 percent and then subtracting the caseload reduction credit, so that the effective standard rather than the printed standard decides whether a penalty under section 609(a)(3) has any application. Section 608 and the regulation at 45 C.F.R. 260.31 turn the word assistance into a boundary, attaching the sixty month limit, assignment and cooperation rules, and sanctions to ongoing basic support and leaving nonrecurrent short-term benefits, work supports, tax credits, and services without basic income support outside that boundary. Section 609 turns state effort into a floor at 80 percent or 75 percent of the fiscal 1994 level and supplies penalties that include a dollar for dollar reduction for a maintenance of effort shortfall, while section 611 turns reporting toward families receiving assistance and leaves families receiving only services with a far thinner federal record. Section 612 then removes tribal programs from the state frame, with work requirements and time limits set through plan approval.

That map also explains the pattern of change between 1996 and 2019. The nominal grant did not move, the purposes did not change, and the assistance boundary did not move, while prices rose, caseloads fell and then faced pressure during the recession that began in December 2007, and Congress recalibrated the credit base year in the Deficit Reduction Act of 2005 and funded the program after fiscal 2010 through a series of short extensions rather than through a comprehensive reauthorization. Each major feature of the reported record by fiscal 2018 traces to a provision rather than to accident: erosion of 36.2 percent in real terms from fiscal 1997, in Congressional Research Service tables updated July 1, 2019, traces to a fixed nominal amount in section 603; a low effective work standard in the first decade traces to a credit measured from a caseload peak in fiscal 1995; growth in child care, prekindergarten, child welfare services, and refundable credits as shares of a combined pool of 31.3 billion dollars in fiscal 2018 traces to purposes without a need condition, to a broad definition of qualified state expenditures, and to spending classified as non-assistance. A disciplined method for how to read a federal statute therefore treats each reported outcome as a question about a clause, a definition, a formula, or a penalty, and returns to the text to locate the provision that produced that outcome. For study and drafting alike, one practical habit is to keep your statute notes, citations, and case chronologies together free on VaultBook, so that the purposes clause, the credit formula, and the spending tables remain in one file rather than scattered across separate records. The block grant rewards that discipline, because the operative law sits less in any single number than in the relation among purpose, definition, formula, and sanction across sections 601 to 619.

Frequently Asked Questions

Q: How is the TANF block grant written?

The block grant is written as Title IV-A of the Social Security Act, codified at 42 U.S.C. 601 to 619, in a sequence that begins with purpose and ends with enforcement. Section 601 states four purposes and denies an individual entitlement, section 602 requires a state plan built on outline and certification, and section 603 fixes the grant amount. Section 607 sets work participation standards and the caseload reduction credit, section 608 attaches time limits and family rules to assistance, and section 609 imposes maintenance of effort and penalties. Section 611 governs reporting, section 612 authorizes tribal programs under a separate plan, and section 604 allows use of the grant in any manner reasonably calculated to accomplish a statutory purpose.

Q: What are the four purposes of TANF?

Section 601(a) lists four purposes in order. The first purpose is assistance to needy families so that children may be cared for in their own homes or in the homes of relatives. The second purpose is an end to the dependence of needy parents on government benefits through promotion of job preparation, work, and marriage. The third purpose is prevention and reduction of the incidence of out-of-wedlock pregnancies, with establishment of annual numerical goals for preventing and reducing that incidence. The fourth purpose is encouragement of the formation and maintenance of two-parent families. The Congressional Research Service, in a primer issued in 2018, presented state flexibility as the overarching frame and the four purposes as the ends toward which that flexibility may be exercised.

Q: Why can TANF money go to families who are not poor?

TANF money can reach families without a test of economic need because the third and fourth purposes contain no need condition. The first purpose covers needy families and the second purpose covers needy parents, while the third purpose, prevention and reduction of out-of-wedlock pregnancies, and the fourth purpose, encouragement of two-parent families, name no needy population at all. The Congressional Research Service, in analysis published in 2018, stated that activities funded to achieve the third and fourth purposes were not limited to needy families, could be provided to the population as a whole, and could reach families without children. A pregnancy prevention campaign for adolescents or marriage education for engaged couples therefore rests on the text without a means test.

Q: Has the TANF grant amount changed since 1996?

The basic grant amount remained fixed in nominal terms from fiscal 1997 through fiscal 2016, apart from separate appropriations for defined purposes, and then fell slightly through a legislated deduction. Congressional Research Service tables updated July 1, 2019 recorded a combined grant of 16.567 billion dollars in fiscal 1997 and 16.512 billion dollars in fiscal 2018, after a reduction of 0.33 percent beginning in fiscal 2017 to finance research and technical assistance. No inflation adjustment, caseload adjustment, or population adjustment was written into the basic grant. Converted with the Consumer Price Index for All Urban Consumers, the fiscal 2018 grant stood at 10.564 billion dollars in constant 1997 dollars, 36.2 percent below fiscal 1997 in real terms.

Q: What is the caseload reduction credit in TANF?

The caseload reduction credit is the reduction in a state work participation target under section 607(b)(3) equal to the percentage decline in the average monthly caseload between the base year and the comparison year, after exclusion of declines caused by federal or state eligibility changes. Subtraction produces the effective standard: a decline of 40 percent lowers the all families standard from 50 percent to 10 percent and the two-parent standard from 90 percent to 50 percent, and a decline of 50 percent or more lowers the all families standard to zero. The Deficit Reduction Act of 2005 recalibrated the base year from fiscal 1995 to fiscal 2005 for standards applicable in fiscal 2007.

Q: What counts as a work activity in TANF?

Section 607(d) lists twelve countable activities, divided in Congressional Research Service analysis from 2019 into nine core activities and three supplemental activities. The core tier covers unsubsidized employment, subsidized private and public sector employment, work experience, on-the-job training, job search and job readiness assistance, community service programs, vocational educational training, and providing child care services to an individual in a community service program. The supplemental tier covers job skills training directly related to employment, education directly related to employment for a recipient without a high school diploma or equivalency certificate, and satisfactory secondary school attendance or study toward a general equivalence certificate for a recipient who had not completed secondary school, countable only after core hours were met.

Q: What is maintenance of effort in TANF?

Maintenance of effort is the minimum state spending required each year under section 609(a)(7) as a condition of a full federal grant, measured against state spending in fiscal 1994 on the predecessor programs, namely Aid to Families with Dependent Children, Emergency Assistance, the Job Opportunities and Basic Skills program, and related child care. The required level stands at 80 percent of that fiscal 1994 amount, reduced to 75 percent for a state that met both work participation rates. Qualified expenditures cover benefits, services, and activities for eligible families that are reasonably calculated to accomplish a TANF purpose. Failure to meet the required level produces a dollar for dollar reduction of the federal grant by the shortfall.

Q: What is the difference between assistance and non-assistance in TANF?

Assistance, under the regulation issued April 12, 1999 at 45 C.F.R. 260.31, means cash, payments, vouchers, and other benefits designed to meet ongoing basic needs such as food, clothing, shelter, utilities, and household goods, plus child care and transportation for families that are not employed. Non-assistance means excluded help, including nonrecurrent short-term benefits for a specific crisis that do not extend beyond four months, work subsidies paid to employers, child care and transportation for employed families, refundable tax credits, Individual Development Account contributions, and services such as counseling and case management that provide no basic income support. The distinction decides whether time limits, work rates, assignment rules, sanctions, and detailed reporting attach.

Q: How is a state grant amount calculated?

A state family assistance grant under section 603(a)(1) was calculated once from predecessor spending and then repeated without recalculation. The amount was the greatest of several measures drawn from fiscal 1992 to fiscal 1995, including a measure based on fiscal 1994 payments with an Emergency Assistance element from fiscal 1995, an average formed from one third of payments for fiscal 1992, 1993, and 1994 combined, and a measure built from the first three quarters of fiscal 1995. The Congressional Research Service, in a primer updated through 2019, identified the predecessor programs as Aid to Families with Dependent Children, Emergency Assistance, and the Job Opportunities and Basic Skills Training program.

Q: What happens when a state misses a participation rate?

Failure to meet an applicable participation rate exposes a state to a reduction of the state family assistance grant under section 609(a)(3), set in Congressional Research Service summaries at a maximum of 5 percent for the first year of failure, rising by 2 percentage points for each subsequent consecutive year of failure to a ceiling of 21 percent of the adjusted grant. Reasonable cause relief and a corrective compliance plan stand between exposure and imposition for the participation penalty, and Congressional Research Service accounts through 2018 documented repeated use of corrective compliance plans and negotiated reductions after recalibration in 2005. Amounts withheld through a penalty also had to be replaced with state funds under the penalty structure.

Q: What does the five-year time limit actually prohibit, and what hardship protection applies?

Section 608(a)(7) prohibits a state from using federal block grant funds to provide assistance to a family that includes an adult who has received assistance for a total of sixty months, whether or not the months were consecutive. The prohibition governs use of federal funds rather than all support to a family, and months of non-assistance add nothing to the count. A hardship exemption in the same provision allows a state to extend federally funded assistance beyond sixty months for up to 20 percent of the average monthly number of families receiving assistance, and the statute also requires disregard of certain months in Indian country or an Alaska Native village where at least half of adults were not employed.

Q: What must a state plan certify?

A state plan under section 602 must contain an outline of the family assistance program and certifications by the chief executive officer of the state. The outline must describe a program serving all political subdivisions, though not necessarily in a uniform manner, providing assistance to needy families and job preparation, work, and support services for parents, and must set forth objective criteria for eligibility and for fair and equitable treatment, including hearing or appeal opportunities. Certifications cover designation of the administering agency, equitable access for members of Indian tribes domiciled in the state who are not served under a tribal plan, and standards against fraud and abuse, including standards on nepotism, conflicts of interest, kickbacks, and political patronage.

Q: What is the contingency fund for?

The contingency fund at section 603(b) supplied extra funding for states that were economically needy, in recognition that a fixed basic grant without adjustment could prove inadequate during a downturn. The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 established the fund at 2 billion dollars. A state had to meet an economic need test expressed through unemployment and food stamp participation and had to spend state funds above the fiscal 1994 predecessor level, after which federal payments matched qualifying expenditure subject to the amount available. The Congressional Research Service reported that the original amount was depleted in early fiscal 2010, after which Congress continued the fund through later appropriations, including 608 million dollars annually beginning in fiscal 2017.

Q: What transfers to child care are allowed?

Section 604(d) allows a state to transfer part of the federal grant to the Child Care and Development Fund and to the Social Services Block Grant, subject to a combined limit of 30 percent of the grant and a separate limit of no more than 10 percent to the Social Services Block Grant, as the Congressional Research Service primer stated the rule. Transferred funds become subject to the rules of the receiving grant, with a condition that funds transferred to the Social Services Block Grant be used for families with incomes below 200 percent of the poverty line. The Administration for Children and Families reported for fiscal 2018 that transfers to the Child Care and Development Fund accounted for 4.8 percent of combined funds.

Q: What counts toward maintenance of effort?

Qualified state expenditures counted toward maintenance of effort include spending for eligible families on benefits, services, and activities reasonably calculated to accomplish a TANF purpose, together with administrative costs subject to separate rules. Counted categories in federal summaries before October 1, 2019 included cash assistance, child care for working families, transportation and other work supports, education and training, family support and preservation services, refundable state tax credits, prekindergarten, child welfare services, and programs addressed to pregnancy prevention and two-parent family formation where statutory conditions were met. Spending through separate state programs financed entirely with state dollars also counted, provided the spending served eligible families and a TANF purpose and was reported as the rules directed.

Q: How do tribal TANF programs differ from state programs?

Tribal programs operate under section 612 through a tribal family assistance plan approved by the Secretary of Health and Human Services, rather than through a state plan under section 602. A tribal grant equals the federal share of fiscal 1994 predecessor expenditures attributable to Indian families in the service area identified in the tribal plan, and the amount is financed by a reduction in the grant of the state or states where that service area lies, rather than by an addition to the national total of 16.489 billion dollars for state grants. The Congressional Research Service primer stated that the Secretary, with participation of the tribes, established work requirements and time limits for each tribe, so tribal programs were not subject to the state participation standards or the state sixty month prohibition in the same form.

Q: What is a nonrecurrent short-term benefit, and why does that benefit escape the time limit?

A nonrecurrent short-term benefit, under 45 C.F.R. 260.31(b)(1) as adopted in 1999, is a benefit designed to deal with a specific crisis situation or episode of need, not intended to meet recurrent or ongoing needs, and not extending beyond four months. Administration for Children and Families examples included arrears and utility help, emergency housing, diversion help, and vehicle repair help where the three conditions were met together. That benefit escapes the sixty month count because the regulation classifies the benefit as non-assistance, and section 608(a)(7) counts only months of assistance paid with federal funds. Receipt alone also places a family outside the work participation calculation and outside the child support assignment condition attached to assistance.

Q: What changed in the work rules under the Deficit Reduction Act?

The Deficit Reduction Act of 2005, Public Law 109-171, signed February 8, 2006, recalibrated the caseload reduction credit by substituting fiscal 2005 for fiscal 1995 as the base year, effective for standards applicable in fiscal 2007, so a credit depended on declines after fiscal 2005. The act also required families receiving assistance in separate state programs financed with maintenance of effort dollars to be counted in participation rates, closing a route that had kept two-parent families and other groups outside the calculation. A further set of changes directed the Department of Health and Human Services to define each countable activity, to establish uniform methods for counting and reporting hours, and to require a state work verification plan, implemented in an interim final rule published June 29, 2006.

Q: Why did unspent balances accumulate under the block grant?

Unspent balances accumulated because a fixed grant continued at the full amount while caseloads fell, and section 404(e) allowed a state to reserve unused dollars without a fiscal year limit for use only within the program. Reserved dollars could not be transferred to the Child Care and Development Fund or to the Social Services Block Grant. The Congressional Research Service reported that on September 30, 2000 states held an unobligated balance of 3.2 billion dollars, apart from amounts obligated but not yet liquidated. An amendment in the American Recovery and Reinvestment Act of 2009 broadened use of carried over funds from assistance alone to any benefit or service that could be provided under the state or tribal program.

Q: What share of combined funds reached basic cash assistance in federal reporting?

Basic assistance accounted for a minority share of combined federal TANF and state maintenance of effort funds in the fiscal 2018 reporting released by the Administration for Children and Families on August 7, 2019. Table A.2 placed basic assistance at 21.4 percent of total funds used in fiscal 2018, at 6.7 billion dollars from a combined pool of 31.3 billion dollars, and at 22.7 percent in fiscal 2017, at 7.0 billion dollars. The Center on Budget and Policy Priorities, in analysis of the same federal spending data, reported that states spent 71 percent of TANF funds on basic assistance in fiscal 1997, the first year of operation, so the fiscal 2018 share represented a fall of approximately fifty percentage points.