Editor’s Note: This chapter is shared here thanks to special permission from Oxford University Press. Matt Weidinger, “Less Activation in US Social Assistance Programs?” In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023.

Abstract:

This chapter reviews work activation in the US cash welfare program currently known as Temporary Assistance for Needy Families (TANF). It focuses on developments surrounding major bipartisan reforms enacted in 1996, when the former Aid to Families with Dependent Children program was replaced with the TANF block grant. Those reforms sought to strengthen the degree to which recipients—primarily single mothers—were expected to engage in work, training, or other activities in exchange for monthly benefit checks. While reforms were followed by significant increases in work and earnings and decreases in poverty and dependence, key features—and how states implemented the new program—subsequently weakened the effects of TANF work activation policies. The chapter concludes that other federal benefit programs—including recent monthly Child Tax Credit checks paid to even non-working parents—have largely eclipsed the TANF program and its work activation policies.

Overview:

Work activation in the US welfare program called Temporary Assistance for Needy Families (TANF) remains a political football over which both US political parties continue to argue. Recent developments in many ways resemble the debate during the now twenty-five years since enactment of the 1996 welfare reform law that created TANF, with most Republicans seeking to strengthen the degree to which the TANF programexpects recipients to work, while most Democrats seek to weaken work requirements, time limits, and funding restraints that were hallmarks of the 1996 law. But those ongoing disputes have proceeded seemingly despite the actual effects of TANF work activation policies, which have continued to weaken with time. That is in significant part due to one of the earliest and most enduring effects of the 1996 reforms: record reductions in welfare caseloads, which remained near historic lows during both the Great Recession and the economic disruptions resulting from the COVID-19 pandemic. As a result of those large and ongoing caseload reductions, states throughout the TANF era have been able to claim “credits” that reduce—often to zero—the share of adults on TANF required to work or train as a condition of collecting benefits.

Since the 1996 law, the partisan contests to, on the right, reinvigorate a significant work activation requirement for the current TANF caseload and, on the left, increase welfare caseloads and funding, have been largely fought to a draw. Neither party, even during periods when theycontrolled both the White House and Congress, has been able to significantly alter the program to match their policy objectives. And now, for over a decade, the TANF program has operated largely without change, based only on short-term extensions of current law that have continued to result in real annual cuts in the value of the TANF block grant, which is not adjusted for inflation.[1] Meanwhile, recent years have seen policymakers on the left as well as the right turn their attention toward expanding other benefit programs. Today, Supplemental Nutrition Assistance Program (SNAP, formerly food stamps), unemployment, and especially Child Tax Credit benefits flow to far more recipient families with children and in far greater amounts than TANF ever provided. Those expanded programs have largely eclipsed the importance of the TANF program and its work activation requirements.

In particular, if revived, the expanded and monthly Child Tax Credit payments made in 2021 will supplant the work activity–backed TANF program with a benefit that, in many states, is more generous than the pre-TANF Aid to Families with Dependent Children (AFDC) program and is also payable without TANF’s work requirements, time limits, or other restrictions. The net effect will be tofurther diminishthe importance of both the TANF program and its work activation policies among the array of US benefit programs. And by once again paying regular government checks to parents regardless of their engaging in work or other activity, this policy also threatens to extinguish any remaining pretense that work activation—a key theme of the major welfare reforms enacted in 1996—remains a significant feature of US welfare policy.

WORK ACTIVATION POLICY UNDER ADC AND AFDC

While the history of work activation in the United States extends to before the country’s founding, the lineage of the current TANF program and its posture toward work activation stems directly from the Aid to Dependent Children (ADC) program, created in 1935, as part of President Franklin Roosevelt’s New Deal.[2] As detailed in a 1934 report of the Roosevelt administration’s Committee on Economic Security, the authors of the ADC program described its proposed benefits as designed specifically to forestall work by recipient mothers: “The theory of the system of aid to dependent children is that the families will be given enough assistance to meet their minimum budgetary needs, without necessitating gainful employment for mothers of young children, which would be detrimental to home life.” (Committee on Economic Security 1934). Especially compared with the far broader eventual scope of the New Deal’s payroll tax–funded social insurance programs like unemployment and social security, ADC benefits were paid to a relative handful of mothers—160,171 in 1936 (Bucklin 1939). The program’s original provision of benefits primarily to widows, along with the attempts by New Deal programs to reserve benefits for prime-age and primarily male breadwinners, explains ADC’s aversion to gainful employment for its original recipients. Other design features, along with the systematic exclusion of minority families, further served to keep recipiency low (Gordon and Batlan 2011).

In the ensuing decades, the ADC program expanded to include some two- parent households and also came to include incentives (andeventually even some limited requirements) for recipient parents to work (Falk 2021). Nonetheless, driven by expanding eligibility—including the ability of black families to enroll—and rising family breakdown, caseloads more than tripled between 1964 and 1973, reaching 3.1 million families (Falk 2021). And as rising numbers of women entered the workforce in the 1970s and beyond, political pressure grew for the singleand often never-married mothers who increasingly dominated the caseload of the now AFDC program to engage in—or at least prepare for—work as a condition of collecting benefits.

By the 1980s, the Reagan administration sought to encourage more state control over welfare programs, for example, by allowing states to require recipients to engage in work experience or community service in exchange for benefits as well as by authorizing demonstrations permitting local testing of welfare- to-work programs. A 1982 proposal to completely devolve cash assistance to states was not enacted (Falk 2021). But the White House was also interested in federal reforms, spurred in part by an influential 1983 study by Mary Jo Bane and David Ellwood (both later welfare officials in the Clinton administration) that noted that while most spells on AFDC were relatively short, 65% of current recipients would eventually spend eight or more years on the rolls (Bane and Ellwood 1985). When the Senate changed hands following the 1986 election, Democrats, now in control of both branches of Congress, took a more active role in reform efforts, crafting what became the Family Support Act (P.L. 100-485), which President Reagan signed into law in October 1988. Among its many provisions, that legislation created the Job Opportunities and Basic Skills (JOBS) program (which provided employment services, education, and training for cash assistance recipients), removed a putative barrier to leaving welfare by establishing the Transitional Medical Assistance program that continued temporary Medicaid coverage for recipients moving from welfare to work, and guaranteed childcare for AFDC recipients “engaged in work activities” along with transitional childcare for those who left AFDC for work (Peskin, Isaacs, and Fairbanks 1989).

That legislation included optional activity requirements, which provided a rhetorical marker on the path to more concrete requirements to come. For example, the JOBS program included state “participation rate” targets that rose from 7% of nonexempt adults in 1990 to 20% in 1995. But, in practice, the law included broad exceptions to those nominal targets. For example, 38% of recipients were expected to be exempt from participation because they were caring for a child under age three. Other exceptions were made for those who were ill, among other categories(Peskin, Isaacs, and Fairbanks 1989). Data from the Department of Health and Human Services (HHS) suggest the years between 1988 and 1994 saw modest increases in full-time work (which rose from 2.1% to 3.2% of recipient mothers) and part-time work (which rose from 4.2% to 4.5%). In keeping with the focus of the program, participation in education and training rose from 2.2% to 13.8%, but that was more than offset by significant declines in the share of mothers seeking work (which fell from 27.6% to 1.7%) (US House of Representatives 1996).

As a result, in the years immediately following the enactment of the Family Support Act, the share of AFDC mothers labeled as “needed at home or not actively seeking work” swelled from 55.5% to 71.7%. And instead of helping to drive caseloads down as the Congressional Budget Office’s assessment of the Family Support Act anticipated, the AFDC caseload increased to record levels.[3] The nonpartisan Congressional Research Service (CRS) notes that “Between July 1989 and August 1992, almost 1 million families were added to the AFDC basic (primarily one-parent) caseload, a 27-percent increase; and by 1992 about one in seven US families with children was enrolled in the program.” (US House of Representatives 1996). A modest intervening recession and the effects of the crack cocaine epidemic are sometimes given as causes for this sharp caseload rise, while studies also point to program and demographic shifts.[4]

Whatever its cause, the rapid caseload growth following enactment of the Family Support Act contributed to calls for more aggressivereforms, such as making eligibility for AFDC more widely conditioned on work or job preparation by able-bodied adults. Building off Bane and Ellwood’s earlier work, other studies that noted lengthy stays on benefits contributed to higher caseloads amplified such calls for reform. For example, a 1995 study by LaDonna Pavetti of the Urban Institute found that the average lifetime stay on welfare, counting repeat spells, for families currently enrolled was thirteen years (Pavetti 1995).

Foundational to those post-FSA calls for change were governors from both parties, including Governor Bill Clinton of Arkansas, who in 1992 made his pledge to “end welfare as we know it” a central element of his successful campaign for President (Washington Post 1992). Clinton argued that the AFDC program and its system of monthly checks for millions of low-incomeadults with children had become “a way of life” for too many.

We will say to those on welfare: “You will have, and you deserve, the opportunity, through training and education, through child care and medical coverage, to liberate yourself. But then, when you can, you must work, because welfare should be a second chance, not a way of life.” (New York Times 1992).

Thus, sixty years after the ADC program was originally designed to replace the need for earnings for a small number of low-income mothers, the US welfare program was now poised to attempt the opposite: to condition receipt of benefits by millions of low-income mothers on work or related activities (Box 6.1).

CREATION OF TANF “WORK PARTICIPATION REQUIREMENTS”

Despite his campaign pledge to reform welfare, once in the White House, President Clinton instead initially focused on controversial budget and healthcare reform legislation. His proposal to reform welfare was unveiled only in June 1994. The Clinton administration proposal “would have phased in a two-year limit on AFDC receipt without work, followed by required participation in a wage paying work program after two years.” (Falk 2021). This proposal immediately drew the ire of liberal Democrats in Congress, who attacked it in committee and blocked its consideration. (Pianin 1994). One key Democratic policymaker, Rep. Harold Ford, Sr. of Tennessee, who chaired the House Ways and Means welfare subcommittee, vowed to oppose any reform legislation that restricted eligibility for AFDC unless it also guaranteed recipients employment paying at least $9 per hour (Haskins 2007). That impasse left the AFDC system intact, despite Clinton’s promises of reform, ahead of the pivotal 1994 congressional elections.

In their campaigns in 1994, House Republican candidates adopted Clinton’s “end welfare as we know it” pledge and made work-oriented welfare reforms a central plank of their ten-point “Contract with America.” Like the Clinton proposal, it would have required work by adult recipients within two years of enrolling in AFDC. But, unlike the Clinton plan, and in a nod to the Bane and Ellwood studies mentioned aboveshowing long-term dependence was a reality for many families, it also would have imposed a five-year lifetime limit on benefits per adult. The legislation also proposed putting limits on the eligibility of unwed teens to welfare checks and a “family cap” that prevented benefits from rising when recipients had more children while on the rolls. In yet another contrast with the Clinton approach, funding for AFDC and childcare would be capped, and states would have the option to receive AFDC funds in the form of a block grant (Falk 2021).

After Republicans claimed the House majority in the November 1994 elections for the first time in forty years (along with control of the Senatefor the first time since the mid-1980s), they worked closely with Republican governors, including those who had previously embraced welfare reforms in their states, to advance federal reform legislation. Consideration of the resulting House bill included dozens of contentiouscommittee hearings and markup sessions, which featured public demonstrations and opponents’ suggesting the legislation’s supporters werecruelly using the nation’s children as “guinea pigs” (Haskins 2007, 150). The legislation ended the entitlement to welfare checks, replaced AFDCwith a block grant, ended the JOBS program, required states to ensure that at least 50% of adult recipients participated in work activities by2003, and included provisions ending benefits for unwed teen parents and requiring states to adopt the “family cap.” It also included new child protection, nutrition, and supplemental security income block grants; significant work requirements for food stamp benefits; and broad restrictions on benefits for noncitizens (Work Opportunity Act of 1995).

Floor debate matched the intensity of committee consideration (Haskins 2007). And, demonstrating the bipartisan appeal of work requirements,Democrats offered a substitute bill, sponsored by Nathan Deal of Georgia (who within a month would switch parties), which supporters argued was“real” welfare reform because it raised the nominal target work participation rate to eventually 52% while increasing funding for education, training, and childcare (Individual Responsibility Act of 1995). All House Democrats who voted supported the Deal substitute, which was defeated before the majority Republican plan was approved on a 234–199 vote on March 24, 1995. In the end, all House Members save one(Democrat Walter Tucker of California, absent during the vote on the Deal substitute) supported overturning the welfare status quo and creating a newsystem of work requirements under either the Deal substitute or the Republican-sponsored bill.

The Senate, led by Republican Majority Leader Robert Dole of Kansas, spent the next six months crafting its alterative. Key issues includedfederal funding for states, the required level of state funding (called “maintenance of effort” or MOE) for the new TANF program, additionalfunding for childcare, and the nature of restrictions on benefits for unwed teens and the family cap. The Senate ultimately rejected what some (including some governors) derided as “conservative strings” on the block grant (such as ending benefits for teen parents and requiring the family cap) and included $3 billion in additional funding for childcare before approving its legislation in a 87 to 12 vote on September 19, 1995.

The subsequent conference agreement was presented to President Clinton twice: once as part of broad budget savings legislation and a second time as a freestanding welfare reform bill. Clinton vetoed both versions in December 1995 and January 1996, respectively, arguing the legislation did “too little to move people from welfare to work” and made excessive cuts in spending on benefits (New York Times 1996b).

Following those vetoes, Congressional Republicans, again working with the nation’s governors, assembled a third version of the legislation. Along the way, presidential politics became more closely enmeshed in the welfare reform debate as Dole emerged as the frontrunner and eventual presidential nominee of the Republican party. Eager to run against a Clinton record of twice vetoing welfare reform, Dole and his advisers were reluctant to offer Clinton a third chance to sign a bill.

As negotiations on a third bill unfolded, a key flashpoint became whether to sever the main welfare reform legislation, including its popular work requirements, from policies such as proposed Medicaid and food stamp block grants, which were strongly opposed by the administration. On the other side of the debate were many of the nation’s governors, including Republicans who supported those block grants. But HouseSpeaker Newt Gingrich jettisoned the idea of a food stamp block grant, insisting that benefit would remain an open-ended federal entitlement. And a revolt led by the large House Republican freshman class—who feared that without a welfare reform law they would have little record to run on that fall—forced the leadership to drop the proposed Medicaid block grant as well. In the end, the House rank and file chose their own political fortunes over those of the Dole campaign, and leadership was forced to go along.

With those changes, and further increases in funding for childcare, the path was cleared for final legislative action. The legislation took the form of a reconciliation bill, which passed the House and Senate in late July prior to an accelerated conference that negotiated the final agreement. On July 31, Clinton, advised by his pollster that a third veto of welfare reform could be “politically catastrophic” to his re-election prospects, issued a public statement that he would sign the legislation (Harris and Yang 1996). The conference report was then quickly approved by large bipartisan majorities in the House (voting 328–101 later on July 31) and Senate (voting 78–21 on August 1). In addition tonearly unanimous support among Republicans, a majority of Democrats in the Senate and half of Democrats in the House supported the conference agreement. Clinton signed the “Personal Responsibility and Work Opportunity Reconciliation Act” into law (P.L. 104-193) on August 22, 1996.

The legislation ended AFDC, marking the first time a major New Deal program had been terminated. It replaced the individual entitlementto AFDC with the TANF program, a broad new block grant to states funding welfare checks but also a wide range of benefits, services, andactivities that states could deploy to assist families with leaving welfare or avoiding it in the first place in favor of work. Compared with the vetoed bills, the legislation included $4 billion in additional funding for childcare, a key administration aim (Government Publishing Office 1996). The legislation included far-reaching changes in child support, children’s disability benefits, eligibility of noncitizens for major benefits, and more. 

In terms of work activation, or what policymakers crafting the new law termed “work participation requirements,” on its face, the new law created the strongest work requirements ever included in federal law. Those included work participation requirements on states, under which states were expected to engage a steadily rising share of all families on welfare with an adult recipient (and, after 2006, what was dubbed a “work-eligible individual”) in one or more of twelve specific “work activities” including unsubsidized employment, job search, and participation in specified education and training activities. By 2002, the share would reach 50% (minus a “caseload reduction” credit). States thatfailed to meet that target would lose federal funds (Haskins and Weidinger 2019).

The loss of funds for failing to satisfy state work participation requirements was designed to be meaningful and increasingly painful for states—and thus encourage them to take effective measures to satisfy the requirements. For the first year of failure, a state could lose up to 5% of its TANF block grant (minus any transfers to other programs), rising in 2 percentage point increments for each subsequent year of failure. HHS could reducethe penalty based on the degree of noncompliance or if noncompliance was due to “extraordinary circumstances.” That financial penalty for failure to satisfy the work requirements reflected the flip side of the financial advantages that states realized from the new block grant itself. Not only did they have new-found freedom to allocate federal and state resources as they saw fit to assist families and otherwise satisfy the federal law’s terms: if they were successful in meeting those objectives without spending the full federal block grant, they also could save—or, as displayed later, divert— federal and some state funds to other purposes.

Alongside the apparent 50% work participation requirement on states, the expressed requirement that any individual adult on TANF must work in exchange for benefits applied only after the adult was on the caseload for two years (or less at state option), and individual states could define “work” for this purpose. The legislation also included “pro rata” penalties against individuals who failed to work as required by states. Subsequent research by the Government Accountability Office revealed that, in 1998, about 5% of TANF families were typically subject to sanction for failing to satisfy work and other program requirements, in most cases experiencing partial sanctions instead of the loss of all TANF benefits (General Accounting Office 2000).

EFFECTS OF THE 1996 REFORMS

Not surprisingly, supporters of the new law predicted a coming era of rising work and personal responsibility and declining poverty and dependence. Reform opponents, on the other hand, predicted that changes in the law would result in disaster. For example, the New York Times opined, “This is not reform, it is punishment.  The effect on cities will be devastating” (New York Times 1996a). The late Sen. Daniel Moynihan (D-NY) predicted children “sleeping on grates” in a dystopian “Grate Society” (US Congress 1996). Future SpeakerNancy Pelosi (D- CA) ventured, “The Republican welfare reform proposal will make the problems of poverty and dependence much worse because it refuses to make work the cornerstone of welfare reform” (US House of Representatives 1996b). Several senior Clinton administration officials resigned in protest of the president’s signing the new law.

The years after 1996 suggest that key effects of the new law were closer to those anticipated by the law’s supporters than its detractors. Contrary to Rep. Pelosi’s forecast of “much worse” dependence, welfare caseloads dropped sharply following enactment of the new law.

As Figure 6.1 displays, the AFDC/TANF caseload rose sharply following the enactment of the 1988 Family Support Act, peaked in 1994, began to decline as federal welfare reforms were being debated in 1995 and 1996, and then plummeted at unprecedented rates in the years following enactment of those reforms.

Only once before had welfare caseloads declined even two years in a row (US Congress 2006). But after reaching its peak, the welfare caseload proceeded to decline for fourteen consecutive years, from just over 5 million families in 1994 to 1.6 million families in 2008, an unprecedented drop of nearly 70% (Haskins and Weidinger 2019). Between August 1996 when the new law was enacted and just five years later in August 2001, the number of families on welfare fell from 4.4 million to under 2.2 million, a decline of more than 50%.[5] The declines were ultimately interrupted by the Great Recession but have continued in the years since, with the TANF caseload falling to 1.1 million families in 2019, immediately prior to the pandemic. As Figure 6.2displays, caseloads remained at that 1.1 million level in June 2020, before subsequently falling to 984,000 in December 2020 (Falk and Landers 2021).

As University of Maryland scholar Doug Besharov, writing in 2008 for the American Enterprise Institute noted, caseload declines were generally unrelated to the nature of a state’s work participation requirements:

Caseloads fell sharply in all states, yet they did so seemingly without regard to whether states developed ambitious programmes or not.They fell in states with strong work-first requirements and those without them; in states with mandatory work experience (workfare) programmes and those without them; in states with job training programmes and those without them; and in states with generous child care subsidies and those without them. They just fell (Besharov 2008)

The widespread caseload declines suggest that federal policies promoting personal responsibility and work over dependence on benefits, while leaving implementation and key policy details to states, had their intended effects. The declines were consistent with fundamental incentives included in the new law, which rewarded states for having smaller welfare caseloads in several ways.

First, the provision of fixed federal block grants reversed a prior financial incentive for larger caseloads embedded in the former open-ended entitlement system—under which states received greater federal funds when caseloads rose. Instead, the new block grant maintained federal funds at roughly 1995 levels, removing the financial incentive for greater caseloads. In its place, and as caseloadsfell, states experienced rapidly rising federal funds per family remaining on welfare. Such funds allowed states to expand spending on work supports such as childcare or nontraditional uses such as child welfare, as the payment of welfare checks fell sharply (Congressional Budget Office 2015).

Second, rapid caseload declines led to many states’ receiving large “caseload reduction credits” toward the new and gradually rising work participation rate requirements (which, for single-parent households that formed the bulk of the caseload, rose in 5 percentage point increments from 25% in 1997 to 50% in 2002). The result was that most states had little trouble satisfying the new law’s work participation requirements in the years following enactment. State policy choices encouraged or required by the new law often pulled in the direction of smaller caseloads aswell. As the Urban Institute noted in 2006, following enactment, “States moved quickly to design welfare-to-work policies that emphasized getting recipients into jobs by shifting to ‘work-first’ welfare systems, modifying program rules to allow more earned income, imposing shorter lifetime limits, and adopting tougher sanctions for families that do not comply with work requirements” (Urban Institute 2006).

The years following enactment of the 1996 reforms also saw a significant acceleration in the already growing share of never-married single mothers (most of whom were not on welfare but comprised the group most likely to enter and stay on the rolls) who participated in the labor force. Figure 6.2 displays how this trend began in 1992 as the economy emerged from the shallow 1990–1991 recession, continued after Earned Income Tax Credit (EITC) expansions enacted in 1993, and then accelerated sharply in the years following enactment of the 1996 reforms.

The figure suggests that mothers who left welfare and similar low-income mothers with backgrounds that made them most likely to experience welfare spells increased their participation in the labor force and then maintained a historically high level of work in the years following enactment of the 1996 reforms (Haskins and Weidinger 2019).[6] As Robert Moffitt and Stephanie Garlow noted in a 2018 review, “In the initial years after reform, many more women joined the labor force than even the reform’s most ardent supporters had hoped” (Moffitt and Garlow 2018).

Earnings from work similarly showed significant improvement in the years following the 1996 reforms, especially when coupled with rising work support benefits such as larger EITC payments that more than offset declining welfare payments.[7] As shown in Figure 6.3, the mean wage and salary income of never-married mothers ages 18 to 54 rose sharply after the 1996 reforms and reached nearly double pre-reform levels prior to the pandemic, in real terms (Haskins and Weidinger 2019).

As a number of observers have noted, the strong economy in the late 1990s and policies designed to make work pay, including expansions to the EITC and the introduction of the Child Tax Credit, contributed to this improving work and earnings picture.[8]

As displayed in Figure 6.4, the increases in work and earnings by unmarried mothers depicted above, accompanied by growing work support benefits, contributed to significant declines in the poverty rates of children following the enactment of the 1996 reforms, as measured by the official as well as the supplemental poverty measures.

As Figure 6.4 shows, 20.8% of children were in poverty under the official poverty measure in 1995. But the rate fell over the next five years to around 16.2% in 2000. It then rose following the 2001 recession, leveled off until the Great Recession that began in late 2007, and then rose to 22% in 2010. Child poverty declined after 2012 to about 18% in 2016. In 2019, 14.4% of children were poor according to the official poverty measure, a decline of more than 30% from the 1995 level (US Census Bureau 2020).

The Supplemental Poverty Measure (SPM), which includes the effect of benefits such as the EITC and food stamps, shows an even greater and more consistently positive improvement in child poverty over the period, falling from 28% in 1993 to 15.2% in 2016 (US Census Bureau 2017). Child poverty as measured by the SPM has since declined further, reaching 12.5% in 2019, marking a more than 50% decline from the 1993 level (US Census Bureau 2020).

Despite such significant declines in child poverty and rising government spending on the working poor, critics of TANF express concern about the continued incidence of deep poverty, often pointing to a declining share of poor families receiving TANF benefits as a contributing factor.[9] Such concerns spotlight the effects of the changing nature of TANF benefits from mostly cash assistance to increasingly work supports and other forms of assistance. As the General Accounting Office (GAO) noted in congressional testimony in 2002,

The increased emphasis on work support and other services for recipients of cash assistance and those not receiving cash assistance represents an significant departure from previous welfare policy that focused on providing monthly cash payments. While the goals andtarget populations of welfare spending have changed, the key measure of the number of people served remains focused solely on families receiving monthly cash assistance. Although this measure provides important information for administrators and policymakers, it does not provide a complete picture of the number of people receiving benefits or services funded atleast in part with TANF/ MOE funds (US General Accounting Office 2002).

That changing nature of benefits suggests that the SPM (which includes growing tax benefits such as the EITC and the Child Tax Credit payablebefore 2021 only when parents work, as well as food stamp benefits) may offer a more complete depiction of need in the post-reform era. As shown in Figure 6.4, child poverty as measured under the SPM was declining or stable in nearly all years since 1996, reaching an all-time low in 2019. This indicates that many parents were able to make up for a loss of cash assistance income from TANF by other means, including wages from work, tax benefits related to work, and other benefits, especially food stamps (Tiehan, Jolliffe, and Smeeding 2013). Further, 2019research by Bruce Meyer and his colleagues finds that more than 90% of those believed to be in extreme poverty are not so once the value of in-kind transfers such as significant food stamp transfers is added to income, survey reports of earnings and transfer receipt are replaced with more accurate administrative records, and the ownership of assets is accounted for (Haskins and Weidinger 2019).

CHANGES IN THE DEFICIT REDUCTION ACT

By authorizing the new TANF program and its many policy changes only through fiscal year 2002, the authors of the 1996 welfare reform lawintended for Congress to carefully review its effects before continuing the program. In considering such effects, Republican policy makers focused much of their attention on the steep caseload declines following the 1996 law and their implications for the state work participation rate requirements. For example, as Congress was considering reauthorization legislation in 2002, the GAO testified that thirty-one states had effectively no work participation rate requirement in fiscal year 2000, when the nominal requirement was 40% (Fagnoni 2002). Republican policymakers thus included policies in welfare reauthorization legislation designed to restore the dynamic the authors of the 1996 law intendedstates to face by 2002. That is, that states either needed to engage 50% of current adults on TANF in countable work activities each month,reduce their current caseload by an additional 50%, or a combination of the two factors equaling 50%.

The House in 2002 and 2003 passed Republican-authored TANF reauthorization legislation that included such updates (Falk 2021). But the legislation failed to gain traction in a narrowly divided Senate (where 60 votes are normally required to pass such legislation), including because states resisted efforts to strengthen the work requirements absent additional federal funding.

The House in early 2005 again initiated action on similar TANF reauthorization legislation, which Republican proponents argued was needed because “the share of current welfare recipients engaged in work or training has fallen in three of the last four years” (Committee on Ways and Means 2005). That legislation was included in budget reconciliation legislation called the Deficit Reduction Act (DRA), which could advance with just a simple majority of 50 votes in the Senate. The DRA included an extension of current TANF funding—continuing the federalblock grant without an inflation adjustment—through fiscal year 2010 (Falk 2021). The DRA also left the 50% state work participation requirement unchanged but updated the caseload reduction credit to provide credit for caseload declines since 2005 rather than 1995. It increased mandatory childcare funding by $200 million per year, even as other welfare provisions in the legislation were projected to save $7.2 billion over ten years, driven mainly by reductions in funding for federal child support administration (Congressional Budget Office 2006). But even though the DRA welfare policy tracked the 1996 law in key respects (fixed funding for TANF, strengthened work activity requirements, additional funding for childcare, and other federal savings), this time support for the legislation was highly partisan. The legislation was narrowly approved in December 2005 by the Senate (51–50, with Vice President Cheney joining 50 Republicans in support) and the House(212–206, with all Democrats voting no) before President Bush signed it into law in February 2006 (US Congress 2006). The bipartisan supportthat characterized the welfare reform law in 1996, bolstered by a Democratic president who promised to “end welfare as we know it,” was clearly a thing of the past.

TANF DURING AND AFTER THE GREAT RECESSION

Critics of welfare reform and its generally fixed federal funding (the program also included a modest “contingency fund” for states experiencing increased need) have often predicted trouble in the event of a recession and presumed rising demand for benefits. While caseloadsremained stable during the generally mild 2001 recession, many were concerned that the Great Recession that began in December 2007 would result in significantly increased demand for benefits. Indeed, anticipating such greater demand for TANF assistance, the American Recovery and Reinvestment Act (ARRA, also known as the Obama stimulus law) in February 2009 changed TANF program funding rules as well as work activation policies.

ARRA . . . included $5 billion for a new TANF Emergency Contingency Fund (ECF) available to be spent in FY2009 and FY2010. The ECF supplemented funding for the regular TANF contingency fund, which itself was depleted in early FY2010. The ECF reimbursed states for 80% of the cost of increased expenditures for basic assistance, short-term emergency aid, and subsidized employment. ARRA also temporarily froze the TANF caseload reduction credit at pre recession levels, through its application to the FY2011 work participation standards (Falk 2021).

ARRA allowed states to substitute the lower of their fiscal year 2007 or 2008 caseloads for the actual caseloads used to calculate the caseloadreduction credit in fiscal years 2009, 2010, and 2011 (Office of Family Assistance 2010). In effect, this allowed states to engage fewer current TANF recipients in work activities than they otherwise would have needed to in order to satisfy the work participation requirements and avoid risking federal financial penalties.

If this policy was designed to promote significant increases in TANF cash assistance caseloads during the recession, a review of TANF caseload dynamics during this period suggests it had only modest success. The average number of families on TANF programs nationwide fellfrom 2.1 million in fiscal year 2005 to 1.7 million in fiscal year 2008, a 19% drop. The national caseload then rose only modestly to 1.8 million in fiscal year 2009 and 1.9 million in fiscal years 2010 and 2011. The GAO reported in 2015 that, in fiscal year 2011, forty-nine of fifty states received some caseload reduction credit, and twenty-two states saw their work participation rate requirement reduced to zero as a result of that credit.[10] By fiscal year 2012 (i.e., once the temporary policy allowing states to freeze caseload reduction credits at their FY 2007 or 2008 levels had expired), the national caseload was declining once again.[11]

Meanwhile, other programs saw far greater increases in recipients and benefits provided. For example, the New York Times in 2012 noted “soaring” food stamp receipt during and after the Great Recession, even as “the number of Americans receiving cash welfare has fallen sincethe 1990s” (De Parle 2012). Unemployment benefits were similarly enlarged, with significant expansions in eligibility for, the amount of, and the duration of weekly checks. As one review summarized, “During the Great Recession, SNAP and Unemployment Insurance programs acted asa counterforce to the economic shock: greatly increasing government spending and expanding income support to families in a way that was at least equal to the cushioning role that they have played in previous recessions, if not more” (Bitler 2019).

What accounts for the difference in demand for benefits? The TANF program’s financing—including its shared financial responsibility with the states—likely played a significant part. As the New York Times noted in 2012, “Since the states get fixed federal grants, any caseload growthcomes at their own expense. By contrast, the federal government pays the entire food stamp bill no matter how many people enroll; states encourage applications, and the rolls have reached record highs” (De Parle 2012). For example, officials in Westchester County, New York, described being rewarded for placing families on the fully federally funded food stamp program rather than the partially state-funded TANF program.[12]

As noted above, many states were able to satisfy the 50% work participation requirement in the early years of the TANF program through theuse of caseload reduction credits. As depicted in Figure 6.5, in each year prior to FY 2016, the share of the national TANF caseload engaged inwork activities for enough hours to count consistently fell short of the nominal 50% goal set by the authors of the 1996 law. During the decade from FY 2002 through 2011, the national average work participation rate fell within a narrow band of between 27.5 and 30.6 each year, before rising considerably in the years following the Great Recession, according to the CRS (Falk and Landers 2021).

The apparent rise in work participation rates in recent years reflects changing policy choices by states, which some describe as work-aroundsdesigned to avoid engaging more recipients in work activities. As the CRS notes, “The increase in the work participation rate has not come from an increase in the number of recipients in regular TANF assistance programs who are either working or in job preparation activities. This increase stems mostly from states creating new ‘earnings supplement’ programs that use TANF funds to aid working parents in the Supplemental Nutrition Assistance Program (SNAP, formerly food stamps) or who have left the regular TANF assistance programs for work” (Falk and Landers 2021). Figure 6.6 displays how those new “earnings supplement” (sometimes called “token payment”) programs are responsible for “(a)lmost all of the increase” in the work participation rate since FY 2012, as CRS describes (Falk 2018).

Congress tried to fix some work participation loopholes in its 2006 reauthorization legislation, but states have used a combination of caseload reduction and additional loopholes since then to significantly diminish the share of current TANF recipients expected to engage in work activities. Since 2010, TANF has operated under a series of short-term extensions without significant policy changes to the underlying program (Weidinger 2018). That has left the program, and especially its intent of engaging a significant share of adults in work and other activities, largely adrift for now over a decade. States continued to claim caseload reduction credits and when necessary tap other loopholes like “excess maintenance of effort (MOE) credits” and the provision of “token payments” to allow them to satisfy the work participation requirements without engaging a significant share of adults in work activities.[13] With Democrats, who as discussed below are increasingly focused on ending work requirements altogether in other programs, presently in control of both Congress and the White House, that status seems highly unlikely to change in the near future.

About half of all states also currently avoid the separate and higher work requirement for two-parent families by putting these cases in a solely state-funded program in which work participation requirements do not apply. Data based on new reporting requirements imposed by HHSshow that only 22.3% of TANF funds were spent on basic assistance (i.e., cash payments to families) in 2019. Another 11% was spent on workactivities and 3% on work supports. Even if the 16% spent on childcare is included under work supports, only a total of 53% of TANF funds is spent on activities directly related to basic assistance and work (Haskins and Weidinger 2019).

To address such perceived flaws, a number of groups and individuals offered TANF reform proposals in the years before the pandemic. These included a comprehensive five-year reauthorization bill crafted by senior Republicans on the House Committee on Ways and Means, which was approved on a partisan vote in committee in May 2018 but was never considered on the House floor. Senior Republicans in the House and Senate introduced related legislation in early 2019, but it was not acted on in either body; a revised version was reintroduced in both bodies in July2021, but has not advanced. This legislation would transition from the process-oriented work participation requirement system to an outcome-focused accountability system for states, require states to engage all work-eligible individuals in work and activities (as defined by each state), and phase out the ability of states to count nongovernmental third-party spending as state spending, designed to minimize the degree to which “excess MOE credits”could reduce the work participation requirements.

As a result, immediately prior to the pandemic in FY 2019, half of all states had effectively no work participation rate requirement(Administration for Children and Families 2020). This was due to the caseload reduction credit, now based on caseload declines since 2005. Participation by adults on TANF in activities other than unsubsidized employment remained rare (Administration for Children and Families 2020). The TANF caseload in 2019 numbered around 1 million families with 3 million total recipients (Administration for Children and Families 2020) In contrast, the SNAP (often called food stamps) caseload before the pandemic numbered 35 million recipients—more than one in ten Americansand ten times the number collecting TANF assistance (US Department of Agriculture 2021). That is evidence of decades of efforts to expandfederal food stamp receipt while TANF caseloads, funded by both state and federal funds, continued to drift ever lower.

TANF IN THE COVID-19 PANDEMIC

Since the pandemic struck the US economy in March 2020, the number of TANF recipients and families has fallen slightly (through December2020), in sharp contrast with other benefit programs.[14] As with prior declines, this likely owes to state TANF policies as well as the fixed block grant funding of the program—to which states are expected to contribute their own funds. That contrasts with rapidly growing and entirely federally funded programs such as repeated large stimulus checks, growing SNAP benefits, and extended and significantly expanded unemployment checks enacted at the start of the pandemic and extended in December 2020 and March 2021.

Similarly, and in contrast with the Great Recession, federal legislative activity involving TANF since the pandemic struck has been remarkably minimal. For example, the March 2020 Coronavirus Aid, Relief, and Economic Security (CARES) Act included only a straight reauthorization of the TANF program through November 2020. Meanwhile, that legislation included historic expansions in eligibility for and the amount of benefits provided to tens of millions of families under unemployment, food, housing, health, and other programs, at a total cost of $1.7 trillion (Congressional Budget Office 2020). Later, the March 2021 American Rescue Plan created the Pandemic Emergency AssistanceFund providing states a share of $1 billion to be used through the end of fiscal year 2022, generally for non-recurrent short-term TANF benefits (Office of Family Assistance 2021). Again, this level of additional assistance paled in comparison with other benefits provided to those in needunder that and related legislation. For example, between March 2020 and September 2021, an estimated $700 billion in temporary federal unemployment benefits were provided through several major bills (Weidinger 2021b).

In terms of work participation requirements, the absence of legislative activity was similarly telling. In contrast with the Great Recession, since the pandemic struck, Congress has not even bothered to act on legislation temporarily minimizing the effect of TANF’s work participation requirements. However, HHS can waive financial penalties for any states that do not satisfy the requirements due to “reasonable cause,” and it has said it would use this authority “to the maximum extent possible” (Falk and Landers 2021). Recent experience suggests that will not be necessary, including because the TANF caseload remains remarkably low. In FY 2020, all states satisfied the main work participationrate requirement, including twenty-eight states whose 50% target rate was reduced to zero due to the caseload reduction credit (Administrationfor Children and Families 2021). Few adults participate in activities other than unsubsidized employment, which includes those collecting token checks; nonparticipating adults rarely complete any hours of activity at all. Absent legislative changes, states will continue to claim such credits allowing them to satisfy the work participation requirements without engaging a significant share of adults now on the rolls in productive activities.

CONCLUSION

Work activation in the TANF program remains a policy football over which the political parties have continued to struggle in recent years, ultimately resulting in few changes despite repeated opportunities to improve the program. Under current program rules, most states have found it easy to satisfy work participation requirements without placing a significant share of current adult recipients in work, education, training, or other activities. Indeed, some states have recently taken to offering token TANF payments to working adults not on the program in an apparent effort to avoid engaging adults on the program in required activities. That development, alongside the greater effects of the decline in the TANF caseload leading to a large caseload reduction credit, has continued to neuter the practical effect of TANF work participation requirements.

It is telling that the TANF caseload has remained near historic lows—with fewer than 1 million families on the rolls, only about half of whom include an adult subject to work requirements—despite significantly increased needs stemming from the pandemic and the resulting economic disruptions since March 2020. That is likely a result of the shared responsibility of paying for TANF benefits between state and federal governments, in contrast with new and expanded federal benefits crafted in response to the pandemic that generally relieved states of significant benefit costs. Those federal programs ramped up quickly and remained significantly larger than pre-pandemic levels in mid-2021 in terms of benefit receipt and spending even after the worst of the pandemic appeared to have passed.

For example, in May 2021, federal SNAP benefits were payable to more than 42 million persons in some 22 million households. In August2021, the Biden administration announced the largest increase in food stamp benefits ever (USDA 2021; Nova 2021). State and federal unemployment benefits, including unprecedented expansions in eligibility for and the amount of federal benefit checks, are likely to total $900billion since the start of the pandemic and, at their peak, were paid to more than 30 million recipients.[15] And fully refundable monthly federal Child Tax Credit payments were flowing to 39 million households with children in mid-2021—nine times the 4.4 million families on the TANF program when it was created in 1996, much less the far smaller caseloads of today.

The recent changes to the Child Tax Credit program reflect the sharpest contrast with the TANF program. The Child Tax Credit program’sbenefits and number of recipients have significantly increased, including through the elimination of the former requirement that parents work to collect benefits. That policy, enacted for 2021, and which President Biden and other senior Democrats would extend permanently, effectively recreates limitless monthly federal benefit checks for parents without any expectation of work—in effect, reviving the defining trait of the pre-reform AFDC program.

These developments reflect the growing ascendence of progressives—who generally opposed the work-based 1996 welfare reforms and have resisted efforts to strengthen the TANF work requirements ever since—among the current governing Democratic coalition. That includesPresident Biden, who was a supporter of the 1996 reforms but now supports eliminating the prior work requirement for adults claiming Child TaxCredits. But instead of directly attacking the TANF program and its work requirements, they have instead converted the Child Tax Credit—formerly a benefit reserved for workers—into the equivalent of welfare checks for nearly all parents, regardless of whether they work or not. Ifextended for additional years beyond 2021, such payments would effectively replace TANF and its expectation of work activity with a newfederal benefit paid to almost all parents regardless of their participation in work, training, or other constructive activity. In many states, the Child Tax Credit benefit for a parent with two young children already exceeds the real value of welfare checks paid under AFDC for similar families (Weidinger 2021b).

It remains to be seen how state TANF programs react to these federal benefit checks, for which all current TANF recipients and millions ofadditional low-income parents were eligible. Especially if this policy is made permanent, some states may further restrict the payment of welfare checks using TANF funds, which could reduce the apparent TANF caseload even more. But regardless of how state TANF programs react, theserecent developments only continue the now decades-long trend of the TANF program becoming a diminishing component of the American social welfare system as a shrinking share of families collect its cash assistance benefits and few of those who do are expected to engage in activities inexchange for that assistance. The absence of significant TANF-related legislation in the pandemic, a period when other programs were significantly expanded to benefit tens of millions of families with children, has already underscored the TANF program’s growing irrelevance within that broader policy landscape.

NOTES

1. For a discussion of TANF extensions and their effect in terms of real TANF funding, see Weidinger (2018).
2. For a useful history of American welfare policy, including on work activation themes, see Olasky (1995).
3. Peskin, Isaacs, and Fairbanks (1989) noted: “By CBO’s estimates, 50,000 families will be off AFDC by the end of the five-year period, a reduction of about 1 percent in caseloads.”
4. See, for example, Blank (2001) and Congressional Budget Office (1991).
5. See https://www.acf.hhs.gov/sites/default/files/documents/ofa/1996_15months.pdf and https://www.acf.hhs.gov/sites/default/files/documents/ofa/2001_15months_tanssp.pdf. The 1996 figure is AFDC cases, and 2000 is TANF and Separate State Program/MOE families.
6. The data in Figure 6.1 are labor force participation (LFP) rates, defined by the Department of Labor as the share of mothers either employed or actively looking for work. The seven-year pe- riod leading up to 2000 saw strong increases in the LFP rates of all single mothers, never-married single mothers (the most disadvantaged group and the most likely to be on welfare), and mar- ried mothers. Despite far smaller caseloads and zero effective work participation requirements in many states, LFP ratesremained well above their averages prior to welfare reform in the years before the pandemic.
7. As Ron Haskins of the Brookings Institution noted in congressional testimony in 2006, “Census Bureau data for female-headed families in the bottom 40 percent of the income distribution for female-headed families (those below about $21,000 in 2000) show that their pattern of income shifted dramatically between 1993 and 2000. In1993 earnings accounted for about 30 percent of the income of low-income, female-headed families, while welfare payments, including cash, food stamps, housing, and school lunch, accounted for nearly 55 percent. By 2000 this pattern had reversed: earnings had leaped by an astounding 136 percent, to constitute nearly 60 percent of income, while welfare income had plummeted byover half, to constitute only about 23 percent of income [Figure 6.2]. As a result of the growth in earnings and legislated expansions of the EITC, income from the EITC more than tripled. Thus with earnings and EITC payments leading the way, the total income of these low-income families increased by about 25 percent over the period(in dollars adjusted for inflation)” (US Congress 2006).”
8. As noted in Haskins and Weidinger (2019), “reviewers across the political spectrum (Blank 2009; Haskins 2017; Grogger and Karoly 2005) have attributed improvements in the rates of work, income, and poverty following the 1996 reforms to a combination of a strong economy, other reforms like the increases in the EITC, and welfare reform policies themselves.”
9. For a discussion of rising spending on the working poor, see Hoynes and Schanzenbach (2018).
10. The GAO also reported that other states made increasing use of “excess MOE credits” as well as “token checks” to satisfy work participation requirements (US Government Accountability Office 2015).
11. See Office of Family Assistance (2016).
12. Personal communication with Douglas Besharov.
13. For example, the GAO in 2012 found that in 2009 “16 of 45 states that met the TANF work participation rate would not have done so without the credit they received for excess state MOE spending.” See US Government Accountability Office (2012).
14. Unemployment insurance, for example, grew from under 2 million recipients prior to the pan- demic to over 30 million by mid-2020. See Administration for Children and Families (2020).
15. See Weidinger (2021b) for a discussion of benefit spending, and US Department of Labor (2021), for peak claims.

REFERENCES

Administration for Children and Families. Temporary Assistance for Needy Families (TANF) and Separate State Programs—Maintenance ofEffort (SSP- MOE) Work Participation Rates and Engagement in Work Activities Fiscal Year 2019. Washington, DC, 2020.

Administration for Children and Families, Office of Family Assistance, “Work Participation Rates: Fiscal Year 2020,” Washington, DC, 2021, https://www. acf.hhs.gov/ofa/data/work-participation-rates-fiscal-year-2020.

Bane, Mary Jo, and David Ellwood, “The Dynamics of Dependence: Routes to Self-Sufficiency,” US Department of Health and Human Services, 1985.

Bitler, Marianne. “Responsiveness of the Safety Net During Downturns: Lessons From the Great Recession,” EconoFact, 2019, https://econofact.org/respons iveness-of-the-safety-net-during-downturns-lessons-from-the-great-re- cession

Blank, Rebecca M. “What Causes Public Assistance Caseloads to Grow?” Journal of Human Resources 36, 1 (2001): 85–118. doi:10.2307/3069671

Besharov, Douglas J. “Two Cheers for American Welfare Reform: Lessons Learned, Questions Raised, Next Steps,” American Enterprise Institute, Washington, DC, 2008, https://www.aei.org/wp-content/uploads/2011/10/ 20081030_BesharovWelfare.pdf?x91208

Bucklin, Dorothy. “Public Aid for the Care of Dependent Children in Their Own Homes, 1932–38,” Social Security, 1939, https://www.ssa.gov/policy/docs/ ssb/v2n4/v2n4p24.pdf.

Committee on Economic Security, “Social Security in America,” Washington, DC, 1934, https://www.ssa.gov/history/reports/ces/cesbookc13.html.

Committee on Ways and Means, Subcommittee on Human Resources, markup documents for HR 240, Washington, DC, 2005.

Congressional Budget Office, “CBO Staff Memorandum,” December 1991, https://www.cbo.gov/sites/default/files/102nd-congress-1991-1992/reports/ 1991_12_aprelimanalysisiofgrowing.pdf.

Congressional Budget Office, “Federal Budgetary Implications of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996,” Washington, DC, 1996, https://www.cbo.gov/sites/default/files/104th-congr ess-1995-1996/reports/1996doc32.pdf.

Congressional Budget Office. Letter from Phillip Swagel, Director of Congressional Budget Office, to Honorable Mike Enzi, Chairman of the Committee on the Budget, Washington, DC, 2020, https://www.cbo.gov/sys tem/files/2020-04/hr748.pdf.

Congressional Budget Office, “S. 1932: Deficit Reduction Act of 2005,” Washington, DC, 2006, https://www.cbo.gov/sites/default/files/109th-congr ess-2005-2006/costestimate/s1932conf0.pdf.

Congressional Budget Office,“Temporary Assistancefor Needy Families: Spending and Policy Options,” Washington, DC, 2015,https://www.cbo.gov/sites/defa ult/files/114th-congress-2015-2016/reports/49887-tanf.pdf.

De Parle, Jason. “Welfare Limits Left Poor Adrift as Recession Hit,” The New York Times, 2012, https://www.nytimes.com/2012/04/08/us/welfare-limits-left- poor-adrift-as-recession-hit.html.

Fagnoni, Cynthia. “States Are Using TANF Flexibility to Adapt Work Requirements and Time Limits to Meet State and Local Needs,” Before the Subcommittee on Human Resources, Committee on Ways and Means, 107th Cong. (2002)

Falk, Gene. “The Temporary Assistance for Needy Families (TANF) Block Grant: A Legislative History,” Congressional Research Service,2021, https:// greenbook-waysandmeans.house.gov/sites/greenbook.waysandmeans.house. gov/files/R44668.

Falk, Gene. “Temporary Assistance for Needy Families: Work Requirements,” Congressional Research Service, Washington, DC, 2018,https://www.everycr sreport.com/files/2018-03-27_IF10856_03e12b7c0bf63e9d8a450990d3efe 9260d437d66.

Falk, Gene, and Patrick Landers, “The Temporary Assistance for Needy Families (TANF) Block Grant: Responses to Frequently Asked Questions,” Congressional Research Service, Washington, DC, 2021, https://fas.org/sgp/ crs/misc/RL32760.pdf.

Linda Gordon and Felice Batlan, “The Legal History of the Aid to Dependent Children Program,” 2011, https://socialwelfare.library.vcu.edu/public-welf are/aid-to-dependent-children-the-legal-history/

Government Publishing Office, “Public Papers of the Presidents of the United States: William J. Clinton,” 1996, https://www.govinfo.gov/content/pkg/PPP- 1996-book2/pdf/PPP-1996-book2-doc-pg1328.pdf.

Harris, John F. and John E. Yang, “Clinton to Sign Bill Overhauling Welfare,” The Washington Post, 1996, https://www.washingtonpost.com/wp-srv/politics/ special/welfare/stories/wf080196.htm

Haskins, Ron. Work Over Welfare: The Inside Story of the 1996 Welfare Reform Law (Washington, DC: Brookings Institution Press, 2007

Haskins, Ron and Matt Weidinger, “The Temporary Assistance for Needy Families Program: Time for Improvements,” Annals of the American Academy of Political and Social Science, 686: 286–309, 2019, doi:10.1177/ 0002716219881628.

Hoynes, Hilary W. and Diane Whitmore Schanzenbach, “Safety Net Investments in Children,” Brookings Papers on Economic Activity, Washington, DC, 2018, https://www.brookings.edu/bpea-articles/safety-net-investments-in- children/.

Moffitt, Robert and Stephanie Garlow, “Did Welfare Reform Increase Employment and Reduce Poverty?” 2018, https://inequality.stanford.edu/sites/default/files/ Pathways_Winter2018_Employment-Poverty.pdf.

The New York Times, “In Their Own Words: Transcript of Speech by Clinton Accepting Democratic Nomination,” July 17, 1992, https://www.nytimes. com/1992/07/17/news/their-own-words-transcript-speech-clinton-accept ing-democratic-nomination.html.

The New York Times, “A Sad Day for Poor Children,” August 1, 1996a, https://www.nytimes.com/1996/08/01/opinion/a-sad-day-for-poor-children.html.

The New York Times, “Welfare Bill Vetoed,” January 10, 1996b, https://times machine.nytimes.com/timesmachine/1996/01/10/058696.html?pageNum ber=1.

Nova, Annie. “Biden Administration Announces the Biggest Increase to Food Stamps Ever,” CNBC, 2021, https://www.cnbc.com/2021/08/16/biden-adm inistration-announces-biggest-increase-to-food-stamps-ever.html.

Olasky, Marvin. The Tragedy of American Compassion (Washington, DC: Regnery Publishing, 1995).

Office of Family Assistance, “Q&A: The American Recovery and Reinvestment Act of 2009 (Recovery Act): Caseload Reduction Credit/TANF Work Participation Rates,” Washington, DC, 2010, https://www.acf.hhs.gov/ofa/ faq/qa-american-recovery-and-reinvestment-act-2009-recovery-act#_Cas eload_Reduction_Credit_Tanf.

Office of Family Assistance, “TANF-ACF-PI-2021-02 (The Pandemic Emergency Assistance Fund),” Washington, DC, 2021, https://www.acf.hhs.gov/ofa/pol icy-guidance/tanf-acf-pi-2021-02.

Office of Family Assistance, “TANF Caseload Data 1996–2015,” 2016, https:// www.acf.hhs.gov/ofa/data/tanf-caseload-data-1996-2015

Pavetti, LaDonna “Who Is Affected by Time Limits?” in Welfare Reform: An Analysis of the Issues, edited by Isabel Sawhill, 1995,http://webarchive.urban. org/publications/306620.html#chap07.

Peskin, Janice, Julia Isaacs, and Alan Fairbank, “Work and Welfare: The Family Support Act of 1988,” Congressional Budget Office,Washington, DC, January 1989, https://www.cbo.gov/sites/default/files/101st-congress-1989-1990/ workingpaper/1989_01_work_0.pdf

Pianin, Eric. “Similarities, Conflicts Arise at Welfare Reform Hearings,” The Washington Post, July 28, 1994.

Tiehen, Laura, Dean Jolliffe, and Timothy Smeeding, The Effect of SNAP on Poverty. University of Kentucky, 2013.

Urban Institute, “A Decade of Welfare Reform: Facts and Figures,” Washington, DC, 2006, https://www.urban.org/sites/default/files/publication/51136/900 980-a-decade-of-welfare-reform-facts-and-figures.pdf.

US Census Bureau, “Income and Poverty in the United States: 2019,” Washington, DC, 2020, https://www.census.gov/library/publications/2020/demo/p60- 270.html.

United States Census Bureau, “The Supplemental Poverty Measure: 2016,” Washington, DC, 2017, https://www.census.gov/content/dam/Census/libr ary/publications/2017/demo/p60-261.pdf.

US Congress, “Proceedings and Debates of the 104th Congress, Second Session, Vol. 142, No. 106,” Washington, DC, July 18, 1996,https://www.congress.gov/ 104/crec/1996/07/18/CREC-1996-07-18-senate.pdf.

US Congress, “S.1932: Deficit Reduction Act of 2005,” Washington, DC, 2006, https://www.congress.gov/109/plaws/publ171/PLAW-109publ171.pdf.

US Congress, “To Review Outcomes of 1996 Welfare Reforms,” Washington, DC, July 19, 2006 (testimony of Ron Haskins),https://www.govinfo.gov/content/ pkg/CHRG-109hhrg30623/pdf/CHRG-109hhrg30623.pdf.

US Department of Agriculture, Food and Nutrition Service, “SNAP Data Tables,” Washington, DC, 2021, https://www.fns.usda.gov/pd/supplemental-nutrit ion-assistance-program-snap

US Department of Agriculture, Food and Nutrition Service, “Supplemental Nutrition Assistance Program Participation and Costs,” Washington, DC, 2021, https://fns-prod.azureedge.net/sites/default/files/resource-files/SNAP summary-8.pdf

US Department of Labor, “COVID-19 Impact,” Department of Labor Blog, Washington, DC, 2021, https://oui.doleta.gov/press/2020/070920.pdf

US General Accounting Office, “States Provide TANF-Funded Work Support Services to Many Low-Income Families Who Do Not Receive Cash Assistance,” Washington, DC, 2002, 13, https://www.gao.gov/assets/gao-02-615t.pdf.

US General Accounting Office, “Welfare Reform: State Sanction Policies and Number of Families Affected,” Washington, DC, 2000,https://www.gao.gov/ assets/hehs-00-44.pdf.

US Government Accountability Office, “Temporary Assistance for Needy Families: An Overview of Spending, Federal Oversight, and Program Incentives,” Washington, DC, 2015, https://www.gao.gov/assets/gao-15- 572t.pdf.

US Government Accountability Office, “Temporary Assistance for Needy Families: State Maintenance of Effort Requirements and Trends,”Washington, DC, 2012, https://www.gao.gov/products/gao-12-713t

United States House of Representatives, Ways and Means Committee, “1996 Green Book,” Washington, DC, 1996a, Section 8, table 8-29,https://aspe.hhs. gov/1996-green-book.

US House of Representatives, “Welfare and Medicaid Reform Act of 1996; Congressional Record Vol. 142, No. 106,” July 18, 1996b,https://www.congr ess.gov/congressional-record/1996/07/18/house-section/article/H7796-1.

The Washington Post, “Bill Clinton in 1992 Ad: ‘A Plan to End Welfare As We Know It,’ ”https://www.washingtonpost.com/video/politics/bill-clinton-in- 1992-ad-a-plan-to-end-welfare-as-we-know-it/2016/08/30/9e6350f8-6ee0- 11e6-993f-73c693a89820_video.html.

Weidinger, Matt. “ ‘Child Allowances’ Revive Welfare as We Knew It,” AEIdeas, February, 2021a, https://www.aei.org/poverty-studies/child-allowances-rev ive-welfare-as-we-knew-it/.

Weidinger, Matt. “Headed for $900 Billion in Unemployment Benefits,” AEIdeas, Washington, DC, 2021b, https://www.aei.org/poverty-studies/headed-for- 900-billion-in-unemployment-benefits/ for a discussion of benefit spending, Weidinger, Matt. “While Congress Differs onBorder Wall, House and Senate Bills Agree on Trimming Welfare Block Grant,” AEIdeas, 2018, https://www. aei.org/economics/while-congress-differs-on-border-wall-house-and-sen

ate-bills-agree-on-trimming-welfare-block-grant/.