Chairman Herger, and Members of the Subcommittee on Human Resources:
Thank you for inviting me to testify on state implementation of work requirements and time limits under the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. My name is Douglas J. Besharov. I am a resident scholar at the American Enterprise Institute for Public Policy Research, where I conduct research on children and families. I am also a professor at the University of Maryland School of Public Affairs, where I teach courses on family policy, welfare reform, and evaluation.
Ask people on the street what "welfare reform" means, and most would probably answer "work in return for welfare." According to Kent Weaver, a senior fellow at the Brookings Institution, public opinion polls conducted between 1993 and 1995, on the eve of welfare reform, revealed that "The clear public favorite among welfare reforms is work requirements, which is consistent with the new paternalism approach to reform."[1] Lawrence Mead of New York University explains: "public opinion polls show that while voters want the government to assist needy families, they also want adult welfare recipients to work, like the taxpayers who support them."[2]
When the Temporary Assistance for Needy Families program (TANF) was enacted, most analysts expected states to develop large mandatory work programs in order to meet its mandatory "participation" requirements. TANF requires states to place an increasing percentage of adults on welfare in work activities. It establishes two separate "work participation rates": (1) an "all-family" or overall rate, and (2) a rate for two-parent families (which is higher than the rate for one-parent families because it is considered easier for one parent in a two-parent household to work than it is for a single mother).
The all-family rate (or overall rate) requires that at least 25 percent of TANF families with an adult (or minor child head of household) be involved for 1997, 30 percent for 1998, 35 percent for 1999, 40 percent for 2000, 45 percent for 2001, and culminating at 50 percent for 2002 and thereafter.[3] In order to be counted as participating, adults in one-parent families must have been engaged in work activities for at least twenty hours per week in 1997 and 1998, twenty-five hours in 1999, and thirty thereafter.[4] (Single parents with a child under six need only participate twenty hours per week to be counted.)
The two-parent family rate is higher (presumably because there are two parents available to care for the children): 75 percent for 1997 and 1998, and 90 percent thereafter. The number of required hours is also higher: In two-parent families, the parents must have been engaged in activities for at least thirty-five hours per week. (The parents can share the hours.)[5]
States that do not meet these participation rates are subject to a financial penalty.[6]
TANF also requires states to reduce or end assistance to people who refuse to engage in such work activities without good cause.[7]
These "participation standards," however, have turned out to have little meaning because of the way the participation rates are calculated and because they can be satisfied by recipients combining welfare with work ("combiners").
Caseload reduction credit. What if a state successfully moves a substantial number of recipients from welfare to work? On the theory that it would be unfair to ignore this achievement, the required participation rates are reduced by the "caseload reduction credit." The credit reduces the state’s required participation rate by one percentage point for each percentage point that the state’s welfare caseload falls below the 1995 level. (Caseload reductions due to eligibility changes, such as full family sanctions, cannot be counted in measuring the caseload decline.)[8] Significantly, thus recognizing "entry effects" gives states an incentive to invest resources and time in helping applicants avoid welfare through various diversion activities and keeping leavers from returning by offering child care and post-employment services.
The caseload reduction credit was established in relation to 1995 welfare caseloads and, because of the sharp decline in the rolls since then, it has all but eliminated the need for states to establish mandatory work programs. For the all-families participation rate, in 2000, thirty-one states did not have to place anyone in a work activity because their caseload declines were so large. In other words their "adjusted" participation rate was zero. Eleven states had "adjusted" all-families participation rates of under 10 percent.[9] Moreover, these participation rates are so low that they are easily satisfied because recipients combining work and welfare (pursuant to earnings disregards) count toward the participation rate. As a result, in 2000, all states and the District of Columbia met the all-families participation requirement.
Meeting the two-parent participation requirements has been more difficult for the states--even though the number of such cases has plummeted nationally[10] (from about 363,000 in 1994 to just 56,000 in 2000, an 85 percent drop)[11]--because both the participation rate and minimum hours of participation are higher. Nationally, in 2000, only about 40 or 50 percent of two-parent cases (with enormous variations among the states) were participating for a sufficient number of hours to meet the two-parent work requirement.[12] However, with the help of the caseload reduction credit, twenty-five states and the District of Columbia met or exceeded their adjusted two-parent work participation rates.[13]
Only seven states did not meet their adjusted two-parent participation rate.[14] In earlier years, some of these states entered into corrective compliance plans with the federal government, and a few states have simply paid the penalty for not meeting their two-parent participation rates. (The penalties tend to be small because they are based on the proportion of two-parent cases in the state, which is generally small.)[15]
But the major reason so few states were not out of compliance is that eighteen or more had, in effect, exempted themselves from the requirement by creating a separate state program for all or some of their two-parent families (or not having a program at all), up from fifteen states in 1999.[16] These separate state-funded programs are not subject to the work requirement (or other TANF provisions such as the five-year time limit).
The growing proportion of the caseload composed of "child-only" cases is also watering down participation requirements. For, there is no work requirement imposed on families that do not have an adult parent receiving aid, even if the parent is living in the same household as the child. In 1997, 23 percent of the national TANF caseload was thus exempt from a work requirement for this reason.[17] By 2000, the figure was up to 32 percent.[18] Some of these child-only cases involve children placed with relatives ("kinship care") because their parents cannot care for them.[19] Some involve immigrant families, where the adult immigrant is not eligible for benefits but their native born children are. Some involve situations where the parent is receiving SSI and is not included as part of the TANF grant (while the child is). And some involve families in which the adult has been sanctioned for some reason and is, therefore, off the grant.
Actual Participation. Despite initial expectations, therefore, participation in the activities counted toward the TANF participation requirements ("countable work-related activities") has been quite limited. In an average month in 2000 (the most recent year with data), only 40 percent of adult TANF recipients participated in a countable activity.[20] And, even that is a misleading statistic, because about 61 percent of those participating are simply combining work and welfare (in large part because of the newly generous earnings disregards described above). TANF calls this "unsubsidized employment," but that clearly is a misnomer since the families continue to receive welfare payments, which can be a substantial portion of their original grants. First the Clinton Administration and now the Bush Administration have helped muddy the waters by repeatedly reporting that large percentages of welfare recipients were "working," when, in fact, the vast majority were taking advantage of earnings disregards to combine work and welfare.[21]
In fact, in 2000, only somewhere between 16 to 23 percent of all adult recipients were participating in activities other than "unsubsidized" employment, [22] and only about 4 percent were in "work experience." Most of the rest of those not combining work and welfare were either in job search (5 percent) or vocational education (3 percent).[23] (See Tables 1 and 1A.)
Importantly, four states--New Jersey, New York, Ohio, and Wisconsin--accounted for over 60 percent of the participants in work experience programs (37,971 out of a national total of 61,643).[24] In these states, the percentage of adults in work experience ranged from 6 percent in New York to 57 percent in Wisconsin. (See Tables 2 and 2A.)
Because most states have had no trouble meeting the all-family participation rate--and because they have unspent TANF funds resulting from the decline in their caseloads--many states have also funded activities and services that may not count toward TANF participation requirements. Sometimes they mandate participation in them. These services include substance abuse treatment, skills assessment, mental health services, domestic abuse services, or adult literacy. Sometimes these services are provided in conjunction with some form of work experience or subsidized employment, and sometimes not.
Such activities are presently not countable toward participation requirements. If they were, they would have added appreciably to the number of recipients in countable activities. (In 2000, they accounted for as much as 15 percent of total participation.)[25] In New York City, for example, in November 2001, adding the participants in normally noncountable activities would increase the number participating there by 7,683 (18 percent). The number of participants would rise from 43,669 to 51,352 (with 1,281 in substance abuse treatment, 1,831 in wellness/rehab, and 4,571 who are "needed at home" to care for a dependent).[26]
A word of warning about these statistics: In conversations with state and county officials, it was clear that many had very poor data on the numbers of participants in various activities, and discrepancies in some states’ data suggest significant inaccuracies. Moreover, the data seems to have little meaning or utility to state officials, and many seem to make little use of the data that they have.
Douglas J. Besharov is the Joseph J. and Violet Jacobs Scholar at AEI.
