SECTION: CAPITOL HILL HEARING TESTIMONY

LENGTH: 6057 words

COMMITTEE: SENATE FINANCE

HEADLINE: WELFARE OVERHAUL

TESTIMONY-BY: STEVE SAVNER, SENIOR STAFF ATTORNEY

AFFILIATION: CENTER FOR LAW AND SOCIAL POLICY

BODY:
Testimony of

Steve Savner Senior Staff Attorney Center for Law and Social Policy

Senate Finance Committee

April 10, 2002

Center for Law and Social Policy

o Create a Career Ladders Fund to enable low-wage workers to upgrade skills and to demonstrate and replicate effective practices for serving them.

o Modify child support assignment and distribution rules so that families receive more child support.

oAllow states to use federal TANF funds to provide ongoing assistance to lowearning working families, without needing to apply a time limit against working families.

oImproving funding for child care and for the basic TANF block grant.

The Administration's proposal would raise TANF participation rates, require 40 hours of participation to fully count toward participation rates, and limit the activities that could count toward the first 24 hours of participation to a set of "direct work" activities. This approach is unwise and unworkable for several reasons:

oIt would significantly restrict state flexibility and compel states to adopt a program model that no state has chosen to implement;

oThe proposal runs counter to what is known about what works;

oWhile the Administration has articulated an overarching goal of improving child well-being, the proposal risks having the opposite effect;

oThe absence of any new funding would force states to cut other programs supporting low-earning working families in order to meet the new costs imposed by the proposal. TANF Block Grant Emphasis on Employment

While much of the current political debate centers around participation rates, the 1996 welfare law sought to promote work by a number of key features:

TANF block grant levels were frozen and states received substantial flexibility in the use of both federal and state maintenance of effort (MOE) funds that became available as cash assistance caseloads fell;

Child care funding was increased, and, in addition, states were allow to transfer up to 30% of TANF funds to the Child Care and Development Block Grant;

Medicaid was "delinked" from TANF ensuring that low-income families could receive Medicaid without participating in welfare;

States received broad discretion to create financial incentives for employment through earned income disregards, and to impose sanctions, including full termination of cash assistance, for noncooperation with work activities requirements;

Time limits were imposed on the use of federal funds for assistance; and

States were required to meet participation rates for families receiving assistance, backed up by the threat of significant financial penalties.

In implementing TANF, states continued a trend begun in the early 1990s of fundamental change in the basic orientation of state systems, as the principal focus shifted from providing income support to an emphasis on requiring and supporting employment.

Employment Outcomes Since 1996

During the 1990s, there was a historically unprecedented increase in employment among lowincome single mothers. Studies of families who have left welfare have consistently found that about 60% of leavers are working. The share of adults working while receiving TANF assistance increased from 8% in 1994 to 28% in 1999." And, ample available evidence points to the dramatic increase in employment among low-income single mothers in recent years. In announcing its welfare reform proposal, the Administration reported that after a decade in which the annual employment rate for single mothers hovered around 58%, the rate had increased every year through 2000, and reached over 73% of mothers heading families in 2000. Moreover, employment rates for never-married mothers increased from under 46% in 1995 to nearly 66% in 2000, an increase of over 40% in just five years. The Administration observed: "These employment increases by single mothers and former welfare mothers are unprecedented. By 2000, the percentage of single mothers with a job reached an all-time high.""' TANF played an important role in the growth of employment among single mothers, but was not the only factor. The "TANF effect" involved both additional requirements and federal block grant funds that became available because of caseload declines, making it possible to increase services.

Beyond this overarching result, however, there are a number of ongoing challenges that need to be addressed:'

Much of the employment has been in low-wage jobs, often without employer-provided benefits.

There is some earnings growth over time, but so far, earnings remain low for most of the affected families.

Many exiting families have low earnings underscores the importance of access to "work supports" - Food Stamps, Medicaid, child care assistance, and child support services - for families leaving assistance. However, participation in Food Stamps and Medicaid sharply declines after families leave assistance, most working leavers do not receive child care assistance, and most leavers do not receive child support.

Families still receiving assistance often have serious and multiple barriers to employment.

It is not yet clear how TANF implementation has affected children, but research on preTANF programs suggests that positive effects may depend on improved family income, and that there may be negative effects on adolescent children that result from increased maternal employment.

Future Directions to Better Meet the Need for Financial Security

These results suggest that, in reauthorization, Congress should continue TANF's strong focus on work but should also communicate to states the importance of.

helping low-income families get better jobs;

helping those with serious barriers to employment prepare for and secure stable employment; and

helping low-income families, including those leaving welfare, gain improved access to work supports, such as child care, child support, Food Stamps, Medicaid.

The following proposals would advance these goals:

States should have the option of using employment outcome measures in lieu of the participation rate process measures in TANF. It is widely agreed that the principal goal of TANF is to increase employment among low-income parents, yet there is no evidence that simply maximizing "participation" is the best way to improve employment outcomes. Participation rates are process measures, and many administrators would greatly prefer to manage programs focusing on desired outcomes. Accordingly, we recommend that states should have the ability to elect to be accountable for meeting a set of employment outcome measures, e.g., employment entries, initial wages, employment retention, earnings gains, rather than participation rates.

Assuming participation rates are maintained in some form, there are four important modifications to the current structure that should be made:

Replace the "caseload reduction credit " with a new employment credit that rewards states when families leave welfare for employment. Give extra credit to states that help families obtain higher paying jobs.

We recommend phasing out the caseload reduction credit and replacing it with a credit based on the extent to which families leaving welfare include a working adult. The participation rate structure should not reward states for caseload reduction whether or not the caseload reduction is attributable to employment. Replacing the caseload reduction credit with an employment credit would reward states for helping families find work, an outcome that is consistent with current TANF goals. Giving states extra credit for placing families in higher-paying jobs would be an important step in reorienting TANF toward the additional goals of poverty reduction and financial security.

Eliminate current limits on vocational educational training.

Since TANF was adopted, new research findings have made it increasingly clear that the most effective welfare-to-work programs have offered a mix of job search, education, job training, and work activities. Some of these mixed-strategy programs have not only increased employment but have also helped welfare recipients find better jobs than they would have on their own. The best example is Portland's Steps to Success program in the National Evaluation of Welfare-toWork Strategies, (NEWWS.)v' Portland provided a mix of services, including job search, life skills, work-focused basic education, and occupational training. Among the eleven NEWWS sites, Portland increased employment and earnings more than the three "work first" programs while also increasing receipt of occupational licenses or certificates and GEDs by as much as the seven education-focused sites." This research suggests that there is ample basis for states so make greater use of education and training, and under the TANF structure states have no incentive to allow such activities unless they are thought to be effective. States should have the flexibility to make these choices without federal limitations.

States should be allowed to have broader discretion to count "barrier removal activities" toward participation rates.

As states have begun working with families with multiple barriers (e.g., health, mental health, disability, substance abuse, domestic violence, lack of English language proficiency), they have typically been unable to count involvement in individualized, barrier removal activities toward the rates. Again, a state has no incentive to allow or pay for such activities unless the state believes it will be an effective means to help a family move toward employment.

The separate two parent participation rate should be eliminated.

The current 90% participation rate for two-parent families has created a strong disincentive against assisting two-parent families in state TANF programs, because a state subjects itself to a greater risk of penalties by assisting such families. There is no need for a separately-calculated rate for two-parent families.

A new Transitional Jobs Block Grant should be created to provide funding to states and localities that want to develop and expand these innovative programs.Since 1997, several states (including Washington, Pennsylvania, Minnesota, and Georgia) and more than 30 cities have implemented transitional jobs programs to help increase employment and earnings of TANF recipients who have been unable to find stable, unsubsidized employment. Such programs generally combine wage-paying jobs with skill development activities and related support services. CLASP has worked extensively with a number of these programs and has provided intensive technical assistance in their development and implementation since 1997. Over 30 programs responding to a CLASP survey reported promising results, two of the largest programs in Washington State and Philadelphia reporting that individuals who complete the program have employment rates in excess of 75%. However, transitional jobs are typically not used in state TANF programs, in part because they are more expensive than other alternatives. While we do not recommend requiring states to adopt such programs, we do recommend providing additional funding to encourage their replication and expansion.

Create a Career Ladders Fund to enable low-wage workers to upgrade skills and to demonstrate and replicate effective practices for serving them. As described earlier, those leaving welfare for work have typically found jobs at below-poverty wages, and the majority of them are not receiving key benefits, such as health care, through their employers. Like other low-wage workers, however, many welfare recipients cannot qualify for higher-quality jobs without intensive services to upgrade their skills and address barriers to employment. Legislation should include additional, dedicated funding directed at research, evaluation, and replication of best practices to improve employment outcomes for families with the most serious employment barriers, as well as to support employment retention and advancement initiatives. Funding should be focused on programs that operate in partnership with employers, especially those in which services are provided at or near the worksite.

Federal funding for child care should be increased. Child care assistance is important to help parents sustain employment, pay for basic needs, and ensure that children are in environments promoting education and healthy development while their parents are working. While the number of families receiving subsidy assistance has grown since 1996, so has the number of lowincome working families. As a result, there remains a tremendous gulf between the number of families eligible for and the number actually receiving subsidy assistance: the federal government estimates that, in 1999, only 12% of potentially eligible families were receiving assistance through the federal Child Care and Development Fund. The share of eligible families receiving help has probably grown since that time, but not enough to change the basic picture: the great majority of potentially eligible low- income families do not receive child care assistance. Since 1996, the principal source of growth in child care funding has come from TANF, as states were able to redirect TANF funds freed up as welfare caseloads fell. However, it seems clear that states will not be able to rely on continued rapid TANF caseload decline as a funding source for child care in the coming years. Accordingly, we believe that Congress should significantly expand the dedicated funding for child care available through the Child Care and Development Fund, with an ultimate goal of making subsidy help available to all eligible lowincome families."'

Child support assignment and distribution rules should be changed so that families receive more child support. Consistent and reliable receipt of child support can be an important work support, as well as play a key role in strengthening the relationship between a noncustodial parent and child. However, when families apply for TANF assistance, they are required to assign (or turn over) to the government their rights to child support. The child support is used by the government to reimburse assistance costs. The basic rule established by the 1996 law is that the government keeps child support owed while a family receives assistance, while the family keeps child support owed when the family does not receive assistance. Congress should eliminate the two exceptions to this basic rule: the tax offset exception and assignment of preassistance arrears. In addition, states that decide to pass through support to families receiving assistance should not have to repay the federal share of the support. States should be given assistance in converting their distribution rules so that program fiscal stability is maintained, for example, by enacting an appropriate effective date, providing systems funding, and allowing use of TANF funds and credit toward TANF maintenance of effort requirements.

States should be allowed to use federal TANF funds to provide ongoing assistance to lowearning working families, without needing to apply a time limit against working families. Under current law, work policies and time limits policies work at cross- purposes with each other. On the one hand, states are often seeking to encourage families to take any available job, and want to provide help to families who are working in low-wage jobs. But, if federal TANF funds are used to provide that assistance, the month counts against the federal time limit and potentially disadvantages the family in the long run. States should not be restricted in their ability to use TANF funds to help working families.

The Administration's Proposal

A common theme of our work-related proposals is that states have made progress on an employment agenda in many areas, but that there are several important challenges that lie ahead. By and large, the specific proposals we make call for changing signals, incentives, and resources to help move state activities in directions that seem most likely to affect more positive outcomes for individuals with barriers to employment and for those who have moved into the low-wage labor market, including those who continue to receive aid, and those who do not. At the same time, our proposals reflect an appreciation of the critical role of allowing state flexibility in determining the most appropriate strategies to accomplish national goals in the TANF structure.

By contrast, the Administration's recently announced framework for reauthorization takes as its starting point that states have failed to implement policies and procedures consonant with the goals and provisions of the 1996 law. Notwithstanding the employment outcomes achieved during the past five years, the Administration apparently has focused on the fact that, according to current federal reporting, a significant number of families are not "engaged" in work-related activities while receiving assistance: according to FY 2000 participation data, 42 percent of adults are reported to have some hours of participation in work-related activities in an average month. And, despite the strong state employment outcomes, the Administration is apparently concerned that most states have generally not opted to operate large-scale unpaid work experience programs.

The 42 percent figure is, at best, incomplete for two reasons. First, states have been required to report engagement in activities counting toward federal participation rates but were not required to report engagement in other activities, and at least half of states clearly have not done so; accordingly, the true number of engaged individuals was surely higher. Second, in any given month, states would never reach 100 percent engagement because some number of families are newly entering or leaving assistance, some are awaiting assessment or assignment to or the start of an activity, some families are under sanction for failing to participate, and some are unable or not expected to participate because, for instance, they have an infant, or illness, or a severe disability restricting participation.

It is certainly true that most states have not elected to run large-scale unpaid work experience programs under TANF. This was their choice, and it reflected their best judgment about the most effective ways to accomplish the work goals of TANF. Given states' success in increasing employment participation, it is hard to see any basis for second-guessing this choice. While we believe that a different mix of program activities with more emphasis on job quality might have helped families attain better jobs, there is no basis for saying that states have not been extraordinarily successful in increasing employment.

The premise that the principal lesson since 1996 is that the states have not taken seriously the challenge of transforming welfare is simply not borne out by the results that have been achieved to date (as described above), and the directions in which the Administration's specific proposals will push states are both unwise and unworkable.

The Administration has proposed an extensive set of new requirements, and the full details are not yet available. However, key provisions would:

Increase the monthly participation rate from 50% to 70% by 2007, while phasing out the caseload reduction credit. Instead, states could count individuals who left assistance due to employment for up to three months.

Increase weekly participation requirement from 20 hours for single parents with children under 6 and 30 hours for other parents to 40 hours for all families with children age 1 or older.

Provide that in meeting the 40-hour requirement, at least 24 hours must be in "direct" work activities - unsubsidized or subsidized employment, supervised work experience or community service programs, on-the-job training, and school completion for teen parents. Vocational training and barrier removal activities would generally not be countable toward the first 24 hours each week. For up to 3 months in a 24-month period, states could count participation in short-term substance abuse treatment, rehabilitation, and work-related training toward meeting the 24- hour direct work requirement.

The Administration's proposed approach is troubling for a number of reasons:

The proposal would force all states to adopt a program model that no state has chosen to implement.

It is widely acknowledged that states have been successful in refocusing the welfare system on work and in increasing employment among single parents. States have done so with a range of approaches, but no state has elected to operate a program that looks like the model the Administration now seeks to impose on all states.

The Administration's approach reflects a particular program model, and every state has been free to adopt that model under TANF, but no state has elected to do so. In structuring TANF programs, some states have placed strong emphasis on job search efforts aimed at connecting families with employment as rapidly as possible. Some have greatly liberalized their policies to broaden support to families who enter low-wage jobs. Most states significantly reduced the role of education and training in their programs (at least in part due to federal participation rate rules), but education and training remains a significant component in some states. Generally, most states have made only limited use of unpaid work experience and community service programs, and even more limited use of subsidized employment and on-the-job training.

The proposal runs counter to what is known about what works.

The clearest finding from two decades of research is that the most effective welfare-to-work strategy is to provide a range of work-focused employment and training services tailored to individual needs, not a one-size-fits-all model. In these "mixed- strategy" programs, the range of services provided typically included assessment, job search, life skills, work-focused basic education, work experience, and job training, with recipients generally participating in only one activity at a time.'x Many other programs that have been studied relied primarily on either job search or basic education and were much less effective.' None of the successful, mixed-strategy programs described earlier had large work experience components.

As discussed above, the most successful site by far in the NEWWS evaluation - Portland, Oregon - stressed moving individuals into the workforce quickly, but it also emphasized finding good jobs and allowed the first activity for each person to vary depending on skills, work history, and other factors. Portland's impacts on employment and earnings are among the largest ever observed in welfare-to-work programs,"' yet states would not be able to adopt the Portland model under the Bush plan. Most of the activities provided by Portland and other such mixedstrategy programs would not count toward work requirements after the first three months. Yet nearly half (49.5%) of recipients in Portland participated longer than three months; about ten percent (9.9%) participated longer than 12 months. Further, there was no standard hourly participation requirement; while staff worked intensively with recipients to help them participate as much as possible, expectations for participation were tailored to each individual.

Moreover, there is no research base for compelling all states to implement large-scale work experience programs, or restricting the use of stand-alone barrier removal activities or vocational training to only three months. The rationale for work experience programs has traditionally been that when the principal barrier to employment faced by an individual is lack of work place experience, an opportunity to gain such experience can affect subsequent employment and earnings. However, research to date has not revealed that these programs have the expected effects.

There is only limited recent research on unpaid work experience programs, because states have generally not opted to implement large-scale programs, so work experience tends to be, at most, a component within a larger program. However, in a review of research conducted in the 1980s, the Manpower Demonstration Research Corporation (MDRC) concluded, "there is little evidence that unpaid work experience leads to consistent employment or earnings effects. More recently, researchers in Washington state,"' were able to isolate the employment and earnings impacts of six different work activities in Washington's "WorkFirst" (TANF) program, including unpaid work experience.""' The study assessed the impact of these components on employment and on earnings. Work experience (together with Job Skills Training and Community Jobs, a Transitional Jobs program offering subsidized employment) was one of three components serving recipients who were relatively less job-ready. The study determined that work experience increased employment among participants but the impacts were substantially less than either of Job Skills Training or Community Jobs. The work experience program had no earnings effects, whereas both of the other two programs serving less job-ready participants had significant positive earnings effects, with the Community Jobs program being the strongest of the three on both measures. Based on the weak performance of the work experience component, in the current budget for the WorkFirst programs, the $3,000,000 allocated for the program was cut, and the program eliminated.

None of this is to say that work experience cannot contribute in important ways to improving the employability of individuals with little labor market experience. The research suggests that when appropriate skill development and barrier removal activities are added to paid work experience, there can be significant impacts. However, there is simply no basis for saying that all states should be compelled to use unpaid work experience programs in instances in which they believe that other program approaches would be more effective.

Further, many of the families still receiving assistance face barriers that make employment more difficult, and programs aimed at reducing these barriers to work will frequently not fit within the three months allowed under the Administration plan for alternatives to "direct work" activities. For example, recent non- experimental findings from a substudy of three NEWWS sites found that, among recipients without a high school diploma, those who participate in adult education and go on to job training or college see a substantially greater increase in their longer-term earnings, earning an additional $1,542 in the third year of follow-up compared to those who participate in basic education only. Similarly, Portland's strong results in increasing earnings of those without high school diplomas are likely connected to its substantially increasing the percentage of high school nongraduate sample members who attained both a GED certificate or a high school diploma and a second education or training credential (such as a trade license or college degree) over the follow-up period." Yet for those who begin the program without even a high school diploma, it is unlikely that they can complete both basic education and job training within the space of three months.

Similarly, a review of model substance abuse programs found that 14 of 20 typically involved participation for longer than three months.' Further, some individuals will need to move in small, incremental steps toward fuller participation in combinations of activities to reduce barriers and to work. This strategy has proven successful in a rigorous evaluation of the National Supported Work Demonstration, which provided a year of subsidized, structured employment (with gradually increasing levels of hours and responsibility) together with on-the-job training and intensive supportive services. The program resulted in large increases in earnings, which persisted even eight years later. And Supported Work was most successful with the most disadvantaged recipients - those who had received welfare the longest, lacked a high school diploma/GED, or had never worked."" Recent reports on best practices for serving those with barriers highlight the importance of combining work with other activities in flexible ways.""' In contrast, no research suggests that the Administration's formula of 24 hours of work and 16 hours of other activities would be effective for most individuals with barriers.

While the Administration has articulated an overarching goal of improving child well-being, the proposal risks having the opposite effect.

The Administration has proposed modifying TANF's goals to articulate an overarching purpose of improving child well-being. And, the Administration has suggested that under its work proposal, a state would be free to treat structured activities that promote child well-being as countable toward meeting the 40- hour requirement, so long as the 24-hour requirement in direct work activities is satisfied.

In many ways, imposing a 40-hour requirement and then allowing activities related to child wellbeing to count as participation seems unresponsive to the central issues that states must address in efforts to simultaneously promote work and advance child well- being. There is a broad consensus that a central goal of TANF is to expect work by those parents capable of engaging in employment. At the same time, much recent evidence indicates that parents are often entering into jobs with low wages, limited advancement, and lacking key benefits such as health care coverage or paid sick and vacation leave. Moreover, working leavers are frequently in jobs with night or weekend schedules or with varying, irregular schedules. And, research evidence suggests that simply going to work or substituting earnings for welfare income does not, in itself, enhance child well-being; rather, it is important for the work to translate to increased family income. Accordingly, one key way to advance the well-being of children should be to help parents get jobs with better wages, health care benefits, and greater potential for advancement; with schedules that allow parents to be at home at night and on weekends; and with paid sick leave and vacation leave and sufficient flexibility to respond to children's needs. Moreover. as parents move into work, it is essential that they have access to a broad range of child care choices, including access to early education programs that can promote the development of younger children and appropriate after-school activities for older children. Therefore, another key aspect of advancing a child well- being agenda ought to include efforts to ensure that TANF recipients and other low-income working families have access to needed child care assistance.

To be clear, we think it would be a positive development if all parents were better able to volunteer at their children's schools or participate in structured activities with their children. But, we also think that it is particularly important that working parents be able to see their children at night, or be able to take time off when a young child is ill, and that allowing parents to count structured activities as work participation is no substitute for helping parents find jobs responsive to family needs.

Moreover, one of the most troubling findings in the recent research is that increased participation in work-related programs by low-income parents appears correlated with adverse impacts on teens' behavior and school performance. At this point, it is unclear whether this adverse impact is principally a function of decreased supervision, increased stress on parents, or increased responsibilities for teens with working parents. This should not be an argument against work requirements and expectations, but, at a minimum, it would counsel for the importance of helping parents find jobs that are consistent with family responsibilities, and against simply mandating 40 hours of out-of- home participation.

The absence of any new funding would force states to cut other programs supporting low-earning working families in order to meet the new costs imposed by the proposal.

It seems clear that the combination of increased numbers of participants and 40-hour participation requirements would result in billions of dollars of increased costs for state efforts, but the Administration is seeking no new funding. Without additional funds, states would face pressure to cut child care and other benefits for low-income working families in order to meet the requirements. Adding new funds won't solve the problems of the basic design, but, without new funding, states would be forced to dismantle many of the innovative initiatives that they've developed in the last five years.

Apparently, the rationale for seeking no additional funding is that TANF caseloads have fallen significantly since 1996, so there is "enough" money to pay for these and other new initiatives within existing funding. However, making such an assertion essentially dismisses the choices that states have made in committing TANF funds as their cash assistance caseloads fell. Nationwide, as cash assistance spending fell, states increasingly redirected their TANF funds to services and supports for low- income families outside the traditional welfare system. The single biggest redirection of TANF funds has been to child care for low-income working families, but the funds have also been used for a broad array of initiatives, such as transportation assistance, state earned income tax credits, employment retention and advancement programs, services for families at risk of entering the child welfare system, help for homeless families and victims of domestic violence, assistance to immigrants ineligible for federal benefits, and others. By FY 2001, states were spending TANF funds at a rate higher than their basic block grant allocations: such allocations are about $16.5 billion, and state TANF spending in FY 2001 reached $18.6 billion.

Since states are currently spending TANF funds in excess of their annualized block grant levels, it seems fundamental that, if one is asking states to do more things with the same amount of money, one must be prepared to articulate what they should stop doing. Yet the Administration has not identified a single area in which it asserts that states are misspending TANF funds. To the contrary, the argument is also being made about the importance of doing more to address marriage and family formation and child well-being. However, any given dollar can only be spent once, and it would literally be impossible for states to redirect existing TANF funds without cutting other low-income benefits and services. And, since many of those benefits are playing an integral role in helping working families sustain work and avoid welfare, cutting such benefits would be both harmful and counter- productive.

Similarly, there are no "extra" child care funds that could simply be redirected to meet the welfare work requirements. It is certainly true that child care funding has grown substantially since 1996, with states' redirection of TANF funds playing a key role in that growth. However, most federally eligible children still do not receive child care subsidy assistance. The precise percentages may be in dispute, but it is clear that, at current funding levels, only a fraction of eligible families are receiving help, and it surely follows that it would be impossible to redirect existing child care funding to meet welfare work requirements without cutting back current funding that is being used to help low-income working families outside the welfare system. Moreover, the Administration has proposed to provide no new federal child care funding in reauthorization. So, even if there were no changes in TANF work requirements, states would still face the specter of needing to cut existing child care slots for low-income working families in order to manage with funding that would remain frozen despite inflationary pressures.

Thank you for providing me with this opportunity to testify.



LOAD-DATE: April 12, 2002