Statement of Steve Savner, Senior Staff Attorney, Center for Law and Social Policy

Testimony Before the Subcommittee on Human Resources
of the House Committee on Ways and Means

Hearing on Welfare Reform and Work

April 3, 2001

Mr. Chairman and Members of the Subcommittee:

Thank you for inviting me to testify. I am a Senior Staff Attorney at the Center for Law and Social Policy (CLASP). CLASP is a nonprofit organization engaged in research, analysis, technical assistance, and advocacy on a range of issues affecting low-income families. Since 1996, we have followed closely the implementation of the Personal Responsibility and Work Opportunity Reconciliation Act, as well as research concerning its implementation and welfare to work programs more broadly. In addition, we often talk and visit with state and local officials, administrators, people affected by welfare policies, program providers, and others concerned about implementation of efforts to change welfare programs.

Today's hearing focuses on work requirements in TANF and other anti-poverty programs, their effectiveness and lessons learned from which recommendations for TANF reauthorization might be drawn. My comments will focus on TANF program requirements and issues to consider in the context of TANF reauthorization, and draws extensively from "Improving Employment Outcomes Under TANF," (Strawn, Greenberg and Savner, February 2001) ⁽¹⁾

My testimony will highlight the provisions of the law that are intended to promote work, and others which may have also created incentives for caseload reduction without regard to employment outcomes; describe common elements to state responses to the law; the effects of initial state choices; how states have further responded in light of these initial effects, and how these experiences should inform our thinking about reauthorization and possible modifications to the statutory framework that is intended to promote and support employment.

Increasing Employment and Caseload Reduction Have Been Two Key TANF Goals

One of the central purposes of the 1996 welfare law was to promote employment among poor parents. In addition, for many, another central purpose of the law was to reduce the number of families receiving cash assistance. It is important to keep in mind that these dual goals of increased employment and caseload reduction are distinct and success in achieving one goal is not always matched by comparable success in the other. Indeed, these goals are sometimes in tension, for example when a state provides a more generous earnings disregard that has the effect of increasing employment but also increasing the caseload. With the exception of limited funding available through the High Performance Bonus, increasing family income beyond the level necessary to leave welfare is not an explicit purpose of the law, (although states have been free to make investments to achieve such a goal as noted above.)

Many key provisions emphasize one or the other, or both of these goals:

• The block grant funding structure allows states substantial flexibility in the use of both federal and state maintenance of effort funds, including the direct use of funds on a broad range of employment-related services and work supports and the ability to transfer a significant portion of TANF funds to the Child Care and Development Block Grant.

• States were given broad discretion in structuring program rules allowing provisions to create significant financial incentives for employment through earned income disregards, as well as time limits and sanctions for noncooperation with work activities.

• The law establishes "participation rates" for families receiving TANF assistance, and provides that states will risk fiscal penalties for failure to meet the required rates. To count toward the rates, an individual must be involved in one of a listed set of work-related activities for a specified number of hours each week. State are given broad authority to count recipients who are employed or participating in a range of subsidized employment and paid and unpaid work experience toward the new federal participation rates. However, education and training activities only count toward the rates to a very limited extent, and generally cannot count for more than twelve months for individuals not working at least twenty hours a week.

• A "caseload reduction credit" specifies that a state's participation rate requirement can be reduced if a state's caseload declines for reasons other than changes in eligibility rules; this creates a strong additional incentive for caseload reduction.

• The block grant funding structure also places a premium on caseload reduction, because a state's federal funding stays constant whether caseload goes up or down. Caseload decline has been seen as necessary to manage within the framework, and the ability to keep and redirect savings from caseload decline to a broad and flexible range of programs and services creates a strong incentive to reduce caseloads.

Initial State Implementation Patterns

Initial TANF implementation in most states solidified a set of Work First policies that states had begun to implement during the early 1990's. These work first polices have emphasized rapid labor force attachment for as many recipients as possible, relying principally upon job search and job readiness activities for most applicants and new recipients, limiting exemptions form participation, and increasing penalties on those found to have refused to cooperate with work requirements without "good cause." ⁽²⁾ A handful of states adopted policies providing for universal or near-universal participation coupled with broad flexibility about the nature of activities in which an individual might be required to participate. More commonly, however, state policies narrowed the range of allowable activities to restrict access to education and training to achieve a focus on rapid job entry, and in response to the narrowly defined set of federally countable activities. ⁽³⁾

States have general succeeded in meeting the "all families" participation rate, and evidence suggests that many adults are participating in employment-related activities that do not count toward federal participation rates.

• In FY 99 over 38% of adults in single parent families were engaged in the narrowly-defined set of federally countable activities. This exceeded the federally required rate of 35% for the year. ⁽⁴⁾ Insofar as the caseload reduction credit has resulted in the reduction of effective participation rates in almost every state, and the elimination of any effective all family participation rate in a number of states, the actual participation rate achieved is all the more striking.

• Data reported from the Urban Institute's 1999 National Survey of America's Families indicates that about two-thirds of all adults receiving assistance were engaged in work-related activities, including federally countable and non-countable activities. ⁽⁵⁾

The picture that emerges here is that some states may have substantially exceeded the effective participation rates than are applicable after taking into account the caseload reduction credit. In addition, they appear to be engaging substantial numbers of participants in activities that do not meet the narrow federal definition of "countable" activities. States' ability to accomplish these results have been due in part to the additional resources generated by caseload declines, and in part directly due to the caseload reduction credit.

Employment And Earnings Among Current And Former Recipients

As reported more fully by my colleague Mark Greenberg in testimony before the Subcommittee on March 15, 2001, since 1996 there has been a significant increase in employment among single female-headed families, and among current and former TANF/AFDC recipients. There is broad agreement that TANF has played an important but not exclusive role in generating these employment increases.

Studies have consistently found that most families leaving welfare have found work ⁽⁶⁾ and that labor force participation has increased among female-headed families. In addition, an increasing share of TANF adults are employed while receiving assistance -- 28% in FY 99, as compared with 8% in FY 94.

Most employed "leavers" are in jobs with low earnings and limited or no access to employment benefits. In the Urban Institute's nationally representative study, median wages for working TANF leavers in 1997 were $6.61 per hour. Moreover, employed leavers are unlikely to receive employer-provided health care coverage or paid sick or vacation leave; in the Urban Institute study, 23% of employed leavers were receiving employer-provided heath care coverage. ⁽⁷⁾ Studies from individual states have reached similar findings.

Prior research had found that employment loss was a significant problem for welfare parents entering employment, and that the limited earnings growth for those entering employment was principally associated with working more hours or weeks in a year rather than with growth in wages. ⁽⁸⁾ State leavers studies provide little information concerning employment retention and advancement; the studies with some longitudinal data typically suggest some earnings growth over time, but that median annual earnings for adults who have left assistance are probably in the range of $8,000 - $12,000.

These data about the employment patterns of recipients and leavers suggest that strategies should be sought during reauthorization to improve the employment outcomes for current and former recipients.

Limited Information Available From Leavers Studies Or Other Sources About Impacts Among Various Racial And Ethnic Groups

Examination of studies designed to track the income and employment status of families who left the cash assistance caseload during the late 1990s suggests differences among various racial and ethnic subgroups. One national study of former welfare recipients shows that whites are more likely to have left welfare compared to Hispanics and non-white/non-Hispanics, and that Hispanics are less likely to have left than whites or non-white/non-Hispanics. Generally, those who have left have more education, and are less likely to face other employment barriers, such as limited work experience, health limitations, etc.

A study of families exiting welfare in Wisconsin in 1995-1996 reported that 61 percent of the white families receiving assistance left the caseload, compared to 36 percent of the African-American families. In an Arizona study of families exiting welfare in last quarter of 1996, researchers found that while African-Americans made up 34 percent of open cases, they were only 8.5 percent of all families that left the caseload during that quarter. The picture for Hispanic respondents is much less clear-cut, with studies from some states showing them leaving the caseload in disproportionately large numbers, while studies from other states reveal opposite results.

Studies in Arizona, Georgia, and Cuyahoga County, Ohio, show that shortly after leaving welfare, the percentages of African-Americans who are employed exceed the percentages of whites who are employed, and results from Arizona, Cuyahoga County and Wisconsin reveal that African-Americans have somewhat higher quarterly earnings than whites. However, studies in those same areas also showed that a much higher percentage of African-Americans returned to welfare within one year of leaving, compared to whites who left. The data for Hispanics vary considerably on all of these measures from one state study to another. ⁽⁹⁾

These studies suggest a pattern in which African-American recipients are less likely to leave welfare than whites, are more likely to be employed shortly after leaving and at somewhat higher wages, but are also more likely to return to welfare within the first year after exiting. Many questions and possible explanations for these findings present themselves. Why are African-Americans leaving more slowly than whites? If whites leave in greater proportions but are employed less, what other sources of income are they relying on to get off welfare and stay off longer? And what are the prospective policy implications of these data? These questions should lead to a much broader research agenda to further explore these issues.

Individuals with Significant Barriers to Employment Represent Ongoing Challenge

For those who are not working, both current and former recipients, evidence continues to show that many have significant barriers to employment including health problems among recipients, health problems among their children that interfere with work, very limited skills, and domestic violence.

According to the 1999 National Survey of America's Families, over one-third of all adults on assistance had health related problems that interfere with work, 5 percent had a child receiving SSI, 27 percent and not worked in the preceding three years, and 44 percent had education less that high school. Among recipients with no identified barriers to employment, 56 percent were employed, while among those with two or more identified barriers, only 20 percent were employed.

State flexibility to both count a broader set of activities in determining their participation rates, and the flexibility to recognize personal circumstances which should excuse participation beyond the limited exemptions provided for under federal law would help states focus resources on those most in need of services and those most able to benefit from participation.

Differential Treatment of Racial and Ethnic Minorities

While we have much information about people who leave and why they leave, we have relatively little information from leavers studies about the kinds of services people received prior to leaving and the connections between those services and activities and their post-program outcomes. Data from several states raise troubling implications of differential treatment of recipients within local welfare systems based on racial or ethnic origin. An analysis by the Chicago Reporter of Illinois data concerning why welfare cases were closed between July 1997 and June 1999 revealed significant differences appear in the reasons for case closings between whites and minorities. A total of 340,958 cases closed in this period, of which 102,423 were whites and 238,535 were minorities. Fifty-four percent of minority cases, but only 39 percent of white cases, closed because the recipient failed to comply with program rules. Though earned income made 40 percent of white families ineligible for support, earned income made only 27 percent of minority families ineligible.

Similar data are reported in a study of recipients in rural Florida who left welfare between October 1996 and December 1998, carried out by the Florida Inter-University Welfare Reform Collaborative. The study sample of 115 former recipients responded to questions about why they left welfare as follows: 53 percent of whites, as compared to 32 percent of African-Americans, found a job; 8 percent of whites and 22 percent of African-Americans were disqualified for non-compliance with program rules; 6 percent of whites and 17 percent of African-Americans voluntarily closed their cases. These two studies raise important and troubling questions about whether African-Americans and Hispanics are being treated differently than whites.

Finally, a study undertaken in two rural counties in northern Virginia focused on the interactions between welfare caseworkers and recipients. ⁽¹⁰⁾ In this study, 39 recipients (22 African-American and 17 white) were interviewed in early 1996 about their interactions with welfare department caseworkers: how frequently caseworkers notified them about job openings, the extent to which caseworkers emphasized further education, caseworker assistance in locating child care, caseworker assistance with transportation, and whether respondents believed that African-American and white clients were treated fairly by caseworkers.

Except with regard to help with child care, respondents' views on these issues varied significantly by race. Fifty-nine percent of whites, but only 36 percent of African-Americans, indicated that their caseworkers were often or sometimes helpful in providing information about potential jobs. Forty-one percent of whites indicated that caseworkers encouraged them to go to school, particularly if they had not received a high school diploma. None of the African-Americans indicated that a caseworker had encouraged them to go to school. One white respondent stated: "They encouraged me to get my GED. I've been in school since October, working on the GED. I hope to graduate in the spring. My worker kept telling me 'You're smarter than you think.' She really convinced me that I could do it." An African-American respondent stated: "They talk to you any kind of way. They say: 'Go get a job.' I told them that I only had two parts left on my GED and I wanted to finish, they said: 'That's not what this program is about.'"

About two-thirds of all respondents in this Virginia study indicated they had transportation barriers, and all respondents indicated that the welfare agency provided vouchers to pay for gasoline to those who needed them. However, 47 percent of whites indicated that caseworkers indicated they would provide additional forms of transportation assistance, while none of the African-Americans reported receiving such offers of help. For example, one white respondent indicated: "I own my car but I need a brake job. I contacted DSS [Department of Social Services] about my car. She told me she will try to come up with some money to get it fixed." An African-American respondent stated: "DSS gives me money for gas. I have a car and a job, but it needs about $300 worth of work, so I can't use it. I asked DSS if they had any funds for car repairs, but she said I should try to use gas vouchers to take a cab or ride with a friend until I save up enough money to get my car fixed." Finally, nearly half (45 percent) of African-Americans--as well as 18 percent of whites--indicated that African-American clients were not treated fairly by DSS.

While this study looked at a very small sample of recipients, it highlights the importance of a range of discretionary actions by caseworkers concerning the availability of services that may significantly affect the well-being of families receiving assistance and the ability of adults in those families to prepare for and succeed in employment. It also shows the potential for differential treatment based on race or ethnicity in the interactions between recipients and caseworkers.

State Responses to Initial Employment Results

Evidence of changing state policies and administration in response to the initial employment results achieved during the first few years of TANF implementation is still limited, but a set of emerging directions appears to be taking shape.

Postemployment Retention Services

States have responded to the low wages and unstable employment of many current and former recipients by expanding an array of services intended to increase job retention, promote rapid reemployment after a job loss, or both. As of October 1999, about two-thirds of the states were providing case management for at least some recipients who became employed or left cash assistance, and a similar number were providing supportive services such as transportation aid, purchase of work clothing or tools, and payment of work-related fees. Half a dozen states were providing short-term cash payments to help cover work expenses, several offered cash bonuses for keeping or finding jobs or leaving TANF, and several provided cash payments to cover emergencies. ⁽¹¹⁾ Many of these postemployment benefits and services are new and little information about utilization exists.

Postemployment Job Advancement Services

As of October 1999, about a third of states (16) had policies to provide post-TANF services aimed at job advancement. These include contracting directly for education, training, employment, and career counseling services; tuition assistance; and individual training accounts. ⁽¹²⁾ A small but growing number of states - about half a dozen - are creating broader initiatives that are designed to serve working, low-income families generally. In some cases, education and training are provided at the worksite, with services customized to employer needs. As with postemployment retention services, it is unclear how many families are actually involved in these initiatives, but numbers appear quite small.

Changes In Strategies For The Unemployed

Beyond creating Postemployment services, a third state response to the problems of low wages and job loss has been to change strategies for unemployed parents to place greater emphasis on helping them access better jobs. Some states are creating incentives for localities to match parents with higher paying jobs as opposed to any job. In 1999 and 2000, a limited number of states expanded access to postsecondary education or training for TANF recipients. These actions included changing work requirements to allow participation in postsecondary education or training to meet all or most of a parents work requirement beyond the twelve months that could count toward federal participation rates; using TANF funds to create additional work-study positions; creating separate state student aid programs for low-income parents funded with state maintenance of effort dollars; and stopping the federal or state time limit clock for recipients who are full-time students. ⁽¹³⁾ While these state actions may suggest an emerging trend, the overall picture remains one of substantial limitations on access to education and training for TANF recipients in most states.

Work Programs under TANF

As in other areas, states have broad authority to structure work programs including providing wage subsidies to employers who hire recipients, creating transitional jobs that offer temporary employment and skill development activity to enhance participants' employability and help them move into unsubsidized employment, and programs in which participants perform work in exchange for their welfare benefits known variously as work experience, community service, or workfare.

Workfare - Limited State Interest and Questionable Results

Many observers predicted that states would make wide spread use of workfare (work experience, community service, etc.) because they are fully countable toward TANF work participation requirements. However, with the exception a few states and New York City, there has been relatively little use made of these programs. While it is difficult to say with great certainty why more use has not been made of these program options, there are a number of factors which may have contributed to state and local decisions not to implement such programs on a large scale.

• There is no evidence that work experience programs are effective in boosting employment and earnings for participants. Research conducted during the 1980's on several work experience programs demonstrated that in every site but one there were no positive employment and earnings impacts that resulted from participation in the programs. ⁽¹⁴⁾

• As discussed above, states' responses to the 1996 law have been predominantly focused on efforts to help participants gain access to unsubsidized employment, not on simply engaging people in activities while they are receiving benefits.

• One of the explicit purposes of many workfare programs has been to discourage families from receiving assistance.

"One major objective of this approach - call it "pure" workfare - has been to reduce welfare dependency by reducing the real benefits of welfare; and this has been accomplished by assigning a work requirement to the receipt of welfare benefits. Thus the work requirement was expected both to deter individuals from enrolling in AFDC, as well as to encourage earlier exits than would otherwise occur." ⁽¹⁵⁾

• Large scale programs can be costly and are difficult to administer effectively. MDRC data showed annualized cost per filled slot ranging from $700 to $8,200. ⁽¹⁶⁾

• Large-scale programs raise critical concerns about the potential displacement of regular employees in the organizations where participants are placed. ⁽¹⁷⁾

Programs that involve the provision of services that are of value to an employer require careful monitoring and oversight to assure that regular employees are not displaced, and to assure that participants' rights under laws to protect employees generally are fully protected.

Transitional Jobs Programs - A Promising Model

For the past several years CLASP has provided technical assistance to a number of state and local TANF agencies and officials to help them design and implement programs that provide time-limited employment in combination with skill development activities and other support services for TANF recipients who have been unsuccessful in finding unsubsidized jobs after participation in job search and other programs. Many of these programs are funded by a combination of TANF and Welfare-to-Work block grant funds, and they are fully countable toward TANF work participation requirements. The potential advantages offered by such programs over work experience/workfare programs is that provide work wages rather than mandating work for welfare in exchange for benefits. This makes it more likely that participants, supervisors, and prospective future employers will attach more importance to these work relationships, and that they will generate much greater skill development than in workfare type programs.

Currently two states, Washington and Vermont, and over 20 cities and other local jurisdictions are operating such programs, typically on small scale. ⁽¹⁸⁾ Washington state, the largest program, currently enrolls about 1,500 to 2,000 participants in transitional jobs at any point in time. Some of these programs have shown extremely promising results in terms of employment outcomes. For example, a study in Washington State revealed that two-thirds of participants were employed after leaving the program, and that there were high levels of satisfaction with the program by participants, site supervisors and program managers. ⁽¹⁹⁾ There is little information yet about job retention or post-program advancement. Nonetheless, this new set of transitional jobs programs offers an attractive model, particularly for jurisdictions that are turning their attention to those clients who have significant barriers to employment, and may also offer an effective model for clients for whom a combination of work experience and skill development may lead to substantially higher paying starting jobs than might otherwise be available. ⁽²⁰⁾ Federal support for research and evaluation, as well as technical assistance for innovative program models such as these should be included as part of reauthorization.

Recommendations for TANF Reauthorization

The following recommendations share a common theme that goes beyond the goals of employment entries and caseload reduction to poverty reduction and the need for each state to assist low-income families to achieve an adequate and stable source of income. These recommendations suggest ways to achieve these goals that combine state flexibility, a clear statement of these expanded purposes, and a meaningful system of accountability that assures good faith state efforts to achieve agreed upon goals and meaningful protections for the fair treatment for those who receive or who seek to receive assistance and/or services from state and local agencies.

First: The purposes of TANF should be revised to include an express goal of reducing family poverty and promoting family economic well-being, and to make explicit that the goal of promoting work includes supporting employment retention and workforce advancement for needy families. The purposes of TANF affect whether particular expenditures are possible and have an important signaling effect in communicating Congressional expectations. Modifying the purposes would provide a powerful statement that the next stage of TANF implementation envisions higher goals than caseload reduction.

Second: States should be required to describe in their state plans how TANF and other resources will be used and coordinated in efforts to promote employment retention and advancement and enhance family economic well-being. This would reinforce the signaling effects, and perhaps help foster coordination. While the federal government should not mandate a single strategy, states should be expected to expressly articulate the strategies that they intend to use.

Third: Measures of state performance in TANF should place a strong emphasis on poverty reduction, higher wages, sustained employment and earnings growth. The law currently provides for $200 million per year for high performance bonuses, and HHS has allocated those funds based on state outcomes including employment entries, retention, and earnings gains. In the context of the overall block grant structure, the existing high performance bonus involves a small amount of money, and generates relatively little attention.

Changes should be considered that explicitly address the extent to which low-income families develop income in excess of the federal poverty level. In addition, rather than framing these performance bonuses as an interstate competition, consideration should be given to a system that more explicitly targets continuous improvement for each state. This could be achieved by measuring each state's performance against its own performance in prior years, and in comparison to benchmarks set for each state by agreement between state and federal officials. State performance in relation to these benchmarks might generate both penalties for extremely weak performance, and bonus for very strong performance. The performance measurement system established under the Workforce Investment Act provides a potential model to be considered in the context of TANF reauthorization.

There are a set of difficult issues to consider in how goals would be set, how performance would be measured, and how adjustments would be made for economic conditions and unforeseen circumstances; at the same time, it is fundamental that in a context of broad flexibility in use of resources, the federal focus should be on measuring and seeking accountability for key outcomes. In any case, performance standards should measure outcomes for families receiving TANF assistance and for a broader group of low-income families. Many of the ways states now use TANF resources involve efforts to build supports outside the welfare system so that families need not seek TANF assistance. A declining share of block grant funds are actually expended on TANF assistance recipients, and measuring state performance should consider labor market participation and poverty status of all low-income families, not just those in the cash assistance system.

Fourth: In the long run, a shift to outcome-based measures rather than participation rates would be desirable, in the interim, if participation rates are continued:

• the definition of countable activities should be broadened by removing restrictions on education and training and by including other activities agreed upon by participants and state and local agencies as being consistent with individual employment plans,

• states should have increased flexibility to recognize that there may be periods of time and circumstances when caregiving for family members may make participation in employment related activities inappropriate, and

• participation rate reductions should be based on states' success in placing individuals into stable employment rather than their success in reducing the caseload.

In a context of a smaller caseload including many individuals with significant employment barriers, the restrictive listing of countable activities works against states' ability to structure services and individualized plans for individuals with multiple barriers and severe basic skills deficits. Because of the caseload reduction credit, many states now have very low effective participation rates. The first impulse for some may be to want to raise rates, but simply raising rates without considering what counts and without addressing the perverse incentives flowing from the caseload reduction credit would only exacerbate the risks that states would not develop effective service strategies for families with multiple barriers. The recommendations noted above will further the goal of providing meaningful and effective employment services to the broadest number of individuals.

Fifth: The federal agencies should vigorously monitor state and local performance regarding implementation of civil rights and employment rights protections afforded under current law, and should assistant participants with vigorous enforcement when appropriate. Several studies have identified troubling and apparently discriminatory treatment of racial and ethnic minorities. More broadly, there appear to be differential results for various racial and ethnic groups and little information as to why these are occurring. A two-fold strategy of further monitoring and research to more clearly understand what is happening is essential to insure that all program participants are treated fairly and equitably. In addition, as the reality of discriminatory treatment has arisen, vigorous enforcement of civil rights and employment rights laws becomes an increasingly important element of federal oversight and this federal role should be highlighted and reinforced during the reauthorization process.

1.Paper prepared for the "New World of Welfare" Conference, held January 31 - February 2, 2001, organized by the University of Michigan's Ford School of Public Policy and the Brookings Institution. The paper will appear in a book of conference papers forthcoming in 2001 from the Brookings Institution Press.

2.State Policy Documentation Project, 2000, www.spdp.org

3.For a more complete discussion of initial state implementation decisions on these issues, see, "Improving Employment Outcomes Under TANF," Strawn, Greenberg, and Savner.

4.Third Annual Report to Congress on TANF, US HHS, August, 2000, Table 3:1

5."Do Families on Welfare in the Post-TANF Era Differ from their Pre-TANF Counterparts?" Zedlewski and Alderson, (Urban Institute-Assessing the New Federalism, February 2001.)

6."Families Who Left Welfare: Who Are They and How Are They Doing?" Loprest, (Urban Institute, 1999).

7.Id.

8."Steady Work and Better Jobs," Strawn and Martinson, (MDRC, 2000).

9.See generally, " 'Leavers' and Diversion Studies: Preliminary Analysis of Racial Differences in Caseload Trends and Leaver Outcomes," Elizabeth Lower-Basch (HHS, December 2000); "Welfare reform and Racial and Ethnic Minorities: The Questions to Ask," Savner, Reprinted from Poverty and Race, Poverty and Race Research Action Council, Volume 9, Number 4, page 3 (July/August 2000) available at http://www.prrac.org/newslet.htm.

10. "All Things Not Being Equal: Differences in Caseworker Support Toward \Black and White Welfare Clients," Susan Gooden, (Harvard Journal of African American Public Policy, Volume IV 1998.)

11. State Policy Documentation Project, (2000), www.spdp.org.

12. Id.

13. See State Policy Documentation Project, 2000 (www.spdp.org) for a full list of state TANF policies toward postsecondary, and Wamhoff and Strawn, forthcoming 2001, for summary of recent developments.)

14. "Unpaid Work Experience for Welfare Recipients: Findings and Lessons from MDRC Research," Brock, Butler, and Long (MDRC 1993).

15.  Lessons for Welfare Reform: An Analysis of the AFDC Caseload and Past Welfare-to-Work Programs, Dave O'Neill and June O'Neill (W.E. Upjohn Institute, 1997) pp. 76-77.

16. "Unpaid Work Experience for Welfare Recipients: Findings and Lessons from MDRC Research," Brock, Butler, and Long (MDRC 1993).

17. Litigation in New York City, Saunders v. City of New York, includes information for example that during The number of City Parks workers declined from 1,251 in December 1993 (the month before the Guiliani Administration took office) to 802 in November 1998, while the number of workfare workers increased from 836 in October 1994 to over 6,000 in September of 1998.

18. See an overview and survey of TANF and Welfare-to Work funded Transitional Employment programs from Richer and Savner, CLASP, forthcoming, Spring 2001.

19. "Community Jobs: Outcomes Assessment and Program Evaluation," Case, Burchfield and Sommers, (Economic Opportunity Institute, 2000).

20. See also, "Transitional Jobs: A Bridge Into the Workforce for Hard-to-Employ Welfare Recipients," Anne Kim, (Progressive Policy Institute, March 2001).