A New Paradigm for Welfare Reform: The Need for Civil Rights Enforcement

A Statement by the U.S. Commission on Civil Rights

August 2002

Passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) in 1996 was intended to drastically transform public assistance in the United States. With it, a new emphasis was established to move public assistance recipients from welfare to work. While a laudable goal, rather than engaging recipients in productive activities that lead to self-sufficiency, the reform instituted tough requirements and restrictions on eligibility (including time limits, work requirements, and participation rates). The law gave states discretion to structure programs, as long as they met basic requirements, and impelled them to enforce strict sanctions.

The Commission has evaluated the 1996 law against new executive and legislative proposals for reauthorization. Because women and people of color are disproportionately affected by public assistance policies, the Commission’s goal is to ensure that civil rights protections are built into welfare reform. The Commission’s analysis resonates with numerous studies, which have found: there are disparities in access to and utilization of services, there is discrimination in the delivery of welfare benefits, whether intentional or not, and civil rights considerations are paramount. Further, the strict requirements imposed by the 1996 law, which would be even more burdensome if the reauthorization proposals before Congress are passed, are most detrimental to the “hard to serve” populations, including recipients with low levels of education, individuals with disabilities, and immigrants, as well as women of color. A significant complicating factor is that many families are just beginning to reach the five-year limit on benefits, therefore, a true evaluation of the reform has yet to be performed.

The Commission found that the proposals before Congress not only ignore some of the negative outcomes of the 1996 reform, but potentially compound the disparate impact of the 1996 law. Without civil rights protections in the legislation, welfare reform cannot lift all Americans out of poverty. Based on its own review and numerous studies, the Commission encourages Congress to promote policies that will alleviate the disparities and advance the objectives of reform. The Commission’s recommendations are offered in three categories: (1) those that will facilitate the enforcement of civil rights laws, (2) those that will safeguard against discriminatory treatment, and (3) those that will prevent future disparate impact.


Neither the 1996 law nor the current proposals adequately define the applicability of civil rights laws to welfare recipients, and there does not appear to be movement to ensure that the civil rights laws are appropriately enforced in the delivery of welfare services. As will be shown, numerous studies prove that welfare reform has done little to eliminate historical discrimination in public assistance. People of color encounter insults and disrespect as they attempt to navigate the welfare system. Women are subjected to sexual inquisitions at welfare offices and sexual harassment at job activities. Individuals with limited English proficiency encounter language barriers. Immigrants are often turned away because of misconceptions about their eligibility status.

Institutional racism and discriminatory practices constitute significant barriers to job security and mobility, and hence earning potential. By promoting “work first” as the central objective of welfare reform, PRWORA assumes that welfare recipients face a level playing field in the labor market, an assumption that has repeatedly proven false. Unlike other employees, welfare workers who experience discrimination often do not have recourse options. The cost for filing a discrimination complaint is much higher for welfare-dependent and other low-wage workers because of the fear that if they file a complaint, they will lose employment and subsequently other benefits.

Furthermore, research has shown that despite the presence of civil rights protections, many individuals in the welfare system are subject to treatment that is discriminatory and illegal. For example, in many cases individuals with limited English proficiency are denied access to service because they cannot communicate with caseworkers or understand written materials. One study found that more than 75 percent of immigrant adults in Los Angeles and nearly two-thirds in New York are limited English proficient; and more than 50 percent in Los Angeles and 38 percent in New York do not speak English well or at all.[1] Another study found that 87 percent of the Vietnamese immigrant recipients and 48 percent of Mexican immigrant recipients in California had limited or no English proficiency. Many were also not literate in their native languages.[2]

In 1999, the Office for Civil Rights in the U.S. Department of Health and Human Services (HHS) found that the failure of welfare offices to provide translation services to non-English-speaking clients has the effect of discriminating on the basis of national origin and is a violation of federal civil rights laws. Yet, it appears that little has been done to enforce compliance, and state agencies have remained inadequate in the provision of services to language minorities.

Civil rights enforcement efforts are hampered by the fact that relatively little data have been collected on the distribution of benefits, sanctions, and access to services by race and ethnicity, and there is no uniform national standard for such data collection. This makes it difficult to draw conclusions about the civil rights implications of welfare reform at the national or state level, leaving policymakers unable to assess program deficiencies and guessing at how to redesign reform.

Recommendation 1: Congress should take the reauthorization of welfare reform as an opportunity to clarify and strengthen the applicability of civil rights and labor laws to participants in TANF (Temporary Assistance to Needy Families) activities and to reiterate the legal requirements of state agencies and employers.

Recommendation 2: Congress should allocate funding for stronger enforcement of civil rights laws, improved training of caseworkers, and outreach to employers of welfare recipients. Congress should give the Office for Civil Rights at the Department of Health and Human Services the resources necessary to investigate allegations of discrimination and monitor the activities of state welfare offices to ensure adherence to federal civil rights statutes.

Recommendation 3: The Administration for Children and Families within HHS should collect and disseminate standardized data by race/ethnicity on welfare recipients, those denied benefits, those sanctioned, and those exempted from work requirements. Data should also be disaggregated by subpopulations, particularly with respect to immigrant welfare recipients, so that state and local agencies can assess usage patterns and better determine the unique needs of various communities. The data should be produced in a uniform and easily accessible format and made available to researchers and practitioners.

Recommendation 4: HHS should conduct regular audits of state welfare agencies. States must be required to adopt grievance procedures. States should be required to develop a plan for dealing with noncompliance with federal civil rights laws and submit it to HHS, and be monitored for a set number of years until the problem is resolved and the compliance goals are met.


Despite the absence of national data, many individual organizations have conducted studies that have resulted in similar findings; there is enough evidence to suggest that there are in fact disparities in access to and receipt of services across racial and ethnic lines. Caseworkers, who have great discretion in connecting recipients with available services, often discriminate, whether intentionally or not, in the services they offer. For example, numerous studies have found that white recipients are more likely to be encouraged to pursue an education, are less likely to be sanctioned, and are more likely to receive child care subsidies than other groups. Other studies have found that welfare agencies are least helpful to blacks in providing job-readiness skills and more helpful to whites, Hispanics, and Asian Pacific Americans. They are also least likely to provide basic academic skills, enrichment, or tutoring services to black recipients.

The National Urban League Institute for Opportunity and Equality found that, generally, minority working mothers on TANF do not receive the necessary subsidies to transition to work—including child care, transportation assistance, and college degree assistance—at the same rate as white working mothers. More than 70 percent of Hispanic and African American women did not receive any subsidies for work-related activities as compared with 62 percent of white women. At the same time, African American and Hispanic women are more likely to work at unpaid jobs for benefits (65 percent and 72 percent, respectively) than are white women (46 percent).[3] The study highlights the need for stronger quality assurance measures in the implementation of TANF to ensure consistency in the distribution of support services across racial and ethnic groups.[4]

According to the Urban League, the most significant disparities exist among support services in which caseworkers are likely to have the most discretion. The study concludes that differences in support service utilization rates may explain differing rates at which racial/ethnic groups successfully leave welfare for work.

Another recent survey of post-1996 welfare recipients in 13 states revealed that people of color have encountered insults and disrespect as they have attempted to navigate the welfare system. The survey also found that women are frequently subject to sexual inquisitions at welfare offices and sexual harassment at job activities, often with no recourse. Individuals whose first language is not English have encountered language barriers, despite federal protections designed to guard against that barrier. Eligible immigrants are often turned away and have been told to “go back where they came from.”[5] Specific findings of the survey include the following:

With respect to service quality, another study found that in two Virginia counties, 41 percent of white recipients, but none of the black recipients, were encouraged to pursue education, and 47 percent of white recipients and no black recipients reported receiving transportation assistance beyond gas vouchers.[7]

Similar findings were made in Mississippi, one of the nation’s poorest states, where black recipients were found to be less likely to receive access to services than whites.[8] Welfare agencies in Wisconsin were found to be least helpful to blacks in providing job-readiness skills and more helpful to whites, Hispanics, and Asian Pacific Americans. They were also least likely to provide basic academic skills, enrichment, or tutoring services to black recipients. These recipients were more likely to have their food stamp benefits reduced, and to have to pay for medical services than any other racial/ethnic group.[9]

Evidence suggests that people of color and language minorities are often disparately affected by welfare rules and restrictions. For example, states with higher percentages of Hispanic and black recipients at the time of welfare reform were more likely to adopt shorter time limits, family caps on benefits, and stronger sanctions than states with lower percentages of minority recipients.[10]

Whites are less likely than other former recipients to leave welfare for administrative reasons, such as not following program rules, administrative mistakes, or reaching time limits on benefits. White former recipients also are more likely to receive help with expenses in the first three months after leaving the rolls than are black former recipients. Blacks report leaving welfare because of administrative problems, time limits, or noncompliance with program rules much more frequently than whites or all other races combined.[11] Nationally, whites leave the rolls at faster rates than minorities, and thus make a faster transition to work. The decline in welfare rolls has been 25 percent for whites, 17 percent for African Americans, and 9 percent for Hispanics.[12] While there may be many causes for this occurrence, researchers have documented that racial discrimination in employment and discriminatory referral policies on the part of caseworkers play a role.[13]

Differential and discriminatory treatment extends beyond the welfare office to employers and agencies that hire welfare-to-work individuals, suggesting further need for increased civil rights monitoring and enforcement. For example, research has found that small and suburban employers are less likely to hire black or Hispanic welfare recipients. Other studies have shown that, even when they have more education than whites, black welfare recipients receive shorter employment interviews (more than half are fewer than five minutes long), and among some temporary employment agencies, there is extensive evidence of racial discrimination in hiring for entry-level jobs.[14] Another survey found that, compared with white recipients, black welfare recipients are also more likely to be subjected to pre-employment tests (usually drug or criminal background checks), are more likely to have to work undesirable evening hours, and are less likely to have a positive relationship with their employers.[15]

In addition to women of color in general, several specific populations have a distinct stake in the reauthorization of welfare reform: immigrants, American Indians, and persons with disabilities. Each has unique needs that remain largely unmet under the current law. The Commission has given careful consideration to possible remedies, out of which flow the recommendations presented here. 


The 1996 law prohibited states from supporting legal immigrants with TANF funds until they have resided in the United States for at least five years. As a result of these restrictions, many immigrants have left the rolls, and the living conditions of these poor families continue to decline. Today, significantly fewer legal immigrants, although eligible, receive TANF assistance, food stamps, and Medicaid. The changes to eligibility had a significant effect on children of immigrant parents; even the participation of U.S. citizen children who live in immigrant families has declined.

The effort to restrict immigrant eligibility was largely premised on false perceptions about the group’s reliance on public assistance. Data show that, contrary to public perceptions, immigrant families are less likely to receive welfare than are citizen families, as was the case prior to 1996. Also contrary to public opinion, immigrants generally do not come to the United States with the intent of receiving “handouts.” In fact, 14 of 19 “new growth” states (i.e., states that have seen a significant increase in immigration) offer no public assistance for new immigrants. New arrivals locate themselves based on the availability of jobs, not on the likelihood of receiving better benefits.[16]

As a result of the 1996 reform measures, 60 percent fewer legal immigrants, although eligible, received TANF assistance in 2000 than in 1995; 48 percent fewer received food stamps; and 15 percent fewer received Medicaid.[17] In other cases where immigrants are eligible for benefits, many do not receive them because they fear retribution from the government, such as deportation.[18] Although, in 1997, Supplemental Security Income (SSI) was reinstated to immigrants who were in the United States prior to the 1996 reform, many immigrants with disabilities lost their coverage and the additional benefits afforded to them through public assistance programs. In many cases SSI was their only means of financial support or medical services.

An Urban Institute-sponsored study, conducted by the Survey Research Center of the University of California at Los Angeles, of immigrants in Los Angeles County and New York City demonstrates reduced benefit use, despite substantial levels of need, among immigrant families in programs directly affected by the eligibility restrictions imposed in the 1996 welfare reform.[19]

As noted above, more than 75 percent of immigrant adults in Los Angeles and nearly two-thirds in New York are limited English proficient; and more than 50 percent in Los Angeles and 38 percent in New York do not speak English well or at all.[20] Adults with limited English proficiency are also poorer than immigrant adults overall, with poverty rates at more than 30 percent in both Los Angeles and New York, despite having higher work force participation rates than poor non-immigrants.[21] Language access in cases such as this is not only vital to welfare participants, but also required by law. Without language assistance, even eligible immigrant families are less likely to receive appropriate services and opportunities to transition to better employment.

For instance, a study of the Hmong community in Wisconsin found that this group faces many barriers to employment, including high rates of illiteracy, cultural and linguistic isolation, and lack of skills. Despite this knowledge, welfare agencies are not addressing employment barriers specific to the Hmong community, are failing to provide specialized training or literacy assistance, and are placing them in work assignments that provide little or no skill development.[22]

The additional barriers faced by immigrants will be compounded by stricter work requirements and definitions of what qualifies as a work activity. Many of the non-cash services, such as counseling, training, English instruction, and education, would benefit new immigrants and help lift them out of low-paying jobs. Further, there is great variation within immigrant communities and among those who receive public assistance, raising the concern that state and local infrastructures may not be equipped to address the distinct needs of each group.

One researcher noted that limited English, lack of education, and limited job skills severely restrict immigrants’ options in the job market, making it difficult for them to comply with welfare-to-work requirements. Language problems also impede their ability to negotiate the welfare bureaucracy, which provides very limited or no translation services, and denies them information about programs to which they are entitled.[23]

Recommendation 5: Congress should immediately restore full benefits to legal immigrants, regardless of date of entry to the United States. Benefits should not be contingent on the financial resources of their sponsors, who may be unable or unwilling to help, especially in times of economic hardship. Congress should also allow access to certain public assistance programs to undocumented immigrants, such as health care, education, and food stamps. For the well-being of these families, particularly their children, all immigrants should have access to the basic human necessities, at the very least.

Recommendation 6: In keeping with the civil rights laws and guidelines already established, HHS must strengthen its monitoring and enforcement of language assistance requirements. Language assistance must be provided to welfare recipients who have limited English proficiency throughout the public assistance process. Congress should require HHS and the Department of Labor to update and circulate guidelines. All states, but particularly those with large language minority populations, should be required to put in place procedures for providing translation services. Welfare offices should partner with advocacy and community groups to ensure that volunteer interpreters are available when needed. In addition, written materials, including program offerings and eligibility rules, should be translated in appropriate languages.

Indian Tribes

Under PRWORA, federally recognized Indian tribes were given the authority to manage their own TANF programs, and they have generally welcomed this discretion. Many tribes have established independent requirements that reflect the unique economic and social conditions among tribal communities. However, despite efforts to stimulate economic development, tribal communities remain poor, and unemployment remains high. Geographic isolation and lack of education and job skills have hampered economic growth on reservations. Furthermore, many tribes lack the infrastructures and expertise to develop programs that will adequately serve the needs of their populations and have received little assistance from either state or federal government. According to one tribal leader, while tribes have admirably assisted their clients in a short time, “it is naturally self-evident that our programs do not have the resources, experience or infrastructure that state programs use, with the assistance of federal funding, in their daily administration of welfare services.”[24]

Between 1994 and 2001, the number of American Indian families receiving cash assistance through TANF state programs decreased to 26,000 from nearly 68,000.[25] It should be noted that a portion of this decline can be attributed to the fact that many Indians participated in tribal TANF programs rather than state programs. However, it is estimated that tribal programs only serve approximately 22,000 families, making the decline in participation still significant but less than the decline among the general population. In six states, the proportion of the caseload composed of American Indians has actually increased since welfare reform. For example, in 2001 in South Dakota, 80 percent of individuals receiving cash assistance were American Indian, despite making up only 8 percent of the state’s population. The U.S. General Accounting Office (GAO) estimates that the overall decline in the number of American Indians receiving cash assistance can be attributed to decreased usage among Indians living off reservations, not those on reservations.[26]

Recommendation 7: Congress should provide resources and technical assistance to tribal TANF offices to assist them in the development of programs and infrastructures. Congress should provide to tribes capacity-building and technical assistance grants, similar to those provided in state programs, so they can improve the administration of their own welfare assistance programs.

Recommendation 8: Congress should render tribal TANF programs eligible to receive performance incentives as an inducement for creating and maintaining successful programs. Tribes should also be provided funding for management information systems, technical assistance, transportation grants, vocational and educational opportunity grants, and community and economic development grants.

Persons with Disabilities

Individuals with disabilities make up a segment of the population that was largely ignored by welfare reform and will be disproportionately affected by provisions built into the reauthorization proposals, such as increased work requirements. While many persons with disabilities are eligible for Supplemental Security Income, the strict eligibility requirements of SSI have forced others to rely on TANF assistance.

It is estimated that more than 40 percent of TANF recipients have impairments or are caring for a child with a disability, compared with 15 percent of the non-TANF population.[27] The National Council on Disability estimates that of the “hardest to serve” individuals remaining on welfare since the enactment of PRWORA, more than half face barriers because of learning disabilities, mental retardation, and emotional or behavioral problems.[28]

Based on analyses of current TANF recipients across several states, it is estimated that a quarter to a third have serious mental health problems; more than 20 percent have physical impairments that limit their ability to work; a fifth to a third have learning disabilities; and 20 to 25 percent have IQs of less than 80.[29] Of former TANF recipients, between 20 and 40 percent of those who left TANF and are not working are not working due to a disability or health condition. Of those, 25 to 50 percent are no longer using TANF due to a failure to comply connected to their disability or health condition.[30] Some studies indicate that those who have learning disabilities or low intelligence are noncompliant because it is difficult to understand the complicated rules of the program. Other studies have shown that recipients with health problems are more likely to be sanctioned for noncompliance than nondisabled recipients (50 percent as compared with 39 percent).[31] Disabilities can make it difficult for individuals to find and keep jobs, thus, making it difficult for them to meet TANF requirements. Increased work requirements, without needed supports, will place these individuals at a greater disadvantage.

A study by the U.S. General Accounting Office, based on data collected in the Census Bureau’s Survey of Income and Program Participation, found that recipients with impairments are half as likely to leave TANF as recipients without impairments. They are also less likely to be employed after leaving the welfare rolls.[32] In their first month after leaving TANF, 36 percent with impairments reported having no earnings (including from SSI), as compared with 23 percent of other welfare leavers.[33] GAO also found that many recipients with impairments are not receiving the assistance needed to move toward employment, and while they are sometimes exempted from work requirements, they are often not exempted from time limits.[34]

Recommendation 9: Congress should encourage states to develop community-based programs that assist individuals with disabilities in finding work. Time limits should be eliminated for individuals with disabilities who have insurmountable barriers to employment, and these individuals should not count against state exemptions for extreme hardship cases.


Unduly restrictive rules for work requirements, participation rates, countable work activities, and time limits established under the 1996 welfare reform law make it difficult for many TANF-receiving women to sustain productive employment. The provisions adopted disproportionately affect people of color, individuals with disabilities, and those with limited English proficiency. In addition to the discriminatory effects of poorly designed and implemented welfare programs cited here, the Commission is concerned about policy proposals that have the potential to disproportionately affect certain populations. Thus, the Commission recommends prophylactic measures that will minimize the projected risks associated with the stricter requirements.

The proposals call for tougher work requirements without providing the help recipients need to find decent employment and benefits. Rather than affording individuals the opportunity to obtain skills and experiences that have the potential to move them into better, living-wage jobs, this approach requires individuals to take low-wage jobs without advancement potential, simply to comply with the definition of work activity.

The restriction on the number of people in each state who can count education as a work activity has prevented states from allowing many welfare recipients to pursue education. This is a disturbing trend considering the proven relationship between education and income. Studies have found that TANF recipients who are not working have significantly lower levels of education than those who are working. Further, skill patterns of women on welfare reveal disparities along racial and ethnic lines. While 51 percent of white recipients have “competent, advanced, or superior” skills, only 17 percent of African American recipients and 16 percent of Hispanic recipients do.[35] At the same time, 46 percent of African American women and 29 percent of Hispanic women on welfare have basic skills and could raise their earning potential though one semester of coursework.[36] However, as noted above, evidence shows that race and ethnicity often determine who is given access to the limited education programs in existence. A recent study of the National Urban League found that PRWORA’s anti-education policies have had a significant impact on African American TANF recipients in particular. State policies that do not allow college courses to count toward work requirements have resulted in significantly lower college enrollment for these women.[37]

In addition, the training and education programs states offer are often inappropriate for the needs of individual recipients. Frequently, states take a one-size-fits-all approach to training. National research has found that the most successful welfare-to-work programs are those that provide services tailored to different recipients’ needs (for example, assistance to overcome language barriers or specialized training to accommodate disabilities).[38]

While it is still too soon to determine the true impact of the initial five-year limit, as many are just now reaching that point, there is widespread concern that many of the individuals who will be forced off the rolls due to time limits are the hardest to employ, those with disabilities, and those for whom the system has failed. Negative and discriminatory treatment of individuals on welfare is also likely to affect their ability to succeed off the rolls, through job retention and increased earnings, which is especially critical in an era of time-limited benefits.[39]

Prior to the 1996 reform, the Urban Institute found that blacks and Hispanics tend to remain on welfare for longer periods of time, and thus would be more seriously affected by time limits. Data at the time suggested that 41 percent of black recipients and 51 percent of Hispanic recipients, while only 27 percent of white recipients, would be forced off the rolls due to time limits.[40] More recent research estimates that black women are 55 percent more likely, and Hispanic women are 90 percent more likely, than white women to spend at least five years on welfare.[41] More than 20 states implemented time limits that were shorter than five years, and there is evidence that the shorter time limits have already had a disparate impact.[42] For instance, in Utah, 53 percent of Hispanics and 48 percent of Native Americans who left the rolls since welfare reform did so due to time limits. Only 24 percent of white recipients left for the same reason. In Florida, 70 percent of the people who have left welfare because they reached time limits are African American.[43]

Finally, numerous studies have demonstrated that navigating the welfare system is often challenging for families, resulting in lack of access to the programs and support systems designed to move them out of poverty. In addition, as this review demonstrates, customer service concerns are often linked with civil rights. It is frequently the case that people of color are subjected to differential treatment in the provision of services, and people of color and language minorities are often disparately affected by welfare rules and restrictions. While the Commission acknowledges that legislating customer service is difficult, measures can be taken to ensure that state and local welfare agencies are adequately serving the communities that rely on their services and that services are delivered in an equitable manner.

Recommendation 10: Given that many of those remaining on the rolls are the hardest to employ, the employment difficulties faced by racial and ethnic minorities, and current economic conditions, which do not favor full-time employment for many, Congress should not increase work requirements, and in some cases reduction, perhaps to 20 hours per week, should be allowed for individuals such as single parents of young children, persons with disabilities, and those with other extenuating circumstances. Requiring only part-time work while providing assistance will better enable these recipients to pursue education and job training.

Recommendation 11: Congress should include a broader range of education programs that meet the work-related activities requirement, such as adult basic education, literacy training, English as a second language classes, GED preparation, and postsecondary education. Allowing English as a second language instruction is necessary to bridge the information gap for immigrants.

Recommendation 12: Congress should not place restrictions on the length of time education can be counted toward work, so that welfare recipients have realistic opportunity to move through education programs that will result in better jobs. Nor should Congress restrict the proportion of state recipients enrolled in postsecondary or vocational education, thus forcing states to limit this opportunity to a select few.

Recommendation 13: States should be required to develop an individualized approach to training, so that appropriate “curricula” can be developed for each recipient based on her or his needs. Individual needs assessments should be conducted and a tailored plan developed for each recipient prior to training or employment assignments. Plans must be reviewed for progress on a quarterly basis and revised if necessary, and there should be in place a review mechanism to determine that all recipients have access to available education programs and that individuals are not being unfairly steered away from such opportunities.

Recommendation 14: Congress should abolish the mandatory 60-month time limit for participation in the TANF program, and recipients should be evaluated for eligibility on a case-by-case basis. Alternatively, Congress could allow extensions of time limits based on assessments of the needs of those perennially on the rolls, who are likely to be among the hardest to employ. If a time limit is imposed, Congress should require states to temporarily suspend participation for recipients who are employed, regardless of whether that employment is full time or part time. Congress should allow states the discretion to extend time for all recipients who face hardships such as a disability, taxing family responsibilities, or loss of employment.

Recommendation 15: Congress, by establishing minimum standards for service delivery, should ensure that states make efforts to better serve welfare recipients, for example, by expanding office hours and locations, and where possible, providing child care during visits. Caseworkers should be required to conduct on-site visits to the hardest to reach individuals, for instance, through monthly visits to low-income housing complexes, local public libraries, community centers, and neighborhood schools. Welfare agencies should hold regular town/neighborhood meetings to provide technical assistance to people navigating the system and to let the public know what services are available. These efforts should provide tailored services to immigrant families irrespective of their levels of assimilation.

Recommendation 16: Ongoing caseworker training is essential to prevent discrimination and to ensure caseworkers understand eligibility requirements and the availability of services. Caseworkers should be held to the same high standards as other professionals, meet standards for competency, and engage in continuing education.

[1] The Urban Institute and the Survey Research Center, University of California at Los Angeles, How Are Immigrants Faring After Welfare Reform? Preliminary Evidence from Los Angeles and New York City, Final Report, Mar. 4, 2002 (hereafter cited as the Urban Institute, How Are Immigrants Faring After Welfare Reform?).

[2] Linda Burnham, “Welfare Reform, Family Hardship, and Women of Color,” The Annals of the American Academy of Political and Social Science, September 2001, p. 45 (hereafter cited as Burnham, “Welfare Reform, Family Hardship, and Women of Color”) citing Equal Rights Advocates, From War on Poverty to War on Welfare: The Impact of Welfare Reform on the Lives of Immigrant Women (San Francisco, CA: The Equal Rights Advocates, 1999).

[3] National Urban League, Institute for Opportunity and Equality, “Differences in TANF Support Service Utilization: Is there Adequate Monitoring to Ensure Program Quality?” June 2002, p. 7 (hereafter cited as National Urban League, “Differences in TANF Support Service Utilization”).

[4] National Urban League, “Differences in TANF Support Service Utilization.”

[5] Rebecca Gordon, Cruel and Usual: How Welfare “Reform” Punishes Poor People (Oakland, CA: Applied Research Center, 2001), p. 5 (hereafter cited as Gordon, Cruel and Usual).

[6] Gordon, Cruel and Usual, pp. 5, 33–34.

[7] Kenneth Finegold and Sarah Staveteig, “Race, Ethnicity, and Welfare Reform,” chapter 11 in Alan Weil and Kenneth Finegold, eds., Welfare Reform, The Next Act (Washington, DC: The Urban Institute Press, 2002), p. 215 (hereafter cited as Finegold and Staveteig, “Race, Ethnicity, and Welfare Reform”); Susan T. Gooden, Center for Public Administration and Policy, Virginia Tech University, “All Things Not Being Equal: Differences in Caseworker Support Toward Black and White Welfare Clients,” Harvard Journal of African American Public Policy, vol. 4 (1998), pp. 23–33.

[8] The Scholar Practitioner Program, African American Leadership Institute, Academy of Leadership, University of Maryland at College Park, “Racial and Ethnic Disparities in the Era of Devolution: A Persistent Challenge to Welfare Reform,” December 2001, p. 26 (hereafter cited as Scholar Practitioner Program, “Racial and Ethnic Disparities in the Era of Devolution”). Findings are based on research conducted under the W.K. Kellogg Foundation’s Devolution Initiative. The Scholar Practitioner Program has undertaken specific studies in five states: Florida, New York, Mississippi, Washington, and Wisconsin.

[9] Scholar Practitioner Program, “Racial and Ethnic Disparities in the Era of Devolution,” p. 23.

[10] Finegold and Staveteig, “Race, Ethnicity, and Welfare Reform,” p. 214, citing Joe Soss, Stanford F. Schram, Thomas V. Vartanian, and Erin O’Brien, “Setting the Terms of Relief: Explaining State Policy Choices in the Devolution Revolution,” American Journal of Political Science, vol. 45, no. 2 (2001), pp. 378–95. See also Scholar Practitioner Program, “Racial and Ethnic Disparities in the Era of Devolution,” p. 37.

[11] Finegold and Staveteig, “Race, Ethnicity, and Welfare Reform,” p. 207. The authors base their data on the Urban Institute’s 1997 and 1999 National Survey of America’s Families.

[12] Burnham, “Welfare Reform, Family Hardship, and Women of Color,” p. 45.

[13] Ibid.

[14] Finegold and Staveteig, “Race, Ethnicity, and Welfare Reform,” p. 213.

[15] Susan T. Gooden, “The Hidden Third Party: Welfare Recipients’ Experiences with Employers,” Journal of Public Management and Social Policy, vol. 5, no. 1 (1999), pp. 69–83 (hereafter cited as Gooden, “The Hidden Third Party”).

[16] See Michael Fix and Jeffrey S. Passel, “Assessing Welfare Reform’s Immigrant Provisions,” chapter 10 in Alan Weil and Kenneth Finegold, eds., Welfare Reform, The Next Act (Washington, DC: The Urban Institute Press, 2002), pp. 193–95.

[17] Michael Fix and Jeffrey Passel, “The Scope and Impact of Welfare Reform’s Immigrant Provisions,” the Urban Institute, discussion paper, January 2002. See also Michael Fix and Ron Haskins, “Welfare Benefits for Non-citizens,” the Brookings Institution, Policy Brief No. 15, February 2002.

[18] Scholar Practitioner Program, “Racial and Ethnic Disparities in the Era of Devolution,” p. 31.

[19] The Urban Institute, How Are Immigrants Faring After Welfare Reform? For this study, 3,447 immigrant families, including 7,843 people, were surveyed in New York City and Los Angeles County. The survey was conducted in five languages and had a response rate of 69 percent. Survey results were compared with data from the Current Population Survey of the U.S. Census Bureau and the Urban Institute’s National Survey of America’s Families.

[20] The Urban Institute, How Are Immigrants Faring After Welfare Reform?

[21] Ibid.

[22] Scholar Practitioner Program, “Racial and Ethnic Disparities in the Era of Devolution,” pp. 23–24.

[23] Burnham, “Welfare Reform, Family Hardship, and Women of Color,” pp. 45–46.

[24] Dallas Massey, Sr., chairman, White Mountain Apache Tribe, testimony before the Committee on Indian Affairs, United States Senate, May 10, 2002, p. 3.

[25] U.S. General Accounting Office, “Welfare Reform: Tribes Are Using TANF Flexibility to Establish Their Own Programs,” testimony before the Committee on Indian Affairs, U.S. Senate, May 10, 2002, p. 8. This includes only the 34 states with federally recognized Indian tribes.

[26] Ibid., p. 9.

[27] U.S. General Accounting Office, “Welfare Reform: Outcomes for TANF Recipients with Impairments,” July 2002, p. 13 (hereafter cited as GAO, “Outcomes for TANF Recipients with Impairments”).

[28] National Council on Disability, National Disability Policy: A Progress Report, November 1999–November 2000, <http://www.ncd.gov>, p. 49.

[29] Eileen P. Sweeney, Center on Budget and Policy Priorities, “Recent Studies Indicate That Many Parents Who Are Current or Former Welfare Recipients Have Disabilities and Other Medical Conditions,” Feb. 29, 2000, <http://www.cbpp. org/2-29-00wel.pdf>, pp. 1–5.

[30] Ibid.

[31] Denise F. Polit, Andrew S. London, and John M. Martinez, “The Health of Poor Urban Women: Findings from the Project on Devolution and Urban Change” (New York: Manpower Demonstration Research Corporation, 2001).

[32] GAO, “Outcomes for TANF Recipients with Impairments,” p. 7. Both findings hold after controlling for gender, race, age, marital status, education, and state of residency.

[33] Ibid., p. 21.

[34] Ibid., p. 9; U.S. General Accounting Office, “Welfare Reform: More Coordinated Effort Could Help States and Localities Move TANF Recipients with Impairments Toward Employment,” Oct. 31, 2002; U.S. General Accounting Office, “Welfare Reform: Moving Hard-to-Employ Recipients into the Workforce,” Mar. 15, 002.

[35] Anthony P. Carnevale and Katherine Reich, “A Piece of the Puzzle,” Educational Testing Service Leadership 2000 Series, <http://www.span-online.org/puzzle.pdf>.

[36] Ibid.

[37] National Urban League, Institute for Opportunity and Equality, “Negative Effects of TANF on College Enrollment,” Special Research Report, June 2002. The Urban League criticized the Census Bureau’s report on the work activities of mothers receiving TANF for not analyzing differences by race and ethnicity, particularly in the key areas of education and access to services. See National Urban League, “Census Report Spins Data, Urban League Contends,” press release, June 6, 2002.

[38] Center for Law and Social Policy and the National Council of State Directors of Adult Education, “Built to Last: Why Skills Matter for Long-Run Success in Welfare Reform,” May 2002, <www.clasp.org>, p. 6.

[39] Gooden, “The Hidden Third Party,” p. 81.

[40] Steve Savner, “Welfare Reform and Racial/Ethnic Minorities: The Questions to Ask,” Poverty and Race, vol. 9 no. 4 (July/August 2000), p. 3.

[41] Finegold and Staveteig, “Race, Ethnicity, and Welfare Reform,” p. 209. See also Scholar Practitioner Program, “Racial and Ethnic Disparities in the Era of Devolution,” p. 6.

[42] States that placed restrictions on or shortened the 60-month limit include Arizona, Arkansas, California, Connecticut, Delaware, Florida, Georgia, Idaho, Illinois, Indiana, Louisiana, Massachusetts, Nebraska, Nevada, North Carolina, Ohio, Oregon, South Carolina, Tennessee, Utah, and Virginia. See U.S. Department of Health and Human Services, Administration for Children and Families, “Time Provisions of State TANF Plans,” <http://www.acf.dhhs.gov/programs/ofa/TIME2.htm>.

[43] National Campaign for Jobs and Income Support, “TANF Reality Check: Time Limits,” <http://www.nationalcampaign.org/download.tlissuebrief.pdf>.