2
Welfare Reform Monitoring and Evaluation: The Current Landscape

This chapter describes the landscape of studies on welfare reform about four years after the passage of PRWORA. It includes both completed studies as well as a larger number that are under way or even in the planning stages. We provide short summaries of each study and describe its goals, evaluation methods, and the data sets used.

The types of studies under way are very diverse. We have found it useful to develop a classification of these studies into categories according to their type and nature—the questions asked, the methodologies used to achieve those goals, and the scope of inquiry. In our schema, studies can be classified as either; (i) descriptive and monitoring studies; (ii) studies of welfare leavers and related groups; (iii) randomized experiments; (iv) caseload and other econometric modeling efforts; (v) process, implementation, and qualitative studies; (vi) other welfare reform studies; and (vii) a variety of studies which do not focus on welfare reform per se but on related policy issues or low-income groups. Table 2–1 summarizes these types. Appendix Table B-1 contains a comprehensive list of the studies under way. In the next few sections we describe each of these types of studies. We withhold any critical evaluation of them until Chapter 4.

Our review of studies-and indeed, our entire report—is primarily focused on the Temporary Assistance for Needy Families (TANF) program. Studies of the Food Stamp, Medicaid, or related programs for the low-income population are not discussed in any detail. The focus on TANF is a partial result of PRWORA itself, which primarily reformed the Aid for Families with Dependent Children (AFDC) program and a partial result of the mission of our sponsoring agency, the Office of the Assistant Secretary for Planning and Evaluation (ASPE) in the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition 2 Welfare Reform Monitoring and Evaluation: The Current Landscape This chapter describes the landscape of studies on welfare reform about four years after the passage of PRWORA. It includes both completed studies as well as a larger number that are under way or even in the planning stages. We provide short summaries of each study and describe its goals, evaluation methods, and the data sets used. The types of studies under way are very diverse. We have found it useful to develop a classification of these studies into categories according to their type and nature—the questions asked, the methodologies used to achieve those goals, and the scope of inquiry. In our schema, studies can be classified as either; (i) descriptive and monitoring studies; (ii) studies of welfare leavers and related groups; (iii) randomized experiments; (iv) caseload and other econometric modeling efforts; (v) process, implementation, and qualitative studies; (vi) other welfare reform studies; and (vii) a variety of studies which do not focus on welfare reform per se but on related policy issues or low-income groups. Table 2–1 summarizes these types. Appendix Table B-1 contains a comprehensive list of the studies under way. In the next few sections we describe each of these types of studies. We withhold any critical evaluation of them until Chapter 4. Our review of studies-and indeed, our entire report—is primarily focused on the Temporary Assistance for Needy Families (TANF) program. Studies of the Food Stamp, Medicaid, or related programs for the low-income population are not discussed in any detail. The focus on TANF is a partial result of PRWORA itself, which primarily reformed the Aid for Families with Dependent Children (AFDC) program and a partial result of the mission of our sponsoring agency, the Office of the Assistant Secretary for Planning and Evaluation (ASPE) in the

OCR for page 24
Evaluating Welfare Reform in an Era of Transition TABLE 2–1 Types of Welfare Reform Projects Currently Under Way Type of Study Description Monitoring and Descriptive Studies that examine adults and children by following families over time, documenting trends in well-being Welfare Leavers and Related Groups Studies that document the outcomes for individuals and families who have left welfare Randomized Experiments Evaluations using randomized experimental and control groups to estimate impact of a specific reform program or feature Caseload and Other Econometric Models Analyses using econometric methods to estimate the effects of welfare reform on caseloads and other outcomes Process, Implementation, and Qualitative Studies Studies using qualitative methods to examine and document implementation of welfare reform, state program rules, or detailed pictures of individuals and families Other Welfare Reform Studies Studies of special populations, the child welfare system, and data collection projects Studies on Topics Related to Welfare Reform Studies of child support enforcement, absent fathers, low-income neighborhoods, low-income children, and other topics Department of Health and Human Services. However, virtually all of our findings on methods and data extend to evaluation efforts for food stamps and Medicaid as well. Before describing the studies, we briefly note who the major supporting funders are and how the types of funders of current welfare reform research differ from those in past eras. Major Supporting Funders The most active federal governmental agencies supporting research on welfare reform are ASPE and the Administration for Children and Families (ACF), both in the U.S. Department of Health and Human Services (DHHS). Both agencies have extensive and detailed research agendas on welfare reform, ranging from broad overview studies to data collection activities to evaluation efforts of specific programs and population groups. They are supporting dozens of research and evaluation activities around the country, both private-sector researchers and state government agencies conducting studies of their own state programs. The breadth of their activities is considerable; Appendix Tables A-1 and A-2 show the research projects currently supported by or being carried out internally at ASPE and ACF.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition However, unlike some past periods of reform evaluation, many other federal agencies are actively involved as well. These include the Food and Nutrition Service and Economic Research Service of the U.S. Department of Agriculture; the U.S. Department of Labor; the Office of Research and Statistics in the Social Security Administration; various health and housing agencies; and several other agencies. The U.S. Census Bureau is also heavily involved in welfare reform because of its responsibility for the Survey of Program Dynamics, which is specifically aimed at providing information for welfare reform evaluation (and is discussed later in this report). The National Institutes of Health is also newly engaged in welfare policy evaluation and is sponsoring several large-budget studies, most of which focus on children and families. State governments are also now heavily involved in sponsoring welfare reform studies, which is a natural result of the devolution of authority over the TANF program to the states. Almost every state in the country has some type of evaluation under way, sometimes conducted in-house and sometimes contracted out. Many of these projects are solely funded by the state; others are partially funded by the federal government or by private foundations. Even newer to the current evaluation funding environment is the active involvement of foundations in support of welfare reform studies. Many large national foundations, as well as many smaller foundations with interests in specific cities or states, have devoted funds to a wide variety of studies. Although there are no figures on the total amount of these foundation funds, the amounts are unquestionably large and often rival those of federal government agencies. DESCRIPTIVE AND MONITORING STUDIES As we noted in our interim report (National Research Council, 1999), descriptive and monitoring studies play an important role in studies of social policies in general and have assumed an important role in the study of welfare reform as well. A monitoring study follows a population of families or individuals over time and tracks their well-being, as measured by economic status and other indicators. However, monitoring studies do not attempt to make formal assessments of the effects of a reform relative to what would have happened in its absence. Monitoring studies are important because they signal whether the well-being of the target population is improving, deteriorating, or remaining the same. They are also useful in identifying specific subgroups that are doing particularly poorly, and may therefore, be in need of additional assistance, regardless of what might have caused that condition. Some “monitoring” studies track families before and after a reform and attempt to make an assessment of the effect of the reform by comparing outcomes before and after. If such an assessment is made, the study is then actually an evaluation study. Tracking families can be valuable from a monitoring point of view, but it is a weak evaluation tool because such before-and-after evaluations

OCR for page 24
Evaluating Welfare Reform in an Era of Transition cannot generally separate the effects of welfare reform from other simultaneously occurring events, such as the improvement in the economy. Although most monitoring studies have focused on individual states and local areas, a few have a national focus. The latter group describes the course of the disadvantaged population over time for the nation as a whole or for major portions of it, sometimes both before and after PRWORA. National-level studies depend, necessarily, on the availability of data that are representative of the country as a whole and so are most easily classified by the data set used. Some studies, for example, have used multiple waves of the March Current Population Survey (CPS) (e.g., Primus et al., 1999) to track the well-being of the low-income population. The Survey of Program Dynamics (SPD) is expected to be used for monitoring as well, but has thus far not been used very much for this purpose. Other national-level studies include baseline and continuing reports on welfare dependency, reports on the development of child indicators by states, general reports on trends in the well-being of children and youth, reports on trends in the economic well-being of low-income Americans, and reports on poverty dynamics and contingent employment. These reports typically also use the CPS, though they are sometimes supplemented by SIPP or other data sets as well. There are three major monitoring studies at a somewhat more localized level: the National Survey of America’s Families (NSAF), the Project on Devolution and Urban Change, and the Three-City Study. The NSAF consists, to date, of two waves of a telephone survey of almost 50,000 families in 13 states plus a small sample of families in the rest of the country. Low-income families and children are oversampled. The families in the two waves are not the same, so the survey essentially gives snapshots of the population at two points in time, in mid-1997 and in 1999. Both survey waves took place after the passage of PRWORA; NSAF thus monitors the population subsequent but not prior to PRWORA. Columbia University is also conducting a study of welfare reform in New York City based on a repeated survey of a population cross-section, the Social Indicators Survey, that closely parallels the NSAF. The Urban Change Study and the Three-City Study involve four urban counties and three cities respectively. Both studies have longitudinal, in-person surveys as a central element, and both have an ethnographic component as well; the Urban Change Study also has an implementation component and a neighborhood indicators component. In the survey components, information is collected on both adult and child well-being, although the Three-City Study has perhaps the most intensive focus on children and includes a special supplemental survey and assessment of a subsample of parents and children. The first wave of the Urban Change Study survey was fielded in 1998, after PRWORA; the first wave of the Three-City Study survey was fielded in 1999, also after PRWORA. The two surveys differ in their sampled populations: the Urban Change Study is drawn from families on TANF at several times, both before and after PRWORA, and both survey and administrative data are being collected on these families; the

OCR for page 24
Evaluating Welfare Reform in an Era of Transition Three City Study drew its sample from the general population of low-income families in the three cities, both those on and off TANF, and is based only on survey data. There are also a large number of monitoring and descriptive studies that are narrower in their geographic and substantive focus than those we have mentioned, usually focused on a particular state or area and sometimes on only some groups of interest. Typically, these studies use administrative data to track certain groups of the low-income or recipient population. These studies, many of which are listed on the web site of the Research Forum on Children, Families, and the New Federalism, are generally more limited in scope (see http://www.researchforum.org). STUDIES OF WELFARE LEAVERS AND RELATED GROUPS There have been several dozen state-level studies of women who have left welfare subsequent to PRWORA. Indeed, studies of these welfare leavers constitute by far the most common type of welfare reform study conducted since 1996. Most of these studies are specific to a particular state or area within a state, and most to date have only measured employment and earnings outcomes of leavers from administrative data; a few, especially ASPE-sponsored studies, have conducted short telephone or in-person interviews as well.1 The goal of these studies is, simply, to track the well-being of women who have left welfare. Existing leaver studies are implicitly designed as monitoring studies, for they do not attempt to determine what the outcomes of leavers would have been if welfare reform had not occurred—or even if those families would have left welfare. They do not assess the effects of the reforms nor explain whether or how much the improving economy has led to exits from welfare. They cannot give a full picture of the effects of policy changes because they focus only on those receiving welfare at a given time, not the entire population that might be affected. Nevertheless, like the more general monitoring studies discussed in the previous section, leaver studies can be very useful in tracking and documenting the well-being of one particular subpopulation—welfare leavers—and determining how well they are doing, even if the cause of those outcomes is not known. Like other monitoring studies, leaver studies can also be useful in identifying specific subgroups of those who have left welfare who have not done well or who have particular special needs that require additional assistance. As we noted in our interim report (National Research Council, 1999), many of the early leaver studies were scientifically low in quality because of data 1   For a review of data, methods, and findings of state welfare leaver studies, see Parrott (1998), Tweedie and Reichert (1998), Brauner and Loprest (1999), Cancian et al. (1999), Tweedie et al. (1999), U.S. General Accounting Office (1999a), Isaacs and Lyon (2000), and Acs and Loprest (2001).

OCR for page 24
Evaluating Welfare Reform in an Era of Transition constraints. Many did not collect survey data, and of those that did, nonresponse rates were often very high. More commonly, leaver studies used administrative data and usually only from a narrow set of sources; consequently, they had only limited information on well-being after leaving welfare, often only employment and earnings from records of the unemployment insurance system. Also, different states used different definitions of leavers and examined different groups, so that the results of the studies are not comparable across states. In fiscal 1998, ASPE funded 14 states and areas to conduct new leaver studies. These studies use a set of definitions that are uniform in some respects and are therefore more likely to be comparable across states. They also collect survey and administrative data on a wider variety of adult and child outcomes (see National Research Council, 1999:Tables A-1, A-2). Most of these ASPE-sponsored studies were designed to examine leaver outcomes for two cohorts of recipients as well—a pre-PRWORA, AFDC cohort and a post-PRWORA, TANF cohort. These two cohorts were to be compared to assess whether rates of leaving and leaver outcomes were different after PRWORA than before, thereby addressing the issue of whether leaver outcomes have indeed changed over time. In fiscal 1999, ASPE funded seven more states to expand significantly on the scope of the leaver studies to examine applicants as well as leavers, diverted applicants as well as other applicants, and eligible nonapplicants. These projects are based on the recognition that welfare reform can affect the rate of entry to welfare as well as the rate of exit, as families who might otherwise have opted to go on welfare chose to attempt self-sufficiency instead. In addition, diversion programs, or programs that seek to dissuade potential recipients or applicants from enrolling in the cash assistance program, directly affect the entry rate. Unfortunately, at this writing very few multiple cohort studies or studies of entry have been completed.2 ASPE has also commissioned a review and synthesis of its leaver studies, which should be completed soon. While most studies of leavers have been conducted at the state level, a few have used national-level data. These studies use cross-sectional surveys, such as the Urban Institute’s new National Survey of America’s Families (Loprest, 1999); the National Longitudinal Survey of Youth (Cancian and Meyer, 2000; Meyer 2   Multiple cohort studies have been completed for Arizona, Illinois, South Carolina, Washington, and Wisconsin (Cancian et al., 2000; Institute for Public Affairs and School of Social Work, 2000; Isaacs and Lyon, 2000; South Carolina Department of Social Services, 2000; Ahn et al., 2000). All of the cohorts in these studies were post-PRWORA except for the Wisconsin study, which included a 1995 cohort (when Wisconsin was operating under waiver authority in 1995) and Washington. No state-level entry-rate or diversion studies have been completed except for a recent study (Mueser et al., 2000), which examined the employment rates of entrants to AFDC-TANF in five urban areas from 1990 to 1997, controlling for business-cycle effects. A Wisconsin study of applicants to TANF in Milwaukee is under way but has not been completed.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition and Cancian, 1998); and the Panel Study of Income Dynamics (Hofferth et al., 2000).3 There are also a few studies that examine “stayers,” or women who remain on welfare, although these studies are not as prevalent as leaver studies. Some have compared characteristics of leavers to stayers (Cancian et al., 2000; Fogarty and Kraley, 2000; Institute for Public Affairs and School of Social Work, 2000; Loprest and Zedlewski, 1999); others have focused on the characteristics of stayers per se, particularly whether they have barriers or obstacles to employment (Danziger et al., 2000; Zedlewski, 1999).4 There are also a number of projects that are tracking the characteristics of welfare recipients over time with administrative data (e.g., the Illinois Panel Study of Recipients, and a recently funded ASPE project on the California caseload). RANDOMIZED EXPERIMENTS Randomized experiments have been a major evaluation method in welfare reform since the 1980s when a number of small-scale experiments of state-level welfare innovations were tested. A larger number of experiments were begun in the early 1990s as states were awarded waivers from the federal government to conduct experiments of various alterations of their AFDC programs, alterations that were related to those later enacted in PRWORA. However, most of these waiver experiments were discontinued after PRWORA. The Administration for Children and Families has continued to fund nine of the waiver experiments in those states that chose to continue to operate their waiver programs (rather than convert immediately to TANF); five of them are funded to measure child outcomes in the experimental and control groups as well. These continuing waiver experiments constitute the bulk of experiments that attempt to directly test packages of reforms that resemble those adopted after PRWORA. A number of experiments that study different aspects of welfare reform are also under way, have been completed, or are in the planning stages. The Post-Employment Services Demonstration was an experiment testing alternative strategies for increasing the rate at which welfare recipients keep jobs once they have obtained them. The Employment Retention and Advancement Project, sponsored by ACF, followed up on this demonstration with a large and more comprehensive experimental test of alternative strategies to assist welfare recipients and welfare leavers in retaining jobs. The evaluation of the Labor Department’s 3   Either panel data or a cross-sectional database with a retrospective history for several years is needed to conduct a leaver study, so the Current Population Survey, for example, cannot be used for this purpose. 4   The Danziger et al. (2000) study includes both leavers and stayers combined and pooled into one sample.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition Welfare to Work Program contains experimental components in some sites and tests the effect of the Welfare to Work Program legislation. The Los Angeles Jobs First-GAIN experiment is also testing work-first versus education-based assistance strategies. The Wisconsin Works Child Support Waiver Demonstration tests the effects of Wisconsin’s changes to the child support system and rules in the state. The National Evaluation of Welfare-to-Work Strategies (NEWWS) is an experiment in a different category, for it began in 1989 as an evaluation of the JOBS component of the Family Support Act of 1988 and hence predates PRWORA and the waiver reforms of the early 1990s. It has been continued, however, because it does test alternative employment strategies—namely, a work-first strategy versus an education-training strategy—and this issue is still relevant to welfare reform after PRWORA. Finally, a number of states are also testing small-scale program features with experimental methods. There are a number of other ongoing or completed experimental evaluations of welfare-related programs that test particular types of welfare reforms other than those enacted by PRWORA. The evaluation of the pilot phase of the Minnesota Family Investment Program, for example, tested a package of reforms containing more generous earnings disregards and some new work requirements.5 The New Hope evaluation and the Canadian Self-Sufficiency Program evaluation both test programs that have a few of the features of post-PRWORA programs, but both differ from welfare reform policies significantly in many important respects and are generally quite different in philosophy, particularly in their heavy emphasis on earnings subsidies. The results from these experiments will be useful if the particular reforms they test arise in future policy discussions in Congress, but they are not directly relevant to the effects of PRWORA. CASELOAD AND OTHER ECONOMETRIC MODELS There have been a number of econometric evaluations of the effects of pre-PRWORA waiver programs and a few of the effects of PRWORA itself. Most of these evaluations have aimed to estimate the effect of welfare reform on welfare caseloads, while a few have examined the effects on earnings, income and other measures of well-being of the low-income population. Most have been nationwide in scope, although in a few cases only a single or small number of states have been examined (e.g., Hill and Main, 1998; Mueser et al., 2000). The majority of pre-PRWORA waiver evaluations used variation across states in the timing of when waiver programs were enacted in each state, as well as the type of program that was adopted, to explain subsequent rates of caseload decline (Council of Economic Advisers, 1997; Figlio and Ziliak, 1999; Moffitt, 1999; Wallace 5   Interestingly, the later post-PRWORA package adopted in Minnesota was quite different.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition and Blank, 1999; Ziliak et al., 1997). The small number of studies that have used post-PRWORA data (Council of Economic Advisers, 1999; Ellwood, 2000; Schoeni and Blank, 2000) have estimated effects of PRWORA either by comparing states that implemented PRWORA later than other states; by simply comparing state outcomes after PRWORA to state outcomes before (i.e., a before-and-after, or pure time-series design); or by comparing trend changes in the outcomes of single mothers before and after PRWORA to trend changes in outcomes for other groups (married women, men, etc.) that presumably were not affected by PRWORA. Most of the studies examining the effect of welfare reform on caseloads have used state-level aggregated caseload totals as their primary data base, while those examining individual and family outcomes have used survey data, most often the Current Population Survey (CPS), which is available annually for many years prior to PRWORA. Probably the most notable feature of the econometric evaluations of welfare reform is that they have attempted to control for the state of the economy—most often using the unemployment rate in a state as a proxy—on caseload and other outcomes in order to estimate the net effect of welfare reform. Because the unemployment rate declined at the same time welfare reform occurred, disentangling their relative influences has been a major challenge. The econometric evaluations are the only type of study that has attempted to estimate these relative influences. They have a number of other strengths as well, though also some weaknesses, which we discuss in Chapter 4. ACF has commissioned a project to synthesize the results of the econometric evaluations. A number of the monitoring and leaver studies already mentioned have some type of implicit econometric evaluation components as well. The most common type of evaluation design is what we termed a cohort comparison study in our interim report (National Research Council, 1999): the Urban Change Study, the Three-City Study, and the NSAF all have plans for this type of analysis, as do some of the leaver studies. In addition, these and other studies will be developing comparison groups within their designs to estimate some effects of welfare reform, but these plans have not been developed to date. Indeed, any study that seeks to estimate the effects of policy alternatives with nonexperimental methods is, in a sense, conducting an econometric study if that term is defined broadly. For example, the Rural Welfare Reform Project sponsored by ACF aims to gauge the effects of welfare reform strategies in rural areas with nonexperimental means. A number of other projects sponsored by ACF in individual states are doing the same. ASPE has funded a number of relatively small research grants that use nonexperimental methods to evaluate the effect of welfare reform. Many nonexperimental studies using various observational designs are funded primarily at the state or local level, such as a large evaluation in California of the CALWORKS program, an evaluation of welfare reform in Los Angeles County, and the large New York State comprehensive welfare reform evaluation.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition PROCESS, IMPLEMENTATION, AND QUALITATIVE STUDIES A major study of implementation of welfare reform in 20 states, the State Capacity Study, is currently being conducted at the Rockefeller Institute of Government, for example. The project aims to describe changing state management systems for social service programs, to describe their goals, and to make recommendations on different management strategies. The Urban Institute, as part of its Assessing the New Federalism Project, is studying implementation and policy choices in the same 13 states that are covered in NSAF, described above. The Urban Change Study has an implementation component that involves studying the implementation of welfare reform in the same four areas where the individual-level data are being collected. In addition, virtually all the waiver experiments have a process analysis component, as have virtually all large-scale experiments conducted in the last 10 years. There are also a large number of implementation studies that have been, and continue to be, conducted in individual states and local areas examining how PRWORA or various welfare services are being provided in the era after PRWORA; they are funded either by state and local governments or by the federal government. Process and implementation analysis, therefore, is thriving in the current welfare reform research scene. Efforts to document state program rules in the aftermath of PRWORA are also under way. Unlike the situation under AFDC, a federally regulated program, all the precise rules that states have adopted no longer need to be reported to the federal government. The Urban Institute, as part of its Assessing the New Federalism initiative, has developed a Welfare Rules Database that contains descriptions of each state’s (and the District of Columbia’s) TANF rules from 1996 to 1999 (the 1996 data include AFDC rules). The 2000 update (funded by ACF) is now being produced. Current plans call for further updates in 2001 and 2002, also funded by ACF. The Center for Law and Social Policy (CLASP) and the Center for Budget and Policy Priorities (CBPP) are jointly collecting information on state TANF and Medicaid policies as part of the State Policy Documentation Project. This information was collected for legislation enacted before and updated through 1999. A joint effort between the U.S. Department of Health and Human Services (DHHS), the Welfare Information Network (WIN), the American Public Human Services Association, and the National Governors Association is summarizing state plans provided to DHHS as part of the final reporting rules of April 1999 (Federal Register, 1999) of PRWORA. These summaries will be entered into a database and made public on the Internet. This information is also being supplemented with information collected through the Urban Institute’s Welfare Rules Database and the State Policy Documentation Project of CLASP and CBPP. The database will be updated every time new state plans are submitted. Finally, the Congressional Research Service has also used state TANF plans reported to DHHS and ACF to produce reports summarizing state program rules on different topics.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition Finally, there have been a number of qualitative studies of welfare or nonwelfare poor families. Most of these studies collect participant-observation or related types of in-depth information on families through an intensive personal contact process, although some conduct shorter unstructured interviews with families. The Urban Change Study conducted an ethnographic investigation of approximately 40 welfare-reliant families in each of its four areas to explore how families coped with the new rules and policies (Quint et al., 1999). The Three-City Study has a similarly large ethnographic project in its three cities to portray welfare reform from the point of view of the affected families. Other ethnographic studies have been a part of the New Hope Project, the Iowa Benefit Plan Project, the Wisconsin Child Support Demonstration Evaluation, and many others. The presence of a fairly large number of ethnographic and qualitative studies is a feature of welfare reform research in the 1990s that was rarely present in earlier periods of research. OTHER WELFARE REFORM STUDIES A variety of other ongoing welfare reform studies do not fall neatly into any of the above categories. One large group of studies of specific populations examine welfare recipients or leavers with problems of substance abuse, domestic violence, mental or physical health, or English-language difficulties. Others focus on specific populations, such as Native Americans or immigrants. Sometimes these studies are purely monitoring in nature, and sometimes they involve an assessment of the effects of particular new policies that apply to the group in question. A second group of studies consists primarily of data collection projects, most often collection of administrative data. These include a national survey of women and children in the child welfare system, as well as data assembly projects at UC-Data in California, a six-city consortium led by the University of Baltimore, and a multiple-state study using confidential data on business firms collected by the Census Bureau. ASPE is also funding a variety of studies to support data collection, including projects to assist states in establishing administrative and survey data bases on welfare recipients and the low-income population, to match federal data bases on employees and welfare recipients, and to use Social Security earnings records in the evaluation of welfare reform. STUDIES ON TOPICS RELATED TO WELFARE REFORM There are many other studies that do not bear directly on reform of the TANF or related welfare programs but rather on other programs that serve some of the same populations. We do not have the space to detail all of them, but we do mention a few of the major ones here. ASPE is funding studies on child support enforcement, child care, abstinence education programs, and child welfare, for example, all of which serve groups heavily overlapping with TANF population.

OCR for page 24
Evaluating Welfare Reform in an Era of Transition ASPE and ACF are jointly sponsoring a project to encourage the measurement of child outcomes in state-level studies, particularly those involving waiver experiments. ASPE is also sponsoring studies of specific groups such as, child-only cases, those with disabilities, and victims of domestic violence. ACF has funded state evaluations of employment retention initiatives and rural welfare reform strategies. Both the Urban Change and Los Angeles Survey of Families and Communities studies have components that study neighborhoods in low-income communities in the aftermath of welfare reform. Yet another ongoing project related to welfare reform is the Fragile Families Study, which is studying the relationship between fathers and mothers of unwed children, as well as the children of such relationships, in 20 cities.