Table of Contents | Previous | Next |
I. INTRODUCTION
The devolution of responsibility for the nation’s primary welfare program from the federal government to state and local agencies was one of the most important institutional changes in federal welfare policy during the 1990s. The Personal Responsibility and Work Reconciliation Act (PRWORA) of 1996 created the Temporary Assistance for Needy Families (TANF) block grant, which gave states considerable latitude to allocate funds across program functions and move functions outside of traditional welfare agencies to other public agencies and private organizations.
Since PRWORA was passed and implemented, cash assistance caseloads have declined by more than half. Policymakers and researchers continue to deliberate about the causes of this decline (e.g., federal and state policies versus a favorable labor market). However, most agree that federal and state welfare policies explain part of the decline, while debating the magnitude. The interaction between policy, local program operations, and program outcomes is not fully understood. Welfare agency structures and management processes express local interpretations of policies and shape critical aspects of their delivery. In some local offices, program staff have considerable discretion in program delivery. Office managers, for example, may make decisions about agency structure and staffing that affect how new policy directives will be incorporated into existing routines and how staff will be deployed. Front-line staff may determine how to manage day-to-day work demands.
PRWORA was not the first reform to affect state and local welfare offices. Rather, welfare reform in 1996 was part of a long series of state and federal changes in cash assistance and other programs. These changes imposed a constant demand on local welfare offices to adapt to new policy initiatives, many of which greatly expanded the responsibilities of state and local welfare systems that were initially established to provide a very specific service—cash assistance to children and their parents.
A. Study Purpose
The Administration for Children and Families (ACF) funded this study to determine how local management of TANF programs has evolved to address changing needs and improve program results. Management efforts to adapt to changing conditions are an understudied aspect of local welfare operations. Rather than viewing TANF implementation as a single event followed by a stable administrative structure and a static array of services and requirements, the project viewed implementation in dynamic terms. It sought to understand how local welfare offices modify their programs and adopt more effective approaches in response to changing circumstances.
The study collected three types of information from five selected sites:
-
An update on the implementation of TANF at the local level (i.e., a description of the practices and operations of local welfare offices in early 2006).
-
A listing and description of innovative practices put into effect after the first few years of TANF implementation, particularly those that were adopted by local managers in order to improve program performance.
-
A more general analysis of the evolution of practices in local welfare offices. For example, how do current practices differ from those found in earlier years of TANF implementation? And to what extent have the changes in local practices been a result of adaptations to local circumstances and learning from local experiences, or a consequence of other factors?
B. Methodology
The project team, composed of senior staff from The Lewin Group and The Nelson A. Rockefeller Institute of Government, gathered information for the study through site visits to five local TANF offices. One issue in using local areas as the unit of analysis is that they are not comparable to one another. One urban or suburban county may include many local and separately administered TANF offices, each with distinct service contractors, while a rural county may be served by only one such office. In order to ensure greater comparability, we defined local offices to include, at minimum, one welfare office (that determines eligibility for assistance cases), as well the offices of the employment and child care agencies serving the welfare clients.
1. Site Selection Criteria
Site selection was guided by three criteria:
-
The availability of pre-existing documentation (i.e., included in a previous Rockefeller Institute or Lewin Group study).
-
Sites that experienced considerable changes relevant to the effectiveness of welfare programs (e.g., economic, demographic, or state policies).
-
Variation in TANF programs (e.g., resource levels, assistance versus non-assistance funding, state versus county administered).
To select five primary and five alternate sites, the team first examined all local sites included in three Rockefeller Institute projects (the field network studies of 1997-98 and 2001-02 and the 1999-2000 frontline workers and management study). All localities with major urban areas were considered. Twenty-one counties satisfied this criterion. Next, data on these 21 counties were reviewed. The localities were then compared and classified according to assessments of the quality of the reports, local focus, and time covered by the studies. Sites with good documentation of older practices were then assessed for economic changes (e.g., increases in unemployment), demographic or social changes (e.g., rapid population changes, increases in immigration) and state policy and fiscal changes. The team then examined each state’s investment in the program relative to need (total spending per poor child); cash benefit levels and eligibility thresholds; and policies regarding time limits, sanctions, and job search before qualifying for assistance.
After 10 sites were selected that represented urban areas for which good data were available, that experienced important changes (mostly increases in demands), and that showed variation in TANF policies and expenditure shifts, the field reports were reviewed for other information about these sites and states. All of this information was considered in determining which sites were recommended for study and which would be suggested as alternates.
2. Local Offices Selected
The five sites ultimately selected were in Phoenix, Arizona; Macon, Georgia; Kansas City, Missouri; Newark, New Jersey; and Milwaukee, Wisconsin. In each city, the team visited one local office. In some sites—Milwaukee, Kansas City, and Phoenix—the office visited was one of a number of offices that administer the TANF program in that city. The findings described in this report relate to the local office visited. For short-hand, the city in which the office is located is used when referring to the local office included in the study.
The sites selected are geographically diverse. Two states are in the Midwest; one each is in the South, the Southwest, and the Northeast. The sites also represent a range of economic and demographic profiles. Exhibit I.1 provides a snapshot of these characteristics in 2005 and compares them to the U.S. average.
| Characteristics of Population | Phoenix | Macon* | Kansas City | Newark | Milwaukee | U.S. |
|---|---|---|---|---|---|---|
| Population (in thousands of persons) | 1,378.0 | 89.8 | 440.9 | 254.2 | 556.9 | 296,410.4 |
| Percent White | 74% | 35% | 61% | 22% | 45% | 75% |
| Percent Black or African American | 5% | 62% | 30% | 53% | 40% | 12% |
| Percent Hispanic (any race) | 42% | 1% | 8% | 33% | 14% | 14% |
| Percent Foreign-born | 23% | 1% | 7% | 30% | 9% | 12% |
| Percent High School Graduates | 79% | 77% | 86% | 65% | 80% | 84% |
| Percent Families Below Poverty Line | 13% | 32% | 13% | 23% | 21% | 10% |
| Per Capita Income | $22,471 | $15,578 | $24,567 | $15,346 | $17,696 | $25,035 |
| Unemployment Rate** | 4.1% | 5.7% | 5.6% | 4.9% | 5.0% | 5.1% |
|
Source: U.S. Bureau of the Census, American Community Survey * Race and ethnicity characteristics for Macon are for 2000 and compiled from the census. ** Unemployment rate is an annual average |
As the Exhibit shows, the proportion of population that is white ranged from 22 percent in Newark to 74 percent in Phoenix; the proportion of African Americans ranged from 5 percent to 62 percent (Phoenix and Macon, respectively). Macon had the lowest proportion of foreign-born residents (1 percent), while Newark had the highest (30 percent). The percent of families with incomes below the poverty line ranged from a low of 13 percent (Kansas City and Phoenix) to a high of 32 percent (Macon). All five cities had lower per capita incomes than the United States as a whole.
States also differ in terms of devolution for TANF program decision-making and resources allocated to the program. Three states—Arizona, Georgia and Missouri—have state-administered programs. In these states, the research team expected that the policies, if not the procedures, would be relatively uniform in the local office visited and other offices around the state. New Jersey and Wisconsin have county-administered programs. The local offices in these sites were thought to have more autonomy in developing TANF policies and processes.
As Exhibit I.2 shows, the five states varied in terms of investment in their TANF programs1 relative to need, as measured by average annual spending per poor child from 2002 to 2004. New Jersey and Wisconsin spent more than the median state on its TANF and Maintenance of Effort (MOE) programs.
| State | Average Annual Spending per Poor Child | State Ranking |
|---|---|---|
| New Jersey | $4,320 | 4 |
| Wisconsin | $2,440 | 17 |
| Missouri | $1,395 | 30 |
| Georgia | $1,257 | 33 |
| Arizona | $1,062 | 37 |
Some of the sites initially selected declined to participate in the study. The team’s first choice for a local site in Georgia was Fulton County (Atlanta). However, due to a number of ongoing policy and programmatic issues unrelated to TANF (primarily associated with the child welfare program), the acting director of the state TANF program requested that the team visit a different site; as a consequence, the site in Macon was chosen.
The team also initially selected a local office in Dallas, Texas. The Deputy Director of the Texas Health and Human Services Commission (HHSC), which oversees eligibility functions, was hesitant about including Dallas in the study because the state was in the process of moving to a call center model and revising other procedures for TANF applications, none of which had been rolled out in Dallas. The HHSC Deputy Director recommended Travis County (Austin), where the new configuration was in place. However, in order to visit the local office responsible for employment and training functions for TANF clients in Austin, it was necessary to secure the agreement of the Texas Workforce Commission (TWC). Staff there indicated that they would not agree to a site visit in Austin because of unresolved issues that had arisen when the revised procedures were implemented. Although the TWC offered other potential sites, they were not in locations with call centers, and were thus not approved by HHSC. As a result, the team was unable to include a local Texas office in the study.
3. Site Visit Discussions
To ensure that the research team collected information that addressed the goals of the study and could be synthesized across major themes for the project final report, the team drafted topic guides. These guided the discussions and were tailored to stakeholders (see Box 1).
Box 1: TANF Stakeholders
|
The topic guides explored a number of themes. Depending on the stakeholder, they focused on:
- Current administrative practices and operations
- Level of local autonomy in setting goals, assigning resources
- Level of coordination among service providers
- Information technology infrastructure and support
- Program, administrative, and resource changes in recent years
- Changes in clients, including demographics, work experience, and problems faced
- Innovations adopted to improve performance
- Changes in data reporting and how the data are used
- Anticipated changes
The site visits were conducted between February and July 2006. Because the guides were drafted and the initial visits conducted prior to passage of the Deceit Reduction Act of 2006, the conversations did not focus on planned changes associated with TANF reauthorization.
C. Organization of This Report
Section II of the report compares state policies of the five study sites, as well as the current program structures and processes.
Section III of the report describes the findings of the study by analyzing the changes in the TANF program at the local sites during the previous five years.
Section IV draws summary conclusions.
| Table of Contents | Previous | Next |