This study was undertaken to offer a preliminary assessment of how housing authorities, their residents, and the Federal government will be affected by welfare reform and to describe the local contexts in which it will unfold. The housing authorities studied here include the Richmond Redevelopment and Housing Authority (RRHA) and the Norfolk Redevelopment and Housing Authority (NRHA) in Virginia; the Cuyahoga County (Cleveland), Columbus, and Lucas County (Toledo) Metropolitan Housing Authorities in Ohio; The City of Los Angeles and San Francisco Housing Authorities in California; and, the Dallas Housing Authority in Texas.

The two housing authorities in Virginia, RRHA and NRHA, are examined here in particular detail because they were the object of a broadly based data collection effort which included field visits by staff of HUD's Policy Studies Division. However, to provide a basis for comparison, as well as some geographic scope for estimating the impact of welfare reform, the study also introduces information on the six other authorities in three states, made possible through telephone conversations and file transfers. Because there is value in comparing different housing authorities within the same state, an unsuccessful attempt was also made to gather information from another Texas housing authority.

The descriptions provided in this report cover a range of features including the size of housing authority populations subject to welfare reform requirements and the demographic characteristics of these residents relevant to their ability to find entry-level jobs. It also includes: the rent revenues at housing authorities (especially the component of revenue likely to be affected by reform); the rent and tenant preference policies of housing authorities; the number and characteristics of those with whom public housing residents must compete for jobs; the existence of entry-level jobs and their location; the wage rates associated with entry-level jobs; and the support networks in place to facilitate work, including child care, transportation, and job readiness preparation.

A. Issues Involving Job Participation

The key to determining the impacts of welfare reform in this study is determining the changes in household income as a result of moving from welfare assistance to possible employment. But, public housing residents required to work have special challenges because they usually did not choose where to live and, therefore, did not account for such practical work-related considerations as proximity to transportation, employment centers, and day care.

Consequently, when considering the likelihood of obtaining a job, their situation is fundamentally different from households against whom they will compete for entry-level jobs, including many with portable subsidies, who have more housing location options. Lack of housing choice, however, might not pose a problem for public housing residents if there were an adequate supply of jobs within reach. But, many jobs in the metropolitan area are not in reasonable commuting distance and hence, are basically inaccessible. And, even though jobs may be accessible, to the extent that public housing residents live within the same commuting zones as other competing for the same pool of jobs, greater competition likely means that fewer jobs would be actually available.

The problems of job access are exacerbated by other obstacles facing public housing residents when searching for jobs. These include insufficient education and skills to be competitive in the job market, lack of affordable child care, and others. Regardless, TANF and alternative State plans under TANF contain time limits and make assistance to residents conditional on work participation.

B. Sources of Information

Some of the information and data required for this study are available for all of the housing authorities included here, but the support network information is available only for Norfolk and Richmond because on-site visits were made there. Labor market information is available only at the metropolitan level for the Virginia, California and Texas study HAs; it is provided at the neighborhood level for the three Ohio housing authorities. The latter were the focus of cooperative agreements that the Office of Policy Development entered into with Case Western Reserve University in Cleveland and the University of Wisconsin-Milwaukee.

A key role in the study was played by participating housing authorities which provided current data covering not only heads of households, but also every family member in residence. In addition, several state agencies responsible for administering the new TANF legislation augmented housing authority information not only for welfare recipients in public housing but also for the remainder of their caseloads in the study sites of interest. This was accomplished through a matching process that permitted the state agencies to identify which of their records related to public housing residents and which did not. The results were the assembling of a richer dataset for all welfare recipients than would have been possible with any other methodology, and the identification of the major population groups that public housing residents would be competing with for jobs as a result of welfare reform.

Information on current and future HA rent revenues and the ability of housing authorities to bolster tenant rent contributions through preference policies, minimum rents, etc., are major components of the assessment of the impacts of welfare reform. The study also draws on information about the work participation patterns of households who do not receive welfare but otherwise resemble public housing residents. Because welfare assistance will soon no longer be available to public housing residents, the work participation levels of unassisted households who resemble them are used as the starting points to estimate the future work participation rates of target populations.

To assess the impacts of welfare reform, information on the location of entry-level jobs relative to the location of public housing residents and other entry-level job seekers is used. In the case of the three Ohio HAs, the additional information on job location gathered through the cooperative agreements has permitted a more fine-tuned assessment of impacts. In these places, neighborhood-level information makes it possible to consider how many metropolitan area jobs are actually available to inner-city public housing residents after accounting for spatial and transportation barriers and for competition from other entry-level job seekers.

C. Assumptions Underlying Study of Welfare Reform Impacts

A major component of this study is a set of estimates of job participation among public housing residents required by the welfare program to find jobs in the study sites and, stemming from these, estimates of the impacts of welfare reform on housing authority rent revenues. Several key assumptions are necessary in order to make these estimates.

This study develops two methods to estimate the job participation rates of mandated residents of public housing developments. The first, "job seekers-to-jobs ratios," relies on predictions of job growth in the study sites in order to estimate the likelihood that mandated residents will identify entry-level jobs and successfully compete for them. This assumes the accuracy of current predictions of available jobs at some future point in time. It also assumes that when welfare benefits terminate and mandated residents enter the job market, the number of entry-level competitors they will face mirrors the number currently existing.

The second method for estimating job participation rates uses 1990 Bureau of the Census Public Use Microdata Samples (PUMS) data and logistic regression models to predict the job participation behavior of mandated public housing residents. This method assumes that mandated public housing residents will mirror the work participation rates of similar people, in the same places, represented in PUMS. The study emphasizes the more conservative estimates that the two methods yield, but presents a range of possible outcomes using both. The conservative approach provides an upper bound on the likely negative effects on housing authority rent revenues.

In this study, the full impacts of welfare reform are assumed to occur in the termination year. Any significant income loss to tenants and revenue loss to housing authorities will become manifest after the welfare reform clock runs out. The clock will run out at different times in different housing authorities. In Virginia and Ohio, the income and revenue impacts of welfare reform should begin to become apparent in the year 2000. In California, these impacts should occur by 2002. In Texas, the impacts could come as early as next year because of the fact that different mandated residents in that state are being held to different work participation timetables, and some have been given just one year to move to work. Federal welfare reform legislation allows for a five-year assistance period to mandated households. However, states have the option of discontinuing assistance for an interval before the full five years of assistance has been received. After the first three years of assistance to mandated households, the Virginia and Ohio waivers programs require a two year interval without assistance before assistance can be resumed. As welfare reform runs its course, other states may decide to apply different timetables to different groups of mandated households, though remaining within Federal guidelines. In these two States, the estimated impacts that are reported in this study are assumed to occur in the year 2000. Obviously, any program decisions to slow the clock for some or all mandated households would reduce the impact in a given year. Therefore, in Texas, although the estimated impacts reported in the sections that follow are assumed to occur in 1999, in fact the impact will be spread out over a several year period.

Finally, it is assumed that, when TANF benefits are terminated, the number and demographic composition of mandated public housing residents will be the same as the current mandated populations. In fact, the number and the mix of characteristics may be different.

D. The Richmond And Norfolk Cases

Norfolk and Richmond Virginia were chosen as the first places to study and as the field sites because they are both convenient to Washington and they have features that make them attractive when considering the context and impacts of welfare reform. To begin with, RRHA and NRHA are large enough for impacts to register; RRHA manages 4,368 units of conventional public housing and NRHA manages 3,575 units. Furthermore, each currently has a substantial number of residents who will be affected by welfare reform. At RRHA there are 1,158 such residents mandated under TANF to find jobs and at NRHA there are 898 mandated residents. Moreover, because Richmond and Norfolk represent contrasting economies within the State -- Richmond's the relatively stronger and Norfolk's the relatively weaker -- they provide an opportunity for exploring the role of the local economy in the impact of welfare reform. Their convenience was particularly important at the initial stage of the study when the value of pursuing a variety of sources and types of information was still being evaluated.

Virginia's reform provisions make it an interesting State to study in depth. It, like many other states, has enacted its own welfare reform legislation, Virginia Initiative for Employment not Welfare (VIEW), which sets the basic parameters of the State program and takes precedence over many features of TANF. Because its waiver limits assistance to two years of cash and one year of non-cash benefits in any three consecutive years (out of a maximum five years in which welfare assistance can be received), the welfare program that is affecting RRHA and NRHA residents is representative of programs that are more stringent than the Federal program; the latter allows participating households to receive assistance for up to five consecutive years. While Federal law allows states to exempt single parents with children under 12 months of age from the work requirement, VIEW is more generous exempting households with children under eighteen months. And, while TANF requires individuals to go to work no later than two years after receiving assistance, VIEW is more stringent requiring all newly enrolled participants who are not otherwise exempted to participate in work activities within 90 days of receipt of assistance and all current households to participate as soon as they enter into agreements of personal responsibility, as all who are mandated must do.

E. Report Structure

The discussion which follows contains six sections. The first covers labor supply and deals with the number of public housing residents who will be looking for jobs, those who will be competing with them for entry-level jobs and the characteristics that affect employability. Labor demand -- the number of entry-level jobs available in the Richmond, Norfolk and other labor markets is then discussed. This is followed by a discussion of the adequacy of support services, including child care and transportation, that facilitate work participation. Next is an assessment of how welfare reform will affect housing authority rent receipts and potential impact on HUD's PFS payments to housing authorities. A discussion of the impact of welfare reform on tenant income then follows. Finally, some potential policy issues are raised in a concluding section.