Endnotes

1 There are several terms used throughout this report that are fully interchangeable. These include "mandated residents," "mandated households," "mandated heads of households," and others. These all refer to the head of household, in public housing, receiving TANF payments, who must participate in work activity according to TANF or state waiver welfare reform requirements. In a few cases, the reference is made to "total persons in a housing authority." This, too, refers only to heads of household, but without regard to the receipt of TANF or the requirement to work. Finally, whenever reference is made to the group of household heads in public housing but not receiving welfare, they are always referred to as "non-mandated residents." In a few cases, the child of a head of household is included among the "mandated residents" because he or she lives with a parent, receives TANF for his or her own child, and is subject to the work requirements of TANF or a state waiver plan.

2 Among others, Ellwood and Bane have shown that level of education influences the jobs and wages attainable by welfare recipients and, therefore, is linked to welfare dependence. David T. Ellwood and Mary Jo Bane, Welfare Realities: From Rhetoric to Reform, Harvard University Press, Cambridge, MA.

3 The Bureau of Labor Statistics and the Dictionary of Occupational Titles link occupations with typical measures of skill and vocational training. In both classifications, there is a clear connection between greater occupational reach and higher educational attainment.

4 On average, Richmond's residents are judged to function at the 8th grade level. AFDC recipients in the suburban counties are believed to function at a somewhat higher level than City recipients, on average closer to the 10th grade level.

5 The JOBS program is operated by the State Department of Social Services. It provides job readiness training, referrals to various job training opportunities, educational opportunities, and other components. Some of the participants are public housing residents but more are not.

6 Information provided by The Virginia Department of Social Services.

7 To the extent that entry-level jobs are tagged by gender, some male entry-level job seekers will not be competing in the same labor pool with female entry-level job seekers.

8 The Bureau of Labor Statistics defines labor market as the Metropolitan Statistical Area.

9 According to staff at the Virginia Employment Commission.

10 U.S. Bureau of the Census, Public Use Microdata (PUMS), 1990.

11 The Retail Merchant's Association.

12 Economic Assumptions for The United States and Virginia: Calendar Years 1997, 1998, and 1999. Virginia Employment Commission. Economic Information Services Division.

13 The three occupations accounting for the majority of placements in the ESP/JOBS program were housekeeping/janitorial, food service, and nurses aid/companion.

14 Mid-Atlantic Guide to Information on Careers. The District of Columbia, Maryland, Virginia, and West Virginia State Occupational Information Coordinating Committee.

15 In the State as a whole, occupations projected to have the largest number of openings through the year 2005 include sales, cashiers, office clerks, janitors and cleaners, waiters and waitresses, food preparation workers, and nursing aides and orderlies -- all entry-level jobs employing large numbers of women. The greatest growth is projected to occur in the services sector with a 40 percent gain from 1990 to 2005.

16 This is the proportion derived from empirical studies covering different cities and states. See, for example, Virginia L. Carlson and Nikolas C. Theodore, Are There Enough Jobs: Welfare Reform and Labor Market Realities, The Chicago Urban League, 1995; Elizabeth McGregor, "Entry Level Jobs," Occupational Outlook Quarterly, Winter 1990-91; Cutting Wages by Cutting Welfare: The Impact of Reform on The Low-Wage Labor Market, Economic Policy Institute, 1995.

17 These ratios assume that all entry-level jobs will be open to female public housing residents. Census occupational data indicate that most occupants of some entry-level jobs are male. If such jobs are removed from the job growth projections for Richmond, as few as 6,000 of the total of 10,000 entry-level job openings may be available to women. Obviously, the ratio of job seekers to job openings is affected by reductions in both the numerator and the denominator. At the entry-level, fewer people will be competing with female public housing residents but fewer jobs will also be available to them.

18 Industry and Occupational Employment Projections: 1990-2005, Virginia Employment Commission.

19 The degree of concentration of mandated public housing residents in Richmond is also true of residents at other housing authorities, and could have implications for their employment prospects. Paul Ong of the University of California found that more Section 8 program participants in the Los Angeles area had jobs than those in public housing. One explanation provided by Ong is that those with a Section 8 subsidy have greater residential choice and mobility. Paul Ong, "Subsidized Housing and Work Among Welfare Recipients, unpublished paper, Department of Urban Planning, School of Public Policy and Social Research, University of California, Los Angeles, CA, February 12, 1996.

20 John Accordino, Trends in The Richmond Economy: Industry and Labor Force Analysis, Department of Urban Studies and Planning, Virginia Commonwealth University, January 1995. Prepared for the City of Richmond Department of Community Development.

21 Between 1979 and 1994, the number of jobs in the City of Richmond decreased from 187,076 to 175,613. Job growth in the MSA went from 355,105 to 476,999. Another 78,000 jobs are expected in the MSA by 2005.

22 Virginia Employment Commission.

23 The concept of spatial mismatch has been used to describe the disjunction between where jobs are now located as a result of economic restructuring and where job seekers live. The construct was first proposed by John Kain in 1968 and has since been elaborated by many others. John Kain, "Housing Segregation, Negro Employment, and Metropolitan Decentralization," Quarterly Journal of Economics, 82: 195-97, 1968. Scholars have not always found evidence supporting the hypothesis, but a multiplicity of such elements characterizes the situation of public housing residents living in inner-city developments. In all of the cities in this study, the trend is toward metropolitan job deconcentration. Furthermore, inner-city residents seeking suburban jobs are often racial minorities living in racially concentrated areas where information about suburban jobs is often restricted. In addition, these residents have less access to automobiles, and even if they were able to reach suburban jobs by public transportation, they would be spending more time commuting for lower wages. One characteristic of the great majority of mandated residents is the fact that they are female. Although the literature finds general support for the mismatch hypothesis, it focuses almost exclusively on the mismatch problems of minority men or minorities as a group. However, the gender difference could be as critical as some of the other elements that characterize the situation of public housing residents, because entry-level jobs that traditionally employ women may be located in different parts of the metropolitan area than entry-level jobs that traditionally employ men. Depending upon the proportion of jobs that are traditionally occupied by women and on where these jobs are located, spatial mismatch may be more significant for women than for men, or vice versa. Hanson and Gender provide information and maps which illustrate that men and women in the Worcester, Mass. area are indeed employed in different locations in the metropolitan area, a fact that has implications for their access to jobs. Susan Hanson and Geraldine Pratt, Gender, Work, and Space, Routledge, New York, 1995.

24 According to John Accordino, the City's loss of competitive share has several causes. The loss of retail, construction and some service enterprises is a result of continuing population movement from the City to surrounding jurisdictions as well as faster growth overall outside the City. The loss of manufacturing may be a result of a lack of appropriate sites within the city for business expansion as well as perceived problems of doing business in the City. John Accordino, Trends in The Richmond Economy: Industry and Labor Force Analysis, Department of Urban Studies and Planning, Virginia Commonwealth University, January 1995. Prepared for the City of Richmond Department of Community Development.

25 This assessment was provided by the Hampton Roads Planning District Commission.

26 Research has shown that access to neighborhood jobs in the Los Angeles area -- jobs in close proximity to where recipients reside -- is negatively related to the percentage of the working age population who rely on welfare. Evelyn Blumenberg and Paul Ong, "Job Accessibility and Welfare Usage: Evidence from Los Angeles," to be published in, The Journal of Policy Analysis and Management, Fall 1998.

27 Neighborhood level ratios have been provided under a Cooperative Agreement between HUD's Office of Policy Development and Research and researchers at Case Western Reserve University and The University of Wisconsin. See Appendices A and B for details of the methodology used to provide neighborhood level ratios.

28 For the purposes of this analysis, neighborhoods are designated by zip code areas.

29 Under the Cooperative Agreement, a 45 minute commute time was the window within which job accessibility was assessed. This time interval reflects the time in transit, that is, the time it would take to go from the center of a neighborhood, in this case the neighborhoods where mandated public housing residents are concentrated, and an employment destination. The interval does not incorporate the time it would take to make ancillary trips that might be required before the journey to work is begun. Thus, it does not include the time it would take to drop children off at day care centers or baby sitters. Nor does it include the time it would take to walk from one's house to a bus stop or train terminal.

30 In Cleveland, two methods have been used to compute seeker-to-job ratios. The distance ring method and the CTPP contour method. For details of these procedures, see Appendix B. In general, the CTPP method is far more time consuming to carry out, although it could be more precise. It is for this reason that one city -- in this case Cleveland -- was used as a test for comparing the two methods. Using either method, neighborhood ratios are generally larger than metropolitan ratios, as would be expected. The two methods do produce different rank ordering of ratios across the City neighborhoods. Assuming access to cars, the CTPP neighborhood ratios are higher than those generated by the distance ring method. However, assuming dependence on public transportation, neighborhood ratios generated by the ring method are higher in two of the four Cleveland neighborhoods of interest.

31 The ratios are for neighborhoods which together include at least three-quarters of the mandated population, listed in order of concentration.

32 It would be fair to expect that most often those with access to automobiles can cover more distance and, therefore, reach more jobs within a given time interval. Although there is disagreement about the proportion, it is generally believed that the majority of mandated residents are dependent on public transportation. Surveys done of other low-income households indicate that they too are often dependent on public transportation. (See p. 21) Because of this high rate of dependence on public transportation, metropolitan and neighborhood ratios assuming the use of public transportation are more often reported here. If it were possible to get more accurate information on the proportion of mandated residents and other entry-level job seekers with access to cars, it might be possible to weight the ratios to incorporate the mode of transportation utilized.

33 For the purposes of this analysis, it was assumed that jobs in which fewer than 15% of the occupants were women were not open to women. This assumption reflects historical patterns of segmentation. Obviously, non-traditional training programs for women and other changes in hiring practices could reduce segmentation over time.

34 The double issue of job deconcentration and the reliance on an inadequate public transit system characterizes the situation of inner-city residents in other places as well. Using PUMS data for Northern New Jersey, McLafferty and Preston showed that African-American women's long commuting times were linked to their heavy reliance on mass transit and poor spatial access to employment. Sara McLafferty and Valerie Preston, "Spatial Mismatch and Labor Market Segmentation for African-American and Latina Women, Economic Geography, Vol. 68, No. 4, October 1992, p406.

35 These are telephone representative positions at the Capital One Company.

36 Housing Authority Staff estimate that about 50 percent of its households have access to cars. According to a 1993 HUD Report, 34 percent of public housing residents in the South have cars. Characteristics of HUD-Assisted Renters and Their Units in 1993, U.S. Department of Housing and Urban Development, Washington, DC, May 1997.

37 TRT now receives a fifty percent subsidy from the Federal government but the amount of subsidy will be cut this year as federal funding phases out. The State is expected to pick up some but not all of the shortfall.

38 In Richmond, there are 1,608 slots available in voluntarily registered day care homes and 1,080 in regulated and registered child care homes. For each registered slot, there may be as many as three unregistered slots in day care homes bringing the total of these slots to about 10,700 according to staff at the Memorial Child Guidance Prevention Clinic and other members of the Richmond day care provider network.

39 There are insufficient slots for pre-schoolers at Creighton, Fairfield, and Whitcomb but Gilpin and Hillside, the other public housing family developments, are well served.

40 These are telephone representative jobs at Capital One.

41 View of the Memorial Child Guidance Prevention Clinic.

42 In suburban locations, the cost of caring for a pre-school age child in a home is nearly double, about $100 a week according to the Memorial Child Guidance Prevention Clinic.

43 The Memorial Child Guidance Prevention Clinic.

44 Child care costs are an eligible deduction from income when computing tenant rent; this child care cost burden partially shifts to the Housing Authority. Virginia does have a fee-based system of child care support available to all low-income households that is administered by DSS. This program requires participants to contribute 10 percent of gross monthly income but is heavily subscribed to and has limited funding. Federal and State, non-TANF funds cover the subsidy costs.

45 Information provided by The Planning Council, Dependent Care Services in Norfolk.

46 Information provided by NRHA.

47 The Planning Council indicates that in 1990 working Norfolk residents had 6,317 children under age 6.

48 Some of the family day care homes that have been certified and subsidized through The Step Ahead Program are in the areas of Calvert Square, Tidewater Gardens, Diggs Town and Oakleaf Forest family developments.

49 Schedule of Fees from Norfolk Day Care Network.

50 In the past, food subsidies were available to licensed providers, but now that such subsidies are no longer provided, there are few incentives for operators to seek licensing.

51 Because of this emphasis, assistance recipients, including public housing residents, who have not yet acquired vocational skills may now have little opportunity to do so.

52 RCAC also provides initial assessments of job readiness, pursuant to Virginia's VIEW program, for existing and new clients of the Richmond branch of the Virginia Department of Social Services. RCAC provides a relatively short program for a subset of those it assesses offering skills that job seekers will need as they enter the labor market. This includes resume writing, workplace behavioral training, classes on personal organization and scheduling, information on how to dress for work, and other components.

53 In December 1994 Norfolk became the first city in the state of Virginia to be designated a Federal Empowerment/ Enterprise Community.

54 Section 3, as amended by the Housing and Community Development Act of 1992, "requires that economic opportunities generated by certain HUD financial assistance for housing . . . and community development programs . . . shall, to the greatest extent feasible, be given to low- and very low-income persons . . . ." 24 CFR Part 135.

55 Eleven percent of mandated residents currently have some income from wages that would not be affected by welfare reform. Furthermore, both mandated tenants who work and those who don't have other sources of income besides AFDC. For example, income from child support would fall into this category.

56 The average tenant rent for all households at RRHA is $156 per month.

57 This report assumes that in the year 2000, the first time that benefit limits will be reached for the current assisted population, the expected population of assisted households will resemble the current population in all relevant characteristics. This allows the maximum impact of welfare reform to occur because all mandated residents would have received assistance long enough for it to have come to an end. The alternative would be to utilize welfare caseload dynamics to predict the number of exits from public housing and the characteristics of those exiting. But historical data on welfare caseloads and on public housing residency do not relate well to current circumstances.

58 Virginia's AFDC payment to a family of three was $291 a month. The majority of states pay more. Even though low, under VIEW guidelines, assistance will continue to those AFDC households who have children under 18 months and/or care for a disabled household member. There are 278 such households living in RRHA family developments.

59 PFS has sometimes funded housing authorities at somewhat less than 100 percent of their total need, but these shortfalls are allocated among housing authorities on a pro rata basis.

60 The rent receipts of 274 AFDC households will be untouched by welfare reform because these households are exempt under VIEW guidelines. The rent receipts of mandated AFDC households are also partially sheltered because 10 percent of these households have some wage income.

61 According to PIH Notice 96-81(HA) of September 30, 1996, housing authorities may set minimum rents of anywhere between $0 and $50. The notice was to have expired on September 1, 1997 but it has been extended for another year under the continuing appropriation. The House bill calls for a minimum rent of $50 with an allowance for hardship exclusions, while the Senate bill calls for a minimum not to exceed $25 that would also incorporate hardship exclusions. The Baltimore Housing Authority is one case where the currently allowed minimum of $0 is in effect. Although a number of housing authorities have adopted a $50 minimum, they do not include the largest authorities. The Administration's Public Housing Reform bill proposes to set minimum rents at $25 per month, and permits either HUD or the PHA to grant hardship exemptions. Under this bill, Richmond's $50 minimum rent would be reduced to $25.

62 Although minimum rents can be waived for three months for hardship cases, RRHA has received no requests for a waiver from households paying minimum rents.

63 Close to one hundred units (of a total of 4,398) turn over each month. This allows the Housing Authority to increase the number of higher rent paying households over time.

64 However, in a number of the larger developments where mandated residents live, the average resident stay is between 13 and 16 years.

65 Over 500 waiting list households are employed and close to 200 have non-employment income from sources other than TANF including Social Security and SSI.

66 Section 402 (b) of the Balanced Budget Downpayment Act, I, 1996 amended Section 3 (a) (2) of the 1937 Act to permit the establishment of ceiling rents that reflect reasonable market value but that are not less than monthly operating costs. Some housing authorities believe that the requirement to cover operating costs pushes rents beyond levels that would make them attractive to working households in certain locales. The Administration's pending Public Housing Reform bill would allow HAs, for family developments, to adopt ceiling rents that reflect market value but that are not less than 75 percent of operating costs.

67 Legislation to create block grant funding of the Public Housing program has been proposed at various times. Although housing authorities might end up with more flexibility with respect to tenant selection under a block grant, the housing authorities are concerned that a capped funding mechanism would also leave them unable to cover part of the income loss when currently mandated residents are unable to replace their assistance income.

68 Welfare officials in the two cities are not sanguine about the prospects of a successful transition from welfare to work on the part of mandated households. It is the judgment of officials at the Richmond office of the State Department of Social Services that about one-half of the mandated population will be relatively easy to place in work situations. The other half will be harder to place because of multiple barriers including educational deficits, criminal records, and substance abuse problems. Officials at the Norfolk office of the State Department of Social Services believe that "large numbers" of clients in Norfolk will not make the transition to work at the 3 and 5 year cut-off points because of their educational and other deficits even though they may be able to find jobs in the near term to fulfill work participation requirements.

69 Easily measured characteristics that have been shown in PUMS to be important for estimating the employability of mandated residents include: educational attainment; sex, race and age of head; age of children; single parenthood; and, marital status. Logistic regressions from PUMS were utilized to determine which characteristics were of importance in each site. The regression coefficients were then applied to the distribution of mandated residents to estimate how many would have been employed for at least 30 hours per week for at least 26 weeks.

70 The Housing Authority in Richmond suspects that many of these households have unreported sources of income because they have no problem coming up with the minimum rent nor do they do what is required to have the sanctions lifted. For such households, participation in work and training could require an unacceptable opportunity cost.

71 Working at least 30 hours a week is the criterion adopted in this study for "having worked" because it is the minimum work requirement to be applied to assisted households under TANF by the year 2000. Households that met the "having worked" criterion also had to have worked for at least one-half of the year.

72 These work participation rates reflect the 1989 economy. Trends in employment since then were checked in each of the sites using the Census Bureau's Current Population Survey (CPS) for the years from 1990 through 1997. For the CPS population similar to the public housing TANF mandated population, the estimated proportion employed at least 30 hours per week for at least 26 weeks showed the following: Richmond, Cleveland and Dallas were virtually unchanged; Norfolk dropped to a level about one third less employed; employment in Columbus rose about five percentage points; Toledo's sample was extremely small for the target group and employment was very volatile, but for larger groups seemed to stay steady; employment in Los Angeles dropped about five percentage points; and, there is a mixed picture in San Francisco, where employment seemed to drop about ten percentage points in the MSA as a whole, but may have increased in the central city based on a fairly small sample size (for a larger group without the target group's age, race, ethnicity, kids and marital status constraints the employment in the central city decreased slightly). The large change in Norfolk's economy from 1990 to 1997 is reflected in the PUMS work participation estimates. Appendix C contains the details of how these trends were developed.

73 In fact, the work participation rates of proxy households may be peak work participation levels since they reflect just a one year period. Over a longer time period, the work participation of households with the characteristics of proxy households should fall based upon the preponderance of evidence about the work participation patterns of such households. Many of these households hold jobs that are at the least stable end of the labor market. Necessarily, the work participation estimates used in this study assume a long-term trend and are, therefore, optimistic.

74 In order to calculate the fiscal impacts of welfare reform, this study assumes that the number of residents now mandated will be seeking entry-level jobs when assistance terminates. In fact, some currently mandated assistance recipients will have left the welfare rolls by then and those who replace them could still be eligible for assistance which they can use to pay rent. Although this study does not utilize a dynamic model to predict the impacts of welfare reform on rent revenues, one possibility is that adverse fiscal impacts would be mitigated in the assistance termination year if some current tenants were replaced by households still entitled to assistance income. However, over a longer time frame, the churning created by welfare reform could reduce HA rent revenues considerably as successful households moved on and unsuccessful households clung to their public housing units. To some extent, housing authority policies affecting tenant mobility including evictions for failure to pay rent, marketing to working households, etc. will affect the longer term fiscal consequences of welfare reform.

75 Richmond is the only one of the eight housing authorities studied in which the estimates of work participation based on predicted job growth are higher than the estimate based on the assumption that mandated residents will demonstrate the same work participation rates as the population they resemble.

76 Selecting an annual income finesses the issue of hours worked and hourly wage rates. There are some residents who will work longer hours for lower wages and others who will work shorter hours for higher wages.

77 Neighborhood estimates are based on taking the weighted sum of the revenue impacts in each of the Cleveland, Columbus and Toledo neighborhoods where residents are concentrated.

78 According to data from the Virginia JOBS program reported to HUD by the Virginia Department of Social Services.

79 The State provides DSS with a list of truants accompanied by social security numbers so that they can be matched with client files to enforce sanctions. Beginning with the 1997-98 school year, NRHA will also receive attendance reports on public housing youth from Norfolk public schools, listing students who have missed 3 or more consecutive days of school during the month.

80 Welfare reform necessarily places the Housing Authority and DSS in opposing camps. From the point of view of DSS, households who lose benefits because of sanctions free up resources for other households. In terms of the impact on the housing authorities, the income reductions as a result of sanctions automatically trigger lower rents. Both DSS and the Housing Authority agree that present rent rules have a perverse affect by rewarding sanctioned behaviors. The Housing Authorities would like to find a way, as the Portland Housing Authority has done, to maintain the rent levels of sanctioned households, and HUD's Public Housing Reform bill contains such a remedy. The bill would disallow reductions in public or assisted housing rents that are triggered by tenant income reductions caused by the application of sanctions for non-compliance with welfare or public assistance program requirements. The bill allows rent reductions related to drops in income that result from the termination of welfare assistance because of time-limits.

81 The fraud unit finds fraud in about one-third of the cases that it investigates. In some cases, people have well paying jobs that they have not reported. But in other cases, the unit finds that the sanctions were unwarranted.

82 However, when the loss of non-cash benefits is factored in, it is less clear whether those with minimum wage jobs have improved their economic situation. In addition to rent subsidy, AFDC recipients were also entitled to Food Stamps and medical benefits which have monetary values. On the other hand, the Earned Income Tax Credit and any HA earned income disregards should add to the net value of income from employment.

83 In 1996 the poverty line was $12,600 for a mother with two children.