I. Introduction

This paper was inspired by major changes in the structure of the U.S. welfare system, as called for by the federal Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, and the important groundwork of two organizations that have examined the concept of wage-based community service employment (CSE). The Center on Budget and Policy Priorities (CBPP) in Washington, D.C., led by Cliff Johnson, and the Center for Law and Social Policy (CLASP), also in Washington, D.C., led by Steve Savner, have produced extensive written material on this welfare-to-work program approach, much of which is summarized in this paper.

The purpose of this paper is to supplement the ideas that have already been developed and to flesh out how wage-based CSE might work from the perspective of the local agencies participating in the implementation of welfare reform. The paper begins with a discussion of the differences between wage-based CSE and other types of community service work. Topics addressed in the subsequent sections are:

1. What Is Wage-Based CSE?

2. Who Can Benefit from Wage-Based CSE? Who Should Participate?

3. Why and How to Assure "Real Work" in Wage-Based CSE

4. Wage-Based CSE as Job Skills Training and/or a Stepping-Stone to Unsubsidized Employment

5. Can Wage-Based CSE Serve as the New Safety Net?

6. Can Wage-Based CSE Benefit Low-Income Communities?

7. Two Program Examples: New Hope and PACT

8. Funding Wage-Based CSE: Should States and Localities Use TANF Monies?

9. Program Costs

10. Options for Administration: Who Should Run Wage-Based CSE?

The paper concludes with observations about the potential of wage-based CSE and the foundation for it that already exists in many communities.

II. What Is Wage-Based CSE?

Wage-based CSE is an idea about how to provide jobs under a reformed welfare system for some welfare recipients who are not able to find jobs in the regular, or "competitive," labor market. It is not defined in law or regulation and has been implemented only in a few locations. In actual operation, wage-based CSE might take a variety of forms, built around the concept’s two essential parts: (1) that jobholders earn wages in a standard employer-employee relationship and (2) that the work performed somehow benefits the wider community or society at large.

Compared to options available in the past, wage-based CSE is a hybrid form of work for public assistance recipients, combining elements of community service — work that would otherwise go undone by paid employees of government agencies and nonprofit or charitable organizations — and subsidized work or on-the-job training, which have used public funds to induce employers to hire and train disadvantaged workers. Wage-based CSE can also take the form of a publicly funded job creation program, including one that helps finance new enterprises (by paying the wages); these enterprises resemble the "affirmative businesses" that employ disabled people nationwide.

The Community Work Experience Programs for welfare recipients (abbreviated as CWEP) that were prevalent in the mid-1980s also had an element of community service. CWEP participants were intended to supplement the workforces of government and nonprofit agencies, performing work that would not otherwise get done; "displacement" of regular workers was forbidden. CWEPs were most often structured as mandatory work in exchange for cash benefits, however, and the government and nonprofit worksite employers provided supervision for the program participants and monitored and reported on their attendance and performance to the welfare agency. In contrast, participants in wage-based CSE are paid wages, rather than given a welfare check; the wages are based on actual hours of work; participants receive all the benefits and protections of regular employees; and they are required to pay taxes on their earnings. In these ways, wage-based CSE looks like "real work" to participants. Participants in wage-based CSE are expected to adhere to the workplace standards of their worksites, which was the case in most CWEP positions as well, but — to complete the "real work" picture — wage-based CSE can be structured so that worksite employers are responsible for hiring and firing participants.

Wage-based CSE is also reminiscent of subsidized work or publicly funded jobs programs, like those operated during the Depression under the Works Progress Administration (WPA) and in the 1970s under the Comprehensive Employment and Training Act (CETA). This is because the money used by employers to pay participants’ wages comes from a public source, in this case the Temporary Assistance for Needy Families (TANF) program,¹ federal Welfare-to-Work grants, or state and local funds.

Finally, wage-based CSE can look like on-the-job training (OJT) for disadvantaged workers. There is a long-standing practice in the publicly funded job training system of giving financial incentives to employers who will take on disadvantaged (untrained) workers and upgrade their skills to entry level or higher for the industry. In the case of wage-based CSE for public assistance recipients, the incentive to the employer is that the welfare system covers the cost of the wages of these employees for a defined placement period.

The Role of Private, For-Profit Employers

One open question about the wage-based CSE concept is whether there are circumstances in which private, for-profit companies should be allowed to be worksite employers. In the past, private employers were excluded from CWEP, and their role in subsidized employment and OJT programs entailed an exchange for the free or discounted labor of program participants they received. In order to ensure that the workers and taxpayers benefitted as well, private employers were expected to commit to hiring participants into permanent, unsubsidized jobs at the end of their training or subsidized tryout.

Under TANF, the never-very-distinct lines between subsidized employment, on-the-job training, and paid community service — as implemented in publicly funded programs for low-income and disadvantaged persons — are even more blurred. States and localities have a great deal of flexibility under TANF to structure work programs and to mix and match elements of past programs. Theoretically, then, private sector employers could be included in a wage-based CSE program. If they are able to commit to permanent jobs in exchange for temporary free labor in a wage-based CSE program, or if they offer some other long-term social benefit, a public/private approach might be even more attractive than a program limited to public agency and nonprofit worksites. By including private employers, wage-based CSE might, for example, gain flexibility, a wider range of worksites, and a larger scale.

Another reason to consider a role for private employers in wage-based CSE programs is that distinctions among nonprofit, for-profit, and public organizations are increasingly unclear in this era of downsized government and privatization of public services. For example, large cities have hospitals and all types of health care facilities of each type. Is a job in a hospital "community service" if the hospital is public or nonprofit, but not if it is a private institution? If private, for-profit contractors to public agencies are delivering services to benefit communities or needy populations, should these organizations be considered for placements of workers in a wage-based CSE program? If program participants provide free, publicly funded labor to private organizations through wage-based CSE but receive job skills training, supervision, and "real work" experience in exchange, does the value received by participants make this an acceptable arrangement?

For state and local administrators of TANF, the central question about extending wage-based CSE to for-profit employers is: Should "community service" be defined by the nature of the work or the nature of the organization providing the service? This question can probably be answered only on a case-by-case basis, in the context of local politics and the image of any particular for-profit provider of public services. Thus, state and local administrators of TANF might approach answering the question about for-profit employers by establishing qualifying criteria that cover community benefits (What is the service to the community?); benefits to participants (such as training, commitments to hire, wage and fringe benefit packages offered at hiring); and the employer’s record on compliance with laws and rules governing its business, including labor practices. TANF officials might also consider setting a fixed length of time for a wage-based CSE placement with private employers that is shorter than placements with public and nonprofit employers.

The Relevant Rules

Wage-based CSE programs for TANF recipients may be operated by many different types of organizations, and the programs are not required by the federal statute or regulations to provide the same basic services, target the same populations, or do anything else in the same ways except with regard to the fundamentals of TANF and labor law. Unlike previous programs that required welfare recipients to work in nonprofit or public agencies in exchange for their welfare benefits, the number of hours a participant works per week under wage-based CSE is not tied to the amount of her TANF assistance. There are, however, rules governing when recipients may be required to work under TANF:

1. Single parents are required to be working or looking for work after they have received assistance for a maximum of 24 months. In order for these parents to be counted toward a state’s TANF participation rate, they must be "engaged in work" at least 25 hours per week beginning in FY 1999. (This weekly work standard goes to 30 hours in FY 2000; for adults in two-parent households, the standard starts and stays at 35 hours per week.) At state option, work requirements may start sooner.

2. Exemptions are allowed for single parents with children under one year of age, at state option.

3. Parents of children under age six must be able to access needed child care for TANF work program participation; sanctions for noncompliance cannot be carried out unless needed child care is available.

The rules governing employers of participants in wage-based CSE are similar to those in other employer-employee relationships:²

1. The Fair Labor Standards Act and other labor laws apply, which means that wage-based CSE participants must be paid at least the federal minimum wage (except in certain training situations).
2. The Occupational Safety and Health Act applies to welfare workers in the same way it applies to other workers. In states where OSHA does not have jurisdiction over public employees, and wage-based CSE participants are employees of public agencies, they are exempt.
3. The U.S. Department of the Treasury has determined that the earnings of participants in CSE programs are usually taxable, and therefore the employers and employees are required to contribute under the Federal Insurance Contributions Act (FICA) for the Federal Old Age and Survivors Insurance and Disability Insurance systems. This means that employers must pay their share of FICA for participants.
4. Wage-based CSE participants generally are covered by the Unemployment Insurance system, although there is an exception for participants working for public or nonprofit agencies and providing services primarily for community benefit or to meet their own needs that would not otherwise normally be provided by other employees. The agency also must provide workers’ compensation coverage.
5. Employers of participants in wage-based CSE jobs must follow antidiscrimination laws that apply to all workers.
   

III. Who Can Benefit from Wage-Based CSE? Who Should Participate?

Wage-based CSE holds out the promise of achieving multiple goals of the newly reformed welfare system. It can serve as a mechanism for welfare recipients to satisfy the societal obligation of work and family support that initially motivated the reforms, and at the same time provide a measure of income security for families with children when parents are not ready for work in the competitive labor market or when and where the competitive labor market is not able to absorb the number of welfare recipients needing jobs under the new TANF rules. Like previous forms of community service for welfare recipients, wage-based CSE offers a pool of labor for getting valuable work done in the community that is not otherwise being accomplished, but with greater benefits and protections than in the past for those who are doing the work. There may also be ways to design wage-based CSE so that it serves as a training ground for unsubsidized employment in the private sector and as a real transitional step toward self-sufficiency.

With all this potential, states and localities may have trouble deciding who should participate in wage-based CSE and/or building and financing a program large enough to serve everyone who could benefit. Given the limited real-world experience with the concept, and limited resources, TANF-implementing organizations might start experimenting by focusing on a fairly narrowly defined population, such as TANF recipients who have participated in work-related activities in the past and have reached the 24-month limit for assistance without work. This group will certainly grow quickly, and a subset will quickly become the group approaching the 60-month lifetime TANF assistance limit established by PRWORA.³ Trying out wage-based CSE on this bellwether group may offer valuable lessons for how to use wage-based CSE as a stepping-stone to unsubsidized work. (The current continuing decreases in TANF cases give states and localities some time to plan and test work options like wage-based CSE before the need for them is critical — which may be in FY 2001 or 2002, when the TANF participation standards are 45 percent and 50 percent, respectively, of the single-parent caseload.)

A Rationale for Targeting Wage-Based CSE

In an untargeted program, everyone potentially eligible for a particular service or activity has an equal chance of being assigned and participating. Selections might be made on a first- come/first-served basis or some sort of lottery. This strategy can make sense when nothing is known about who benefits from the service and how much. It does not make sense for wage-based CSE, however, given that the program resources are likely to be limited and that a great deal is known about what works in helping welfare recipients find work and become self-sufficient. Although there are no specific research findings about the results of wage-based CSE, and there have been only a few studies of work programs operated in the context of time limits on the receipt of public assistance, many other work options for welfare recipients have been studied, including some that are similar to wage-based CSE. Taking that accumulated knowledge together with some overall TANF goals, it is possible to derive some broad principles for targeting wage-based CSE and to draw from them inferences about program design.

First, to use program resources most efficiently and avoid wasting any TANF recipient’s lifetime reserve of assistance, those who can get a job in the competitive labor market should not be given wage-based CSE jobs. The widely used and documented test of recipients’ job-getting ability is a supervised, assisted job search activity. Typically, group job search is run as a "workshop" that lasts for two to four weeks and provides classroom training in presenting oneself to employers, practice in interviewing, help with résumé preparation, supervision and resources for making telephone calls to potential employers, and job development by workshop staff, who try to sell employers on the benefits of hiring from the workshop, or who market the qualifications of individual participants. Thus, in order to be considered for a wage-based CSE position, most TANF recipients should have participated in a job search activity recently, have a usable résumé, but have been unsuccessful in landing a job with job search assistance.

Second, to use wage-based CSE worksites most effectively, program staff need to try to understand why candidates have not succeeded in finding unsubsidized work. Lack of work experience, or lack of recent work experience, are employability problems for which wage-based CSE is a particularly good solution, and recipients who need better résumés and job references can be helped by such placements. However, poor communication skills; poor reading, writing, and computational skills; or hostile attitudes are not employability problems that can be solved in a typical wage-based CSE job. In fact, it is as unlikely that recipients with these problems will work out in a CSE worksite as in a competitive labor market job unless they are being provided other services simultaneously. This suggests that some sort of assessment is necessary in order to develop a pool of candidates for wage-based CSE.

Third, under TANF’s time limits, some subgroups of the recipient caseload are going to be "needier" than others and might be given priority for wage-based CSE, consistent with the welfare reform intention of providing work, not cash, as the fallback for recipients who cannot find their own jobs. The neediest group of all, under TANF rules, will be those who use up their lifetime limit of cash assistance and still have no job prospects. Recipients who have been out of the job market for many years and are closing in on the lifetime limit might be targeted for wage-based CSE in order to help them make the transition from cash assistance to self-support deliberately, rather than precipitously.

Fourth, in localities where openings for entry-level or slightly above entry-level jobs are scarce and the TANF population eventually needing to work is larger than the local economy can absorb, wage-based CSE should be used as a business development strategy, targeting participants who can be matched with employers to create or expand enterprises that will sustain them after a wage-based CSE position ends. For example, three or four TANF recipients with sewing skills matched to an employer with the ability to produce and market a mail-order line of sewn products could generate enough demand during the wage-based CSE placement period to build a business able to employ them as regular workers.⁴

Fifth, states and localities might want to experiment with the potential of wage-based CSE as a skills upgrading approach. If a wide range of worksites were available in a wage-based CSE program, the program could be used as a kind of apprenticeship training for occupations that require previous work experience in the occupation. This suggests selecting some candidates who have specific occupational goals that they cannot fulfill without more preparation.

General Assistance Recipients

Wage-based CSE is appropriate for recipients of state- and locally funded General Assistance (GA), as demonstrated by New York City’s Parks Career Training (PACT) program, which is described in Section VIII. GA is a last-resort form of welfare for people who do not qualify for TANF or Supplemental Security Income (SSI) or Unemployment Insurance; it is provided by most, but not all, states and by some localities where there is either not a state program at all or not a statewide program. Most GA recipients are adults not living in families, and many are the noncustodial parents of children receiving TANF benefits in another household. Assistance is more limited than AFDC was or TANF is now — with lower benefits, stricter time limits, and more stringent work requirements — and coverage has been shrinking for about a decade.⁵

GA recipients are likely to have poor literacy skills, uneven work histories and/or extended periods of unemployment, health problems and/or minor disabilities (that is, not severe enough — or not well-documented enough — to qualify them for SSI), and substance abuse problems. They are also likely to have children who are not living with them. Thus, many, if not most, are likely to need and can benefit from the potential transitional, income, and job training benefits of wage-based CSE.

The issues for states and localities about offering wage-based CSE to GA recipients are primarily financial. What funding sources are available? What priority should GA recipients have for wage-based CSE in relation to TANF recipients? What proportion of the GA caseload should states and localities aim to reach with wage-based CSE?

Should Wage-Based CSE Be Voluntary? Options for Participant Choice

During the 1980s there was considerable debate within the employment and training field about whether welfare-to-work programs should be voluntary or mandatory. Proponents of voluntary programs argued that participants would be more likely to succeed at something they chose, and proponents of mandatory programs argued that many welfare recipients who could be successful at finding and keeping jobs would not initiate the process and thus needed a "push." Examples of both types of programs were shown to be effective in increasing the employment and earnings of participants, although the mandatory programs generally operated at a much larger scale than the voluntary programs. Evaluations of a range of welfare-to-work programs from the 1980s also showed that the distinction between mandatory and voluntary was not clear-cut. Rather, "mandatoriness" could be defined in degrees, depending on the specific activities that recipients were required to participate in, the intensity and duration of their required participation, and the sanctions that were enforced for noncompliance.⁶

For wage-based CSE, a sharp distinction between mandatory and voluntary approaches is not helpful in thinking about program design. A more salient question is: How much choice should candidates for wage-based CSE have about which aspects of their participation? Some choices, for example, depend on whether wage-based CSE is embedded in a system of activities for which there is a clear progression and criteria for "who goes where" or whether the program is intended to be free-standing. In a system with a defined progression, TANF recipients may have little choice about whether to accept an offer of a wage-based CSE job (if they do not find their own jobs) but may be able to select the type of CSE jobs they prefer. Where wage-based CSE is free-standing, the staff may have the most information, control, and choice — resulting in a more directive, if not more mandatory, approach.

Another issue with implications for participant choice is whether the program aims for the best possible match between a worksite and a participant or whether the assumption is that it will operate on the basis of trial-and-error. If an optimal match is desired, participants and worksite employers need to be provided variety, choice, information, and preparation. If a wage-based CSE program operator leans toward the trial-and-error method, then the program must be flexible enough to help participants quickly change worksites and must apply broad criteria for when, why, and how often this will be allowed.

Following are some considerations for wage-based CSE program design that involve more or less choice on the part of participants:

• Program goals: Who will be targeted? If long-term welfare recipients who have not been active in work activities are targeted for wage-based CSE, some may need a "push." Recipients who have already participated in job search and other work activities may be motivated enough to volunteer for wage-based CSE, especially if it promises either job skills training or an unsubsidized job at the end. (New York City’s PACT program was marketed to participants in unpaid work experience with excellent response, even though it entailed twice the commitment of work hours, because a "real job" — although temporary — was to follow the unpaid training period.)

• Preceding program activities: What will happen for participants before they start at a CSE worksite? Does job search assistance include some job readiness training (for example, about employer expectations, handling conflict and stress, and making contingency plans)? Is there any assessment of occupational interests? Are candidates for wage-based CSE able to visit potential worksites before they interview with employers? Are they able to choose worksites based on their interests? Programs providing more work preparation prior to wage-based CSE and more information about work options are in a better position to match participants with worksite employers.

• Simultaneous program activities: Will participants be involved in classroom training while they work in a wage-based CSE job? Along with work in a wage-based CSE position, TANF recipients may elect or be required to take basic skills classes, English as a Second Language, classes to prepare for the GED exam for high school equivalency, or other types of literacy and job training. In this case, some directive matching is required to ensure that participants who need to improve their literacy skills can function adequately for worksite employers while they learn. However, an "optimal match" to satisfy participants’ interests may be less important for such participants than a wage-based CSE job that works, in terms of location and scheduling, for the participant to take classes simultaneously.

• Work hours: Can wage-based CSE participants choose to work more than the minimum requirement? In order to count toward TANF participation goals (and to qualify for TANF funding), the TANF recipients working in wage-based CSE jobs must work at least 25 hours per week in FY 1999 and 30 hours per week in FY 2000. Should they be able to work longer hours if their employers provide the work and the participants want it? This is, in part, a financial question for TANF— that is, can the program afford to pay for more work per participant? — and a question of how much choice participants should have about their schedules. If administrators of wage-based CSE programs decide that they can afford to subsidize more paid CSE (and the related child care expenses), they may want to establish criteria for full-time or almost full-time positions that assure that participants make a free, informed choice about these schedules and are not pressured by employers, that participants choosing full-time or nearly full-time work are monitored carefully for adjustment, and that participants who cannot handle increased hours are able to cut back without penalty.

• Alternatives that do not pay wages: If wage-based CSE is targeted to the TANF recipients who cannot get a job in the competitive labor market — at a point in time, anyway — under what circumstances might a work experience position that does not pay wages make sense? The discussion of "real work" in the next section of the paper points to some advantages of the wage-paying relationship between employers and employees as well as some disadvantages when participants are not ready for real work. CWEP — work in exchange for benefits — may provide a "less real" work tryout that is useful for some TANF recipients in a progression toward wage-based CSE and unsubsidized, competitive jobs;⁷ the choice of unpaid CSE might be offered to participants who are unable to find a satisfactory waged CSE placement or who have been fired from a paid CSE position.

Implications of Targeting Policies for Program Capacity and Operations

If TANF administrators prefer to serve a varied population in wage-based CSE and to help some participants upgrade their skills — rather than to target the program narrowly to those on the verge of losing their benefits — it will be important to start early on the task of assessing, finding work options for, and managing the group of recipients approaching their lifetime limit on TANF cash assistance. This is because, in order for wage-based CSE to fulfill its potential as a skills upgrading option, the rest of the TANF program must be under control and working smoothly so that the CSE component is not overwhelmed with last-resort cases. In general, to succeed at this, TANF work programs need:

1. Management information systems in place to track recipients according to their cumulative assistance, the work-related activities they have been involved in over time, and their TANF status (exempt from work requirements, exempt from time limits, etc.).

2. Enough job search assistance capacity to "test" most of the TANF caseload in the labor market over time.

3. Postplacement services to help recipients retain their jobs and move up as well as to counsel and re-place them when they lose jobs.

4. A variety of options to help recipients who are approaching their time limits improve their self-sufficiency prospects. Some options may be designed to help with personal problems rather than employment-related skills. (TANF programs also need the capacity to determine when personal issues, such as abusive relationships, are holding people back.)

5. Clear criteria for deciding which recipients will be among the 20 percent exempt from 60-month time limits.

6. Non-TANF funding streams that can be used to support a portion of wage-based CSE slots so that participation in these placements does not count against TANF recipients’ lifetime assistance limits, and clear criteria for deciding which recipients will get extra time to prepare for the end of assistance that non-TANF funding provides.

7. Job developers to work with people approaching their limits (and at earlier stages).

Where state and local TANF administrators decide to reserve wage-based CSE for those recipients who are up against their lifetime cash assistance limits — offering this option as a last resort for people who are likely to have a lot to deal with in their lives — it will be important to plan for extensive support services, staff the program adequately to maintain regular contact (preferably on site about once a week) with all worksite employers, and recruit worksites where the work and the work expectations are somewhat flexible and the regular contact is acceptable. Also, job developers should work with participants throughout their wage-based CSE placements in order to help them move into unsubsidized work as quickly as possible. Worksite employers for these participants should be given notice that the placements may not last the maximum time allowed. (The New Hope initiative in Milwaukee, Wisconsin, followed this policy for all the participants working in community service jobs, with no apparent loss of employer interest.)⁸

IV. Why and How to Assure "Real Work" in Wage-Based CSE

Why Real Work Matters

The criticisms by participants of workfare or Community Work Experience Programs (CWEPs) of the 1980s centered on the terms and conditions that made CWEP jobs unlike and less desirable than real work. Even though participants believed themselves to be performing jobs very similar to their co-workers’ in terms of the jobs’ value to the organizations where they were placed, and CWEP sponsors agreed that participants’ contributions were meaningful, the participants were not compensated for real work, and in some instances they were not treated by co-workers as real employees.⁹

There is also some evidence that uncompensated community work experience does not provide the best conditions for participants to learn work skills — possibly because the employers, supervisors, and co-workers in these situations do not always set the same standards for an unpaid CSE worker as they have for a trainee who is on their organization’s payroll. In this relationship, no one has the economic stake in the outcome that paid employment presumes: that a worker’s productivity is equal to or greater than the money he/she receives in wages and benefits. Furthermore, community service jobs have often been characterized as "make work" — activity of little economic value designed to keep the worker occupied. Although the "make work" allegation was not substantiated in the CWEPs of the 1980s that were evaluated, the research on these programs showed that participants did not acquire new skills in CWEP jobs. In other contexts, anecdotes abound about unpaid community service and public service jobs in which participants are asked to perform tasks without meaning or value to the employer, intended mainly to keep them busy, and that supervision is inattentive or even hostile, reinforcing a perception that it does not matter to the employer whether the CSE work gets done.¹⁰

A more relaxed attitude toward employee productivity and workplace behavior on the part of employers and supervisors can sometimes be an appropriate on-the-job training approach for people who have little or no work experience if the standards for first-time workers are gradually increased to simulate the conditions of real work. This is a commonplace work adjustment strategy in vocational rehabilitation and has been adapted to welfare-to-work programs. Most often, however, when less-than-standard performance is expected from community service work participants at the beginning of their placements, less is still expected later. It is unusual for a community service work placement to be structured to encourage and reward gradually increasing capabilities of participants; when this happens, it is most often because an individual supervisor is a "natural" teacher and/or is personally interested in the success of his/her charges.¹¹

The main advantage that an expectation of "real work for real wages" brings to CSE is that a successful paid CSE placement generally constitutes a better reference on a résumé than an unpaid placement. Employers recognize the terms of the labor-for-cash relationship in work; previous paid work signals to someone looking to hire that a job seeker met a certain standard of performance because he/she was good enough to get paid.¹² There are immediate advantages to paid CSE participants as well. In this country, welfare recipients who attain the status of worker tend to experience others as being more respectful of them as individuals, feel better about themselves, gain confidence, and view their prospects with more optimism, which in turn can help land their next job. Whether the psychological benefits to participants of getting paid for their work outweigh the well-documented stresses of low-wage work is an open question.¹³

Delivering on the Promise of Real Work

To succeed in providing real work in a CSE program, administrators need to select worksite sponsors carefully and to be prepared to compensate for some sponsors’ shortcomings. At a minimum, worksite sponsors need to be able to devote enough supervisor time to the tasks of incorporating assigned CSE workers into the routines, culture, and procedures of the workplace and of eliminating any barriers to treating assigned workers like regular workers. Very thinly staffed organizations are not recommended as worksite sponsors; neither are those with a "sink or swim" culture for new employees, or those with very entrepreneurial or competitive cultures (except for the unusual few TANF recipients able to handle little structure and information and lots of discretion). The placing agency for the wage-based CSE program needs to have the capacity to provide everything the worksite sponsor cannot in the way of orientation, information, connections/mentors, etc. As noted above, another condition of "real work" is that employers or worksite sponsors (which may not be the same organization) should have the ability to hire and fire CSE workers.

Administrators also need to attend to the broader TANF context in which wage-based CSE programs will operate. If wage-based CSE is intended as a step in a progression of welfare-to-work activities and/or is reserved for TANF recipients who seem to be ready for the conditions of real work but cannot find a job, a system must be in place in the broader TANF work program for determining who is ready for "real work." How will it be determined that a recipient is ready for wage-based CSE? For example, will participants need a reference or an indication of having successfully completed another work activity? What assessment procedures will apply? What degree of accuracy is necessary in predictions of readiness for real work, and what amount of risk is acceptable?¹⁴

One approach to handling the problems of risk and unpredictability is for CSE programs to develop an array of worksites with a range of workplace expectations — some more flexible than others, some supervisors more patient than others, some kinds of work more demanding than others — and with part-time to full-time schedules and varying wage rates. Such variation is possible, especially in large communities, because "community service" is not necessarily less demanding than unsubsidized private sector employment across all type of jobs. For example, "real work" in office environments and customer service depends on employees showing up on time, being at their work stations, attending carefully to their tasks, and being able to interact with co-workers and customers pleasantly. Inventory work and a variety of back-office jobs, on the other hand, often have more flexibility in terms of when the work needs to start and end (if not in terms of accuracy). Many kinds of jobs involving outdoor work require less concentration and attention to detail.

If wage-based CSE programs are designed to accommodate a range of real work opportunities, a system for matching job candidates with positions will be needed. Because job placement has not received much attention in past community service/work experience programs, there are few lessons from this field. (In fact, workfare participants were often placed in the jobs closest to their homes or the jobs that suited their children’s school schedules.)¹⁵ The simplest and most beneficial approach for everyone involved may be for the program to make a market. As recommended above, potential worksite employers should be empowered to hire and fire participants. In a market, participants would likewise be empowered to select which employers to interview with, based on job descriptions (which the CSE program should help employers develop).

Real working conditions for CSE participants may be more difficult to achieve in public sector agencies than in nonprofit organizations because of Civil Service hiring procedures. Unless public agencies have a noncompetitive classification for temporary workers that can accommodate paid CSE workers, it will be difficult for these agencies to treat CSE participants like real workers, at least in terms of the methods needed to get them started working.

Dealing with the Consequences of Real Work

CSE participants who pass worksite employer screens but are not truly ready for the expectations and conditions of real work may miss days, be late, lose pay, receive reprimands, and get fired. Wage-based CSE programs need to have in place monitoring systems and counselors who can work with participants to try to avert the final unsuccessful outcome of real working conditions. Programs also need policies and procedures to deal with participants who do not do well in wage-based CSE assignments. Under what circumstances should they be placed in another wage-based CSE position? What is the fallback work activity for people who do not do well in wage-based CSE? Should the program do anything about income lost when participants miss work? Should there be wage-based CSE jobs with different levels of expectations (perhaps signaled by the wage rates) in order to have more options?

Real Work Versus Desirable Work

In designing wage-based CSE, it is important to take into account TANF recipients’ perceptions of the "comfort level" or desirability of different jobs and job settings. For example, entry-level work in the public sector and in nonprofit organizations is sometimes believed to be less demanding than entry-level work in for-profit companies. Before the mid-1970s, job security and fringe benefits for entry-level workers were attractive in many large manufacturing and transportation companies; then, government began to be seen as a source of more "good jobs" from the perspective of security and benefits, although the private sector continues to be seen as the best environment for opportunities to move up quickly. Similar generalizations can be found in popular culture about the working conditions and desirability of jobs in different industries — "fast food" being a frequently disparaged industry, whereas any "high tech" or computer-based industry is frequently believed to be lucrative.

For wage-based CSE, the implication of such perceptions of economic sectors and industries is that participants might have weaker or stronger incentives to move from CSE jobs into the competitive labor market depending on how their experiences of their specific CSE jobs compare with their perceptions of their opportunities in the market. There are circumstances in which wage-based CSE might be more attractive in the short run than the unsubsidized job a TANF recipient might be able to land at a particular point in time. In addition to job security, fringe benefits, and advancement opportunities, like other workers, CSE participants may be equally or even more concerned about work hours, supervision, location, job strenuousness or stress, and the social environment of a job. In fact, to some recipients as to some other workers, these may matter more than the job’s auspices and more than the pay.

As a result, operators of wage-based CSE programs may need to be creative in adjusting incentives (and reducing disincentives) among the varieties of work available locally — for example, by offering cash bonuses to participants who take regular jobs, perhaps varying the bonus according to the type of job, and by working with employers in high-turnover industries to make their jobs more attractive to TANF participants. Also, without some entry criteria for wage-based CSE — such as having failed the job search test — or fixed-length CSE work assignments, participants who are able to find regular jobs but perceive CSE jobs as more desirable might resist moving on to the competitive labor market when they are ready. (The New Hope initiative in Milwaukee set a policy of allowing participants to work in paid community service jobs for six months; a participant could be placed in another community service job for a second six-month stint during the demonstration if she/he was unable to find unsubsidized work.)¹⁶

V. Wage-Based CSE as Job Skills Training and/or a Stepping-Stone to Unsubsidized Employment

Evidence of the Potential

The Community Work Experience Programs (CWEPs) of the 1980s were clearly not either a route out of welfare or a route out of poverty, and there is little evidence from past unpaid work experience programs for welfare recipients that transferable occupation-specific skills were taught.¹⁷ But in the wider field of publicly funded jobs, there is evidence that supports the prospect of wage-based CSE serving as a transitional step for TANF recipients into unsubsidized employment. Cliff Johnson and Ana Carricchi Lopez reviewed the record of public job creation initiatives and found many examples from the 1970s and 1980s, and even before, that reduced unemployment and increased the income of those who participated, both while they took part in the initiatives and subsequently.¹⁸ Also, a significant proportion of the people who were provided first jobs through publicly funded jobs programs like CETA went on to work regularly in other jobs.¹⁹

The minimal regulatory direction in PRWORA for job programs for TANF recipients provides new opportunities for CSE to take advantage of the best concepts in employment and training for disadvantaged workers, as implemented by the most successful employment and training organizations. Thus, wage-based CSE could be structured to provide occupation-specific skills training based on publicly financed wage subsidy models, such as the AFDC Homemaker-Home Health Aide Demonstrations operated between 1983 and 1986 to provide "job ready" welfare recipients four to eight weeks of training and up to 12 months of subsidized employment.²⁰ Alternatively, it could be structured like apprenticeship models in the private sector, such as those used for centuries to train new workers in the skilled trades. Another model is the combination public-private approach, such as America Works, which uses an intermediary job training organization to screen, recruit, train, and place workers in private sector jobs.

Work that is performed in crews is particularly amenable to worksite teaching and management approaches, including work in construction, building maintenance, parks maintenance, and other unskilled or semi-skilled labor. YouthBuild uses such an approach for building or renovating affordable housing in low-income communities and also provides education and leadership development activities for the young adult participants (most of whom are high school dropouts) in order to increase their employability and help the young people see and appreciate the value of sustained effort as they learn specific skills and provide tangible benefits to communities.²¹ The New York City Parks Career Training (PACT) program is a rare example of a community work experience program for welfare recipients that did teach job skills (although not uniformly).²²

Providing Occupation-Specific Job Skills Training in Wage-based CSE

To build occupational skills training into wage-based CSE, program designers first need to know something about the range of career aspirations of the whole TANF population early in the program implementation period. This can be accomplished in a superficial way through a survey, in a somewhat more complete way through one-on-one counseling and assessment sessions, or, most comprehensively and accurately, during the course of career education classes in which potential trainees are exposed to information about types of work and careers they may not be familiar with or have not considered for themselves. Other essential elements of a wage-based CSE program that offers occupational skills training are:

• Agreements between the worksite sponsors and the placing agency for the CSE program about the job skills that participants will be taught, along with appropriate competency measures and time frames, frequent benchmarks, and a system for both the sponsoring agency and the placing agency to monitor participants’ progress;²³

• Agreements about who (which agencies and individuals within the agencies) will be responsible for teaching, monitoring, and supervising participants;

• Contingency plans and agreements for when the program is not working for an agency or a participant; and

• Enough job developers for the CSE program to do the work of defining and describing jobs at worksite locations, including the skills participants will learn, benchmarks for their progress, and measures of their skill achievement. (If these tasks are left up to worksite sponsors or employers, either the program start-up will be delayed or the documentation will often be incomplete.)

Additional incentives — beyond the free labor — for worksite sponsors to take on the job skills training role in wage-based CSE are useful, but not always essential. For some worksite sponsors, payments in recognition of their increased supervision, mentoring, teaching, and other efforts will matter. For others, it will be more helpful to have an on-site supervisor from the program. Extra outside training (such as driver education) or other supplies and equipment for the participants that will make them more effective employees will be most appreciated by other employers.

Tryout Employment

As work program options are redefined under TANF, there is no reason why a wage-based CSE job could not be used as a tryout for a "competitive" job either with the worksite employer or with other employers in the same field. For example, even though CSE (or CWEP) in the welfare context has not been based on the expectation that participants will be hired into regular jobs with their worksite organizations at the end of their assignments, this does occasionally occur because the community service/unpaid work experience situation offers employers the opportunity to look over a worker for several weeks to several months without risk — not unlike the circumstances in the temporary worker industry. If the employer likes the participant’s performance and a position becomes available, the community service/work experience assignment might result in a regular job. The chances that a temporary, fixed-duration assignment in wage-based CSE will turn into a permanent job can be increased if the program:

• Helps worksite employers create positions for wage-based CSE that have not previously been performed as a single job. The CSE participants occupying those positions thus will be uniquely qualified for any such permanent positions created. Another good strategy is to create positions that are designed to showcase the particular qualities or skills needed for another class of jobs in which there is regular turnover.
  For example, the competition in the health care industry is providing many opportunities for new functions that improve customer service and response. CSE "patient greeters" who are outgoing, relaxed, and personable might create a demand for their service and thus create their own permanent jobs. In the same hospital environment, many clerical and maintenance jobs require minimal qualifications but demand reliability and attention to detail. A CSE participant who demonstrates these qualities during a different type of assignment might be next in line when a permanent job opens up.

• Keeps in touch with worksite supervisors about participants’ work and work habits — if employers agree to this — in order to intervene early in problem situations and avert dismissals.
  Participants who are not working out in one wage-based CSE job can be placed in another one before a problem situation ends in a firing, which can be a setback for participants (as well as a useful learning experience) and troublesome for the worksite. In a "real work" situation, this type of intervention would not necessarily occur, but worksite employers should be sold on the CSE program’s temp-agency-like customer service for employers.

• Works with participants and worksite supervisors during CSE assignments to develop résumés, references, and work samples for participants that they can use later to market themselves for permanent jobs.
  Trying to recollect and accurately describe a past job is more difficult than doing this for a current job. Similarly, it is easier for a job candidate’s references to write letters describing the candidate’s performance and skills when the experience is fresh. Participants in wage-based CSE particularly need to do a good job translating their job experience into a usable résumé and references if they have spotty work histories, little work experience, or no recent work experience and recent references.
  One way to structure this task is to have participants keep a log or journal of their work activities with daily entries on what they did, comments on new skills learned and new tasks performed, feedback from co-workers and supervisors, and interactions with co-workers and supervisors. These journals can be used to construct a description of the wage-based CSE job for a résumé while the participant is still working, which can then be checked with the supervisor to make sure that it fits his or her understanding of the job. One or more letters of reference can be developed in the same way during the participant’s time on the job and checked with supervisors.

Creating Supports During Transitional Jobs

A key benefit of transitional jobs is the opportunity they provide for people who have not been in the labor force to become accustomed to the routine and discipline of getting to work every day. This includes planning for and handling the situations in their lives that make it difficult to be reliable, such as health issues, children’s needs and schedules, transportation, adult household members who are demanding, concerns about personal safety, and lack of pocket money. Welfare recipients who have little work experience or who have been out of the workforce for a long time are likely to need at least a six-month transitional job. This is long enough for their child care and/or transportation arrangements to fall apart at least once and, ideally, to be repaired in a way that will support permanent employment. This will also give them time to sort out other living arrangements and reorder their lives for work.

During the transitional period, it is helpful for participants to have resources for problem solving and psychological support for change. Groups of participants going through the same adjustments serve this purpose well because they are able to give advice based on firsthand knowledge of each other’s situations, but such groups must be set up in a way that does not add to the demands and stress of work schedules for CSE or reduce participants’ earnings. The best arrangement is a designated time set aside during regular work hours (during lunch breaks, for example) at a convenient location with a group facilitator who can direct the discussion lightly but can let the participants problem-solve together as much as possible.

VI. Can Wage-Based CSE Serve as the New Safety Net?

PRWORA leaves very little income security in place for low-income families whose adult members cannot or will not work enough or well enough to support themselves and their children. While disagreements will continue about whether a safety net of cash assistance is an inducement to sloth or a basic protection for children, in our system of shared public responsibilities, state and local governments inevitably bear at least some of the burden of dismantled federal social programs.²⁴ Thus, an important reason for states and localities to adopt wage-based CSE might be to avoid the fiscal and social consequences of family impoverishment, such as increased homelessness, increased need for emergency shelter and foster care, and other emergency services. Wage-based CSE can be a partial solution to the lost safety net.

Wage-based CSE can be designed to accommodate TANF recipients who use up their lifetime allotment of TANF cash assistance without finding family-supporting work and to provide a fallback for TANF recipients who cannot find any unsubsidized job at the point in their TANF careers when they are required to be working (or looking for work — that is, after 24 months of assistance). It can also be a second part-time job, to supplement a part-time unsubsidized job, when TANF recipients are supposed to be working but cannot make ends meet. Wage-based CSE can pay in wages at least the equivalent of the TANF grant — and may garner considerably more family income via the federal Earned Income Credit — keeping money for basic living expenses flowing to the family households that might otherwise be sanctioned or have their assistance cut off entirely.

Income Issues

One of the most attractive features of community service work — if it is properly structured to be "real jobs for real wages" — is the opportunity it provides for participants to increase their income compared to welfare even if they are paid only at the minimum wage level. This is because low-earning employees are eligible for the federal Earned Income Credit (EIC; also called the Earned Income Tax Credit, or EITC), which can effectively increase their income by more than a third. A family with one child receives a credit equal to 34 percent of the first $6,500 of earnings, and a family with two or more children receives 40 percent of the first $9,140 of earnings.

EIC eligibility does not depend on a worker’s earning enough to owe taxes. It is a payment similar to a negative income tax which is paid to the worker either by the federal treasury at the end of the tax year or by the employer as part of the worker’s regular paycheck. (Many employers are not familiar with the version of the EIC that enables a participant to have a projected EIC added to his or her paycheck, although the procedure is relatively straightforward. Information about how to do this and the forms required may be obtained from the Internal Revenue Service: Circular E, Employer’s Tax Guide, Publication 15, Revised January 1998.)

Because the current federal law and regulation on the status of wage-based CSE in TANF with respect to the federal EIC is complicated and has not been tested, TANF program administrators should be certain they are up to date on interpretations of the law before proceeding with a plan that depends on trying to qualify participants for the EIC. In addition to the federal government agencies involved and the Center for Law and Social Policy, affinity groups and professional associations — such as the American Public Human Services Association,²⁵ the National Governors’ Association, the National Conference of State Legislatures, and the National Association of Counties — are usually good sources of the status of legislative interpretation as it affects programs like TANF.

Finally, states and localities can enhance the income of wage-based CSE program participants through their decisions about wage rates and work hours, although tradeoffs are sometimes involved. For example:

• If the policy aim of wage-based CSE is to replace TANF income, most states will get close to this by providing 25-hour-per-week jobs at minimum wage, assuming that participants will receive the EIC as an income supplement. If states or localities aim to increase family income compared to TANF assistance, however — for example, to achieve approximate parity with the working poor — wage rates and/or hours will usually have to be increased.

• Maximizing the income of each participant will require that most states and localities keep their wage-based CSE programs small. However, if the policy objective for wage-based CSE is to maximize the number of participants who can be placed in these jobs given a level of program funding — that is, to spread the funding across as many people as possible — wage rates and hours of work must be set at the minimum.²⁶
  A useful reality check on policy preferences in this area is to estimate the percentage of TANF household heads who will come within 12 months of exhausting their lifetime assistance in each of the next two to three years and then compare the resulting number of TANF cases with the number that can be provided wage-based CSE jobs under two scenarios: (1) maximizing the number of participants or (2) maximizing the income per participants.

• States and localities will certainly need to compensate worksite employers for their share of the cost of FICA for participants in wage-based CSE in order to induce the employers to provide the jobs. (This amount must be figured into the total per participant cost, which will determine the potential scale of the program.) It may make sense as an income enhancer for states and localities to compensate participants for their share of FICA, too, particularly if the required amount of work yields no more than TANF assistance.

• Within the parameters of the TANF requirements for work hours per week, work hours of participants in wage-based CSE jobs can be set to achieve several different objectives, including family needs for income — for example, based on family size. If the CSE program is TANF-funded, ensuring that participants work enough to be counted toward TANF participation goals (25 hours per week beginning in FY 1999) might be important to states and localities. In some cases, worksite employers will have specific needs for work hours. The wage-based CSE program might adopt a case-by-case approach in order to provide the amount of work that each participant wants or can handle.

• Wage rates might vary across CSE jobs as well — according to the jobs to be performed, for example. (Uniform wage rates for all participants would make wage-based CSE more like a program; leaving the decision to employers would make wage-based CSE more like "real work.") Another "real work" approach would be to create a wage structure for CSE that provided incentives for performance.

VII. Can Wage-Based CSE Benefit Low-Income Communities?

Income and services can flow into poor communities as a result of higher rates of employment among the people who live there. However, there is a long record of disappointment in policies designed to achieve this purpose through government incentives to business and industry.²⁷ Further, in the field of community economic development, there have been few attempts to implement geographically targeted employment strategies. Consequently, the potential of wage-based CSE as an economic development strategy for a neighborhood is virtually untested. Theoretically, wage-based CSE could work as a community investment mechanism by delivering: (1) earnings at least equal to but probably greater than welfare benefits that could go to neighborhood families and then into the neighborhood economy and (2) services that would not otherwise be available for the benefit of neighborhood residents if CSE jobs were located in the neighborhood.

Stimulating the economy during periods of recession has long been the rationale for federal job creation programs. For wage-based CSE, these attempts offer a "good news/bad news" story. While countercyclical job creation efforts have most often failed to boost the economy (because new government-funded jobs have usually arrived too late to influence the business cycle), the in-program effects and post-program effects on the income of participants — economically disadvantaged individuals — and their work careers have often been substantial.²⁸ Poor neighborhoods have fared less well as a result of policies to stimulate economic activity.

Given the experimental nature of community economic development via wage-based CSE programs, these programs might productively partner with community-based organizations or coalitions of community-based groups that have already developed their agendas for community improvement but lack the resources to take on more than one activity at a time or to devote enough time to any single activity to keep it moving along rapidly.²⁹ Some community-based organizations and coalitions that experienced major reductions in local government service contracts during the last few years are still operating, but at a much reduced level of service, in neighborhoods where there is a large gap between service needs and availability. A targeted strategy for wage-based CSE might create jobs in these organizations.

Another strategy for using wage-based CSE to generate community benefits is aimed at keeping income that is derived from job creation within the community. There is a research literature that indicates that the money available to poor people in poor neighborhoods does not stay long in those neighborhoods but quickly gets into the control of businesses, corporations, and individuals who spend it elsewhere.³⁰ More recently, however, community economic development strategies have tried to keep the assets and resources of community residents in the community — for example, through development of neighborhood credit institutions, "buy local" campaigns, consumer cooperatives, and tenant-owned and -managed housing. Wage-based CSE programs might target their job creation efforts to community organizing and community enterprises that serve these purposes.

Administrators of wage-based CSE programs should be alert to the consequences of success in this economic arena: If wage-based CSE becomes a valuable tool for community investment, local politics may also become a central concern for the program. When community service was defined as whatever jobs might be performed by a public or nonprofit agency — as it was under the rules for operating community work experience for AFDC recipients from the early 1970s until the enactment of PRWORA — policymakers and program administrators did not have to examine whether any particular CWEP assignment was a real service to the community, with the attendant questions: Which community? Who benefits from the service? Who decides whether the service is needed and/or a priority? Now, however, wage-based CSE might emerge as a valuable resource for neighborhoods, and thus a potential focus of political competition. One approach to managing political issues about the allocation of wage-based CSE is for the program to work through many neighborhood organizations and coalitions to determine good uses for community service workers, keeping in mind that the best strategy for allocation from the program perspective may be to have many types of jobs and worksite sponsors to choose from, either neighborhood by neighborhood or across a locality.

VIII. Two Program Examples

New Hope

A recent example cited by Johnson and Carricchi Lopez is the New Hope initiative in Milwaukee, Wisconsin, for which new outcome data suggest that paid community service can be a stepping-stone to unsubsidized employment for some. Based on an evaluation of New Hope participation and outcomes — not impacts, which compare the program-eligible group’s outcomes to a control group’s — 43 percent of community service jobholders in New Hope moved from unemployment or unsteady employment (before enrolling in New Hope) into unsubsidized work after their community service placements.³¹ Measures of job-keeping for all the New Hope participants who used community service jobs indicate that about 30 percent worked 30 hours or more in unsubsidized jobs in most of the 12 months after they left community service jobs; about 30 percent had a short tenure in their unsubsidized jobs following community service (working in 2˝ to 7 of the next 12 months), and about 40 percent either did not get a foothold or did not stick in unsubsidized employment (working less than 2˝ months during the year after their community service).³²

New Hope also offers instructive lessons about the role that community service jobs might play in welfare reform. It was designed principally as an incentive-driven program to test the notion that work and household income in targeted poor neighborhoods could be increased by offering residents who worked at least 30 hours per week an earnings supplement to bring their household income at least to poverty level along with income-based subsidies for health insurance and child care. In other words, New Hope was envisioned as one way to "make work pay" for a range of low-income workers — not the limited population of welfare recipients. In the New Hope package of benefits, community service jobs were intended as a fallback for people who were not working when they enrolled and could not find jobs on their own or who did not succeed in getting a job during an initial eight-week job search. (Community service jobs in New Hope were also intended to give part-time workers an opportunity to "top up" their work hours in order to meet the 30-hour-per-week standard for qualifying for earnings supplements.)

During the first 12 months of program operation, about one-quarter of the New Hope participants worked at community service jobs at some point. (New Hope’s community service job placements were limited to no more than six months, and each participant was allowed to do two community service stints. These jobs paid minimum wage.) Compared to TANF recipients, New Hope was aimed at a relatively advantaged group of low-income people — those who were already working as well as those who wanted to work and/or thought they could work — and the local economy was very strong. The fact that about 25 percent of program participants could not work enough on their own to meet the 30-hour-per-week standard under these favorable conditions should be a signal to state and local TANF administrators to expect a relatively high interest in and demand for wage-based CSE.³³

PACT

The New York City Parks Career Training (PACT) program offers a recent promising example of a part-paid/part-unpaid community service program designed to move participants into unsubsidized jobs. In the case of PACT, the participants were adults receiving the New York equivalent of General Assistance, called Home Relief (HR), a population that is predominantly male and predominantly single living without children. PACT’s on-the-job training was both intensive and relatively long term — 35 hours per week for up to 12 months (but usually 10 months) — and included paid seasonal jobs with the Parks Department that included office work, skilled trades, horticulture, park maintenance, vehicle maintenance, security, and custodial work. These jobs paid $8.40 per hour. PACT’s outcomes were similar to those for New Hope: About 40 percent found regular full-time jobs while participating in the program, and 77 percent of those who went to work kept their jobs for at least 90 days.³⁴

The PACT experience confirms a high degree of interest among its target group of very low-income single adults in the prospect of real training that culminates in a real job. As noted above, although the required work hours for PACT were double those of the participants’ workfare obligation in New York City at the time the PACT program got under way in 1994, two-thirds of the HR recipients who attended PACT recruitment presentations for the "class" that started in the fall of 1994 volunteered for the program. The opportunity to receive the types of job training offered by PACT and hopes that this would lead to employment were the most common reasons reported by participants for their choice of PACT over the mandatory workfare alternative, called WEP, which required half the work hours but offered no training, job counseling, job placement services, or tryout employment in real jobs that paid real wages. PACT’s program retention rates are equally impressive. More than 60 percent of participants stayed in PACT for the full 10 months, or left because they had found a job; only 15 percent left voluntarily without a job.³⁵

Participation in PACT was a two-way street. Not all HR recipients who were obligated to participate in workfare volunteered for PACT, and not all the volunteers for PACT were selected for the program. The Parks Department rejected about one-third of the PACT applicants — most frequently, those who had less than three years’ work experience, but also many who had received public assistance for five years or more, had only fair or poor English proficiency, or had a felony conviction. Thus, PACT worked with a motivated group whose members were relatively employable compared with the whole HR population; compared with the general population, however, this was a disadvantaged group: predominantly middle-aged, 84 percent minority, 40 percent lacking a high school diploma or GED credential, 20 percent with criminal records, and 40 percent with less than five years of full-time work experience.

During their work/training experience, about half the PACT participants worked in crews of eight to 10 persons supervised by a Parks Department worker, and the other half were individually assigned to jobs in the department. With few exceptions, PACT participants viewed the program as real, challenging work and a valuable opportunity, and most appreciated the structure and predictability of the 35-hour week (remembering that Home Relief is provided to indigent people who do not have children living with them and are, thus, less likely to appreciate flexibility in a job as much as working parents).

IX. Funding Wage-Based CSE: Should States and Localities Use TANF Monies?³⁶

The total amount of funding that supports welfare-to-work activities is greater by far than the federal monies provided to states for the TANF program. Other sources include the state and local funds required as a maintenance-of-effort level for the TANF program. Beyond these federal and state TANF funds, states and localities, Private Industry Councils, and nonprofit organizations can receive competitive Welfare-to-Work grants authorized by the Balanced Budget Act of 1997 and administered by the U.S. Department of Labor. States and localities may also choose to spend public funds that they control on welfare-to-work programs, which may include amounts budgeted for employment and job training services, vocational education, vocational rehabilitation, and adult education. Some states and localities also have access to private funding to support some components of their welfare-to-work activities.

With multiple sources of funding going into all efforts to move welfare recipients into the workforce, states and localities have some choices about how to allocate and account for funding for various welfare-to-work purposes. Two major choices that affect wage-based CSE are (1) whether wage-based CSE is financed under TANF, federal Welfare-to-Work grants, or some other funding arrangement and (2) whether states commingle federal TANF funds and the state maintenance-of-effort funds — as opposed to segregating them while operating a single program or operating a separate program with the state funds.

The whole financing picture for the workforce development in a jurisdiction should be a major consideration in deciding on an appropriate answer to the question "Should states and localities use TANF funding for wage-based CSE?" State and local planners should ask: For every type of welfare-to-work activity planned or desired, what is the most suitable funding stream? "Suitable" might mean "most flexible" when dealing with the hardest-to-serve segment of the TANF recipient group, for example.³⁷ Alternatively, if wage-based CSE is intended for people who can work but cannot find a job because of spotty work histories, lack of experience, or lack of references, the most suitable funding stream might be the one that enables the jurisdiction to finance wage-based CSE on as large a scale as necessary to create a transition for a relatively employable group.

The federal Welfare-to-Work grants are targeted to hard-to-serve TANF recipients and the noncustodial parents of children in TANF families. Allowable activities specifically include "the conduct and administration of community service or work experience programs" and "job creation through public or private sector wage subsidies."³⁸ This funding source may appear to be a better fit than TANF for wage-based CSE, but decisions about the use of Welfare-to-Work funds will be coordinated at the local level, whereas in many states, decisions about the use of TANF funds will be made at the state level.

The advantages and disadvantages of commingling federal TANF and state maintenance-of-effort funds for the purpose of operating wage-based CSE are not clear-cut. Rather, there are tradeoffs. If TANF and maintenance-of-effort funds are mixed, then TANF rules apply to all the activities funded by the total and to all persons served in these activities. Of particular importance is the TANF prohibition on providing assistance to families who have already received assistance for 60 months (or the state-established lifetime limit). If funds are segregated, some of the TANF rules apply to the segregated state funds, including work and participation requirements. However, if states choose to operate a separate state program either with their maintenance-of-effort funds for TANF or with other funds, they do not have to abide by TANF rules (but neither can they count participants in these programs toward the TANF participation goals).

In most cases, decisions by states and localities about how to finance wage-based CSE interact with decisions about whom it will be designed to serve. For example, wage-based CSE programs that operate in depressed local economies where jobs are scarce might use TANF as the principal funder because, in these areas, TANF recipients are unlikely to be able to get the part-time, unsubsidized jobs that those living in better local economic conditions would take to satisfy their work requirements. Thus, the number of people for whom wage-based CSE would be a good choice might be large enough to significantly help the jurisdiction meet its TANF participation goals (if these goals are passed through from states to local jurisdictions without adjustments for economic conditions). At the same time, the jurisdiction would be providing a measure of income security for people who are unemployed through no fault of their own.

X. Program Costs

In the 1980s, unpaid community work experience programs for AFDC recipients were attractive to many jurisdictions because they were a relatively inexpensive way to get people working. Where unemployment rates were high and jobs for low-skilled people scarce, CWEP was a form of inexpensive job creation that also served to reinforce the value of work. Where the "social contract" idea was the paramount policy agenda,³⁹ CWEP was a valued program because it could be operated on a larger scale than education and job training and it could keep participants involved in work longer than job search activities — again, for relatively low cost per person (less than $1,000 per person, and usually much less, in 1985 dollars).⁴⁰

It is important to note, however, that these low costs for CWEP cannot be directly compared to present-day costs for a wage-based CSE program. During the many years that AFDC cash assistance was an entitlement with open-ended expenditures by the federal and state governments (for eligible individuals and their families), funding for work programs for AFDC recipients was a separate, close-ended appropriation by Congress with varying requirements for state matching funds. Planning and accounting for cash assistance and work program expenditures at the state and local levels were also separate and thus the AFDC cash assistance and Food Stamps and related expenditures for people participating in CWEP were not considered program costs along with administration, child care, and transportation — even though, in one sense, AFDC was the source from which CWEP participants were "paid" for their work; or, from another perspective, AFDC was the reason that CWEP participants did work without pay.

Now, with TANF block grants that combine multiple forms of assistance to indigent families, including cash and work program expenditures, wages in a CSE program must be considered "on the same side of the ledger" as cash assistance in accounting for program costs. Thus, a current comparison of CWEP and wage-based CSE costs would add the wages paid to CSE participants (and related benefits) to program administration costs and compare the result with the sum of cash assistance (and related benefits) for unpaid CWEP participants and the administrative costs for that program.

Even with such a revised accounting perspective, however, wage-based CSE, as described in this paper, is bound to be more expensive than CWEP as it was run in the 1980s, when most benefit-cost evaluations of this model were conducted. This is because those CWEPs generally were minimally staffed. For example, the task of developing worksites and worksite agreements tended to be labor-intensive during the planning and program start-up period, but then it slacked off. Assigning eligible participants to worksites was often a perfunctory task based on simple rules of thumb requiring little interaction among participants, staff, and worksite personnel. Finally, tracking attendance and compliance and monitoring worksites were often handled on an exception basis; rather than a proactive or preventive approach, reactions to problems and responses to documented noncompliance were considerably lagged in order to keep staff costs down.

In contrast, most of the wage-based CSE options described here call for proactive, continuous staff involvement in order to ensure that this work program meets ambitious goals for preparing TANF recipients for work in a policy environment where the cost of failure is high — particularly for those TANF recipients who run out of assistance before they are able to support their families. The final section of this paper deals with how the various functions required to operate wage-based CSE might be divided up among public and private organizations, but the program suggested here requires, at the local level:

• job developers to work with all participants;

• staff (perhaps job developers) to develop a range of worksites, including job descriptions for the CSE positions at the worksites and special agreements for training worksites that specify the job skills for training, benchmarks for performance, and measures of competency;

• computer systems development staff to design procedures and technology that enable CSE program staff and worksite supervisors to monitor and report on the attendance and earnings of participants; and

• computer systems development staff to establish protocols for identifying appropriate TANF recipients for referral to wage-based CSE based on the target group criteria selected for the program, if this capability is not already available.⁴¹

XI. Options for Administration: Who Should Run Wage-Based CSE?

With rare exceptions, the assumptions of the 1970s and early 1980s about publicly funded programs were that they would be administered by the employees of public agencies and that contractors to these agencies might be engaged, generally on a competitive basis, to deliver services, either to the agency itself or to the persons and organizations that the agency was intended to serve. The administrative paradigm of the past suggested a clear division between public responsibilities and areas in which competition among service providers might yield lower costs, efficiencies, innovations, and better performance. While there is some academic and considerable political debate about when the paradigm shift occurred, whom to attribute it to, and whether it is a good or bad development for the governmental "customer," the 1990s assumptions are not only that there are alternatives to having public services delivered by public employees but also that there are administrative alternatives.⁴² Thus, there are many possible answers to the question: Who should run wage-based CSE under TANF?

One reason to ask the question is that potential worksite employers of CSE participants may be more willing to provide work opportunities for the program if they do not have to assume responsibilities beyond those that would be usual for a new, temporary employee. Better still would be if they do not have to assume even these usual employer responsibilities. Another reason to ask the question is that many local public agencies are downsizing and looking for ways of lowering the cost of public services and increasing their productivity in a "smaller government" movement that has led to numerous experiments in "outsourcing" and other arrangements for private entities to provide public services. It is no longer assumed that a wage-based CSE program can be run only by a government agency, nor that it can be run best by a government agency.

Potential Performers of Program Functions in Wage-based CSE

Not only do the ground rules for welfare-to-work programs no longer dictate how responsibilities for the various funding, administration, direct service, and other aspects of such programs should be allocated, but there are a number of new potential players as well. Under the ground rules of the 1980s, CWEP functions were primarily divided between local welfare agencies and public or nonprofit agency worksite sponsors. Under JTPA, responsibilities for subsidized employment and OJT for disadvantaged workers were primarily divided among the public agencies that were the local Service Delivery Area (SDA) administrative entities, the SDA contractors who provided client assessment and training, and private employers.⁴³ These arrangements began to diversify in the early 1990s as government was "reinvented." In particular, for-profit companies were invited to train, place, and support disadvantaged workers under contracts with public agencies.⁴⁴ Now, a much larger number of public, nonprofit and for-profit organizations might be involved in wage-based CSE, including, but not limited to:

Public agencies

• State welfare agency

• State workforce agency

• State education agency

• State vocational rehabilitation agency

• Local welfare agency

• Local PIC/JTPA/WIA agency

• Local employment service office

• Local vocational rehabilitation office

• Local provider of counseling or treatment services

• Community college/adult school

Private entities

• Individual employers

• Chamber of Commerce/business association/staffed employer group

• Labor market/workforce intermediary (for-profit or nonprofit)

• Business service company (for payroll services, temp support, and the like)

• Individual provider of professional services (such as specialized testing and assessment, therapy, computer network design, and procedures manual writing)

• Nonprofit provider of counseling or treatment services

• Community-based organization (CBO)/incorporated neighborhood group

• Job training provider (nonprofit or for-profit)

• Educational institution

• Labor union

Which Organizations Are Best Suited for Which Functions?

The Full Range of Options

The guidance offered here to program designers thinking about which organizations and institutions might be best suited to run a wage-based CSE program, or be a partner in it, is based on the assumptions that there are many ways to build a program out of local public and private organizations and that TANF allows states and localities considerable discretion in deciding what will work best for their conditions. Table 1 arrays program functions in a wage-based CSE program against a few of the potential institutional players in order to demonstrate the wide range of options for building a program function by function. In this table, a "Yes" at the intersection of an organizational type and a program function means that either past practice or the current capabilities of such organizations suggests that they could perform the function — and that there is no apparent legal restriction or inherent conflict in such a choice. In this scheme, four of the six organizational types listed as column headings in the table are theoretically able to perform a majority of program functions. In other words, putting aside for the moment local conditions and local history, these organizational types offer many possibilities for structuring a wage-based CSE program.

The designation "Maybe" in Table 1 means that the performance of the program function by the type of organization indicated has not been common in the past and/or that expertise in this functional area is not common now, but there is no legal or other apparent barrier to the development of expertise and capability by the type of organization. For example, although it has not been common for labor unions to provide employment and training services to nonmembers, it is possible for these organizations to develop the expertise and capacity to work with CSE participants both before they are placed and while they are working. (In fact, some unions are trying to organize workfare laborers.⁴⁵)

"No" appears only a few times in the table, most prominently to indicate that "Business Services Companies" generally have capabilities to perform only a few — although critically important — functions in wage-based CSE programs. "No" in other cases suggests that political accountability can be assigned only to an entity that is wholly or partly public and that "assuring rights of CSE workers" is an inherent conflict for employers/employer organizations.

Options Based on Factors Affecting Program Performance

The apparently wide-open possibilities for structuring a wage-based CSE program suggested by Table 1 must be considered in light of factors or tendencies that might be expected to affect program performance across localities. (Local history and present conditions are also suggestive of real options; these are discussed in the next section.) For example, organizational specialization is a major limitation on possible structures for a wage-based CSE program: A provider of counseling or treatment services might be an essential partner in preparing participants for work and supporting them during their CSE experiences, but such a organization would not usually be qualified to plan, design, or lead a wage-based CSE program. Similarly, business services companies have their own market niches so that a company that designs computer networking applications would not usually be able to provide payroll services. (Both organizations might be able to train and employ one or more CSE participants, however, and partners in operating the program should not be overlooked as sources of wage-based CSE jobs.)

Program performance considerations include economies of scale, which might matter, for example, when a large number of participants in wage-based CSE will qualify for the Earned Income Credit (EIC). In this case, contracting with a private payroll services company accustomed to handling employee groups with large numbers of low-income workers might be a better solution than trying to teach each worksite employer’s payroll staff how to do advance payments of the EIC. (Tradeoffs are involved in this choice, however, because the "real work" quality of CSE is diminished if participants receive their pay in a different form or on a different schedule from their co-workers.)

When personal relationships, interpersonal skills, or cultural competency can make a difference in the performance of a wage-based CSE program, such as in persuading worksite employers or CSE participants to take risks in the employment situation, it may be important to choose an organizational partner based on the specific individuals who will perform the persuasive functions or based on relationships of familiarity and trust. Community-based organizations are particularly important intermediaries if people from minority communities are being asked in a wage-based CSE program to cross geographic, ethnic, and cultural borders in order to participate and succeed.

When expertise matters, specialized or niche organizations that have performed a specific task over and over again might be selected. For example, it saves time and money to know what the usual glitches are in creating a computer network linking CSE employers to a central station for collecting time sheets and, from the data, developing specifications for a payroll services company to issue paychecks, action reports for worksite employers and CSE managers, and notices about pay situations for CSE participants.

Similarly, when efficiency and/or speed matters, as in the process of reassigning or re-placing dismissed CSE workers or getting them out interviewing with employers again immediately, it may be important to work with an organization, such as a labor market intermediary, that is accustomed to assessing and adjusting its work priorities every day. When consistency and/or reliability matter, large public bureaucracies are often a good choice, because they are best when establishing and following a set procedure. This could be important in all areas where the workers’ rights and protections are at issue, or where eligibility for work-related benefits, such as support services, needs to be determined.

These program performance considerations result in a smaller number of possibilities for the organizational structure of wage-based CSE programs. In fact, they point to a basic form for the program, which may have several variations, rather than a very large number of selections. The analysis presented in Table 1 suggests that no single type of organization is best suited to perform all the functions for a wage-based CSE program. Table 1 also suggests a basic form of the program, which would include at least five of the six organizational types listed: a public or public/private entity for political accountability; one or more organizations capable of delivering employment and training services to CSE participants (drawn from labor market/workforce intermediaries, CBOs or employer organizations); CBOs for community-sensitive design and promotion; worksite employers; business service companies to handle some of the complicated financial tasks. One variation on the basic form of wage-based CSE is regional: Several labor market intermediaries and/or several constellations of CBOs might work different territories of a county or state in a program structure that was headed by a public or public/private entity and serviced by one or more business services companies.

Function-by-Function Choices of Organizations

Following are discussions of each function for operating a wage-based CSE program that present the justification for the "Yes," "No," and "Maybe" entries in Table 1 and illustrate how local history and conditions might affect the choice of organizational partners.

            a. Overall coordination, ultimate accountability, and public visibility. While there are many examples of privatization in the delivery of publicly financed services, it is difficult for public agencies to delegate all responsibility and accountability to private entities. Typically under state and federal laws, public agencies must establish a contractual relationship with private entities in order to delegate public functions and transfer public monies to them. Then, the public agencies must oversee the financial and technical performance of the contractor and take steps to correct or remove the contractor for poor performance or excessive spending. In the welfare field, since TANF replaced AFDC, most of the limits on what can be done in operating the program derive from state government (with the important exception of the federal government’s determination that both wage-based CSE and programs that provide benefits in exchange for work which operate under TANF will be subject to the Fair Labor Standards Act and other national labor laws). Thus, TANF funds block-granted by the federal government to a state might theoretically be transferred to a private entity to operate a wage-based CSE program in a way that also delegates almost all coordination, accountability, and public visibility functions to that entity.

In most states, ultimate accountability for the expenditure of public funds still must reside with a public agency. However, overall coordination and the primary public visibility functions of a wage-based CSE program (or TANF as a whole) might be taken on by a private entity. History and local politics figure largely in the benefits of such an arrangement. If it is important to create a sense of the new among stakeholders — a departure from the past — especially among potential employers for the CSE participants, then a new organization formed for just this purpose might be the best choice. In localities where a previous "workfare" (work-in-exchange-for-benefits) program met opposition from a substantial segment of nonprofit employers, a sense of the new and a new organization at the head of wage-based CSE might be the best choice to get maximum participation and cooperation from potential worksite sponsors and service providers. On the other hand, if a known, reliable, trusted agency matters most to the CSE stakeholders, recreating institutional relationships of the past to operate wage-based CSE might be the best local answer.

In a locality where there is no recent experience with workfare or anything like CSE, then "stance," skills, expertise, and capacity may the be most important criteria for choosing who should coordinate and be the public face of a new wage-based CSE program. The coordinating entity needs to have a relatively neutral stance in the locality in order to successfully acquire support and get timely responses from all the stakeholders: potential CSE worksite sponsors and those who sign up and employ participants; TANF recipients and their representatives; public agency managers and workers and the workers’ union representatives; low-income community leaders, including CBO managers and staff; local politicians; providers of education, employment, and child care services for TANF recipients; business service participants in the program (who provide payroll, messenger, information technology, or other services); and state agency overseers.

Relevant organizational skills, expertise, and capacity for this lead role include:

• the ability to communicate professionally and sensitively to all stakeholders using the media, methods, and language most acceptable and effective for each;

• an understanding of the problems and solutions involved in designing and operating data systems that require multi-organizational input and feedback;

• the ability to produce decisions from a multi-organizational group and track actions needed to implement decisions;

• the ability to effectively respond to and generate the interest of newspaper, television, radio, and other journalists;

• the ability to design systems to assess and report on the progress of implementation of a multi-organizational initiative that has multiple goals;

• the ability to identify needs for and acquire expert help quickly;

• access to adequate meeting space and administrative support needed to schedule and conduct regular meetings, produce and distribute materials in a timely manner, and communicate regularly with all stakeholders; and

• an in-place, modern computer and telecommunications infrastructure.

From a legal perspective, the organization that coordinates a work-based CSE program need not assume the role of employer nor any of the financial responsibilities of the program. On the other hand, as described below, some of the other functions of the program that have legal and financial implications may fit well with the coordination role. For example, if a labor market/workforce intermediary organization is selected or established to operate the program day to day and to play the principal employer role, such an organization would be in the best position to coordinate it as well.

            b. Recruiting worksites and or/worksite employers for CSE workers. The nature and difficulty of the task of recruiting worksites and/or worksite employers for a new wage-based CSE program under TANF will be influenced in large part by whether there is recent local experience with a "workfare" or CSE program (paid or unpaid) in the locality and, if so, how satisfied worksite sponsors and welfare recipients have been with the program. This experience will, in turn, suggest who should undertake the recruitment task, how it should be structured, and how much is asked of the worksites. Consider three localities:

Locality A — where there has not been any significant CSE since the CETA Public Service Employment program and substantial opposition to workfare kept the county government from approving a work experience component under AFDC/WIN in the 1980s and under JOBS in the 1990s.

Locality B — where the welfare department ran a relatively large-scale workfare program in the 1980s, but in the 1990s has concentrated its welfare-to-work efforts on job search/placement, remedial education, and job training, with only a small work experience component that includes occasional subsidized placements in the private sector through grant diversion.

Locality C — where the PIC/JTPA agency has successfully marketed subsidized jobs in the private sector, via the OJT component, for 15 years, on a small scale, but the welfare agency has no recent experience with community service or work experience.

In Locality A, a key task for recruiting worksites and/or employers for wage-based CSE would be to convince both potential worksite sponsors and the opponents of workfare that wage-based CSE is not the same program. Thus, a new player in the role of worksite recruiter might be the best choice. An organization formed for the purpose of operating wage-based CSE is one way to go, but it may be important to have a board of directors or other advisory group for a new organization that include previous opponents to workfare. It may also be important to minimize the responsibilities of the worksite in the program. This can be accomplished by having a labor market/workforce intermediary be the "employer" for CSE participants, which would entail, at a minimum, selecting their worksites (with worksite sponsor concurrence); taking care of most employee compensation functions; assuring compliance with labor laws; and providing some monitoring, job supervision, and dismissal and re-placement services — much like a temp service agency.

In Locality B, much of the relevant expertise and experience for starting a wage-based CSE program was in the local welfare agency. Depending on how much staff turnover has occurred since the operation of the large-scale workfare program in the 1980s, it may be important to involve this agency in a worksite recruitment role again. In this locality, worksite sponsors might be given choices about the extent of their roles in hiring, supervising, and paying participants.

In Locality C, wage-based CSE is going to be starting with a "clean slate," including the task of recruiting potential worksites, so the choice of an organization to play the recruitment role should be based on broader goals of the program — for example, if community investment is a prime goal, a community-based organization might be best — or on the marketing strengths of organizations playing other roles in the program.

            c. Setting expectations for training at worksites. This function goes along with those involved in recruiting worksites for a wage-based CSE program, although it may entail more communication and "hands on" work with the people who will supervise CSE participants and is an ongoing function, whereas worksite recruitment may be accomplished with an early big push and essentially completed after program start-up begins. The chief attributes of an organization that will be successful at this task are: customer orientation, flexibility in scheduling staff assignments, and good infrastructure and systems for tracking the status of worksites and participants. The individuals who perform this function need good customer service skills, experience in informal negotiation, and expertise in the processes of work and work-based learning relevant to the range of the program’s worksites. (Job developers, "industry reps," or vocational training specialists are potential candidates for this work, for example.)

If the organization(s) responsible for worksite recruitment for a wage-based CSE program does not have the number or types of staff necessary to visit worksites designated as "training worksites," develop training plans for participants in conjunction with their supervisors, and monitor participants’ training progress (or the necessary systems for tracking), the alternative(s) should also be selected based on their ability to stay in communication with the worksite recruiting organization and the CSE participant placing organization. For a relatively small number of training worksites, this function might be subcontracted to an appropriately qualified individual.

            d. Setting expectations for CSE workers on the job. Worksite employers must play a key role in setting expectations for participants’ performance, attendance, and workplace behavior in wage-based CSE. At a minimum, employers need to work with program staff (perhaps job developers) to develop written descriptions of the CSE positions, including policies regarding attendance and workplace behavior. (For example, are CSE participants expected to call the worksite employer if they will be absent or late? How are they expected to dress? What are the rules about lunch?) To the extent that the program is designed to simulate "real work," employers might have the only role in setting expectations, and the program might rely on them to hire, supervise, and fire participants according to their own personnel policies. (In this case, there would still need to be another organization responsible for assuring workers’ rights.)

The disadvantage of setting up such a strong employer role is that there is wide variability in the quality of supervision of entry-level workers, as well as in the consistency of human resources work among employers. Small, nonprofit organizations, in particular, may not even have written personnel policies or job descriptions for their own employees. Another way to set expectations for participants in wage-based CSE is to assign this aspect of worker supervision to a labor market/workforce intermediary organization — one that might also serve as the employer for all program participants, like a temp agency — that would ensure fairness and consistency for CSE participants, regardless of where they are placed. The organization selected for this role might be the same organization that recruits worksites because it is a natural outgrowth of that relationship with employers and calls for knowledge of the worksites and worksite personnel.

            e. Participant intake and assessment. The tasks of deciding which TANF recipients should participate in wage-based CSE and getting them started in the program can be carried out entirely by the public welfare agency or involve that agency only a little. The local partner selected for this functioning wage-based CSE depends, first of all, on the degree of choice recipients have in their participation and which recipients the program is intended to serve. For a program that is entirely voluntary and not narrowly targeted, for example, intake and assessment functions might be turned over either to community-based organizations or to a labor market/workforce intermediary organization. These organizations could then recruit participants directly from low-income communities or use mailing/contact lists provided by the welfare agency. Recipients who respond to marketing messages would be assessed for skills and experience, but much more emphasis might be placed on their vocational interests and aspirations than in a wage-based CSE program that is both mandatory and narrowly targeted. In this circumstance, the TANF recipient is the customer — or at least one of them.

If wage-based CSE is meant to serve either as a last-resort work option for TANF recipients who are supposed to be working but who cannot find or keep an unsubsidized job, or a résumé-creating work experience for long-term recipients who are nearing the end of their lifetime assistance period, the intake function will usually involve the local welfare agency more substantially. Where these welfare agencies have the computer capability to select potential candidates according to targeted characteristics, such as length of TANF assistance or previous work activities, the welfare agency might refer candidates to another organization — such as the local PIC, employment service office, or an independent labor market/workforce intermediary — for intake and assessment. (If the entire TANF work program is being operated by or coordinated by a nonprofit or private organization, of course, then that entity should also perform intake and assessment functions for wage-based CSE.)

The key tasks in assessment are to develop enough of a picture of each participant’s work-related capabilities to narrow the possibilities for a CSE assignment (assuming that there are many choices); to identify her needs for support services; and to begin the process of getting the support services in place. Where wage-based CSE is designed to upgrade the job skills or the earning potential of some participants, the assessment process will need to include consideration of their career interests and aspirations as well. One way of structuring the "front end" of wage-based CSE is to market the option to people who are still unemployed at the end of a job search training program. Alternatively, candidates can be selected for wage-based CSE at that point on the basis of a match between their skills and interests and available worksites.

            f. Assignment, job placement, re-placement, and/or hiring of CSE workers. The choice of an organization to handle this function again depends, in part, on whether participation in wage-based CSE is entirely up to TANF recipients and the extent to which it is intended to simulate real work. If recipients and worksite sponsors have complete freedom to choose, the program may be designed to create a marketplace for them. In this case, the role of the organization responsible for the job placement function would be to ensure that both buyers and sellers have complete information about each other and to facilitate searching and matchmaking.

Alternatively, if wage-based CSE is designed to "push" long-term welfare recipients into workplaces where they can learn and succeed, the function needs to be performed by an organization that has counselors experienced in matching low-skilled workers to jobs on the basis of personalities, skills, and worksite conditions. In different localities, any one of the following might have this capability and experience: welfare agency, community college, PIC/JTPA agency, PIC/JTPA contractor for employment and training services (including CBOs that have this experience), employment service, vocational rehabilitation agency, or labor market/workforce intermediary.

Assignment or job placement functions will need to be performed more than once for some participants. Some may choose to change worksites (if this is permitted); some may experience a change in their support services, health, or other personal circumstances that require a change in worksites; and some may be dismissed by their worksite employers for not meeting expectations. The critical capability of the organization responsible for reassignment and re-placement is keeping up-to-the-minute information on the status of participants at their worksites. Whichever agency performs the placement function should also have responsibility for re-placement, but this may not be the same agency that performs the case management/counseling and job development roles discussed below. If not, the placing organization must have an effective and time-sensitive information link to the case management organization.

            g. Orientation, preparation, and/or training of workers for their CSE jobs. This function should be located as close to the actual CSE jobs as possible in order to ensure that it accurately reflects workplace expectations and conditions. In other words, if worksite sponsors are performing the function of supervising CSE participants, it is most useful for the supervisors to get them ready for the jobs. If a labor market/workforce intermediary is acting as the employer and supervising participants on-site, the orientation/preparation/training function should be performed by that organization. If there is an option for participants to request a change of assignment, the orientation/preparation/training process may stimulate requests for changes as participants learn more about the work and worksites. Thus, it will be important for participants and worksite sponsors to be informed about how to exercise the option.

            h. On-the-job supervision and job skills training. This is a function that should be performed by worksite sponsors most of the time, because they are most familiar with the demands of the CSE jobs. However, an on-site supervisor employed by another organization (such as a labor market/workforce intermediary or another local provider of employment and training services) might be placed at some CSE worksites — for example, if there is a large number of CSE workers at the worksite who could be managed as a crew. Job coaches from another organization could perform the supervision and training functions on-site as well.

            i. Dismissal of workers for not meeting expectations. In a "real work" environment, this function will be performed by the organization that supervises participants’ work at the worksite. Many people are uncomfortable with this task, however, and would prefer to let a worker (especially one who is not costing the worksite sponsor money) continue to perform inadequately as long as she or he is not making trouble, rather than talk to the worker directly about poor performance. Thus, it is better for the worksite sponsors, participants, and the CSE program to offer some worksite sponsors the service of intervening with a CSE worker who is performing inadequately than to allow the CSE experience to degenerate into "make work." Dismissals are an opportunity for participants to learn about the workplace and should be "processed" with a counselor. (It is also important to note that the ability to dismiss a worker is one of the key indicators of who is the employer in an employer-employee relationship for the purposes of defining some legal and financial responsibilities of employers.)

            j. Case management, counseling, and job development during CSE placement. These three functions logically go together as a set of activities that must be performed to keep CSE participants on track toward a positive reference for work and an unsubsidized job. They do not need to be performed by the same individual, but it is better for coordination and communication for them to be performed by a single organization for each case. That is, CSE participant cases might be divided among two or more organizations responsible for these activities — for example, on the basis of geography in a large program region so that case management organizations concentrate on clusters of worksites that are in close proximity to each other or neighborhoods where participants live — but each participant should deal with only one organization during her CSE placement.

These activities must be organized so that feedback from worksite employers (about participants’ attendance, performance, and any problems on the job) and from participants are channeled to the organization(s) responsible for worksite recruitment, participant selection, matching participants to worksites, and program planning. Although any number of employment and training organizations might be selected to perform case management, counseling, and job development during CSE placements, there are some important considerations:

• Job development is a specialized skill and cannot be performed by just any counselor.

• These activities are a logical extension of the assignment/job placement function; if they are performed by an organization that is different from the one that has responsibility for assignment/job placement, very careful attention to communication issues is needed.

• There are differing professional identity issues associated with the functions of job developer, case manager, and counselor, depending on the training and experience or the staff who fill these roles. Most often, these identity issues work against program goals when job developers identify with employers and case managers and/or counselors identify with clients. Organizations that employ people performing all three functions need to be prepared to address this issue.

            k. Promoting CSE to TANF recipients as an opportunity. In the context of work requirements for a larger proportion of welfare recipients than ever before, lifetime limits on cash assistance, and welfare-to-work programs that are investing less in education and job skill training and more in "work first" strategies, recipients might respond to the idea of wage-based CSE as just another way of pushing them into the workforce. Or, wage-based CSE might be seen as a promising opportunity to prepare for competing in the labor market. Because it offers many advantages compared to unpaid community service for people who cannot find jobs on their own — especially for those who have poor work records — wage-based CSE might be successfully promoted as a useful step toward self-sufficiency.

The credibility and "stance" of the organization doing the promoting is likely to be important to how the message about wage-based CSE is received by TANF recipients. Organizations that have a local track record advocating for the interests of low-income people are the best choice for this role in getting a wage-based CSE program started. However, it is important that such organizations speak from direct experience, perform ongoing and substantial functions in the program, and are positioned in the program structure to receive feedback on participants’ progress and to continue to advocate for the interests of participants.

            l. Financial functions involving CSE workers. Keeping track of participants’ actual hours worked, paying them (including tax withholding and EIC advances), and keeping track of and providing their fringe benefits, if applicable, are best done by worksite employers when the purpose of wage-based CSE is to simulate "real work." However, as an inducement for employers to participate, the program might offer centralized payroll services, which would require only that employers submit time sheets regularly. In this case, the local organization that has responsibility for coordination of wage-based CSE might engage a payroll business service company to handle the task.

There are other circumstances when it might be appropriate to choose centralized payroll services. For example, it may be difficult for public agencies to handle payment to CSE participants who are not their employees, or there may be small organizations serving as worksites that do not have the accounting capability to handle the EIC. The most important considerations for deciding how to handle these functions are:

• Reliability — CSE workers are financially vulnerable and should not have to cope with missed paydays.

• Accuracy — Similarly, CSE workers cannot afford to be paid less than they are owed or overpaid in a manner that creates repayment liability.

• Scale — It will be difficult for a very large program to manage different pay arrangements for different worksite employers. If there are more than a few worksite sponsors that cannot or would prefer not to handle payroll for their CSE participants, the programs might do payroll services for all.

            m. Financial functions involving employers. The extra costs to employers of workers’ compensation, unemployment insurance, FICA, and fringe benefits for CSE participants can all be reimbursed by the program — a program feature that can be a significant boost in marketing to and recruiting worksites.⁴⁶ However, this is an area is which wage-based CSE programs may wish to distinguish between public or nonprofit employers of CSE participants and private, for-profit employers, if the latter are included. One way of balancing the benefits of the participants’ free labor to the private, for-profit employers with benefits to participants and society of the work experience these employers offer in a wage-based CSE program is to require that the for-profit organizations pay for the "extras" beyond the cost of wages.

Employers’ concerns about their experience ratings for the Unemployment Insurance system, which determine the amount they must contribute, might be a reason to create an intermediary organization to serve as the employer of record. Alternatively, employers might be compensated variably, depending on how much a wage-based CSE position that turns over every few months costs them in UI contributions.

            n. Arranging and paying for support services. Arranging for support services logically fits with participant intake and assessment, and with in-program case management and counseling, because it entails sensitive one-on-one interaction with participants. Thus, the arranging (if not the paying) part of the function might be assigned to the organization with responsibility for one or both of the other in-person services. Paying for support services, on the other hand, can be a straightforward check-writing service performed by a business services company or accounting department of a public agency, handled in the same way as payroll as long as there is a timely and effective link between the program staff who approve the payments (presumably the "arrangers") and the check writers.

Recent very difficult experience in a few states suggests another important proviso: If the arranging and paying functions are carried by a single organization and that organization has performed poorly in a similar role in the past, it is better to create the capability from scratch in a new organization because inadequate performance on arranging and paying for support services can shut down a program like wage-based CSE. (Organizations with a record of poor performance on any of the functions needed for wage-based CSE should be considered very carefully; arranging and paying for support services is a prerequisite to participation, however, and good performance is thus more important than in some other areas.)

            o. Assuring the rights of CSE workers. Table 1 indicates that most of the service delivery functions of a wage-based CSE program can be performed by several types of organizations, including employer organizations. Assuring the rights of CSE workers is one function that probably poses conflicts of interest for employer groups, however, and possibly for other partners in wage-based CSE programs as well if another organization serves as the employer of record for CSE participants. The regular system of protections for workers under the Fair Labor Standards Act would apply to participants in wage-based CSE, making the employer incapable of assuring workers’ rights in a dispute. It is also important that one or more organizations in the program partnership have the responsibility for looking over the whole of the program’s processes and outcomes from the perspective of the rights of workers. Again, that responsibility should not be assigned either to an employer group or to the employer of record for the CSE participants.

            p. Transferring and accounting for public funds used for wage-based CSE programs. The most important issue here is that the public agency with control over TANF or other funds to be used for wage-based CSE needs to work closely with the organizations that will have overall coordination responsibility (if they are different) in order to estimate the amounts and purposes of funds needed and to develop appropriate accounting procedures and contractual arrangements with service providers. Because TANF funds and Welfare-to-Work funds flow to different state and local entities with different accounting and performance standards, it may be simplest in accounting terms for localities that intend to use both sources to divide them according to purpose. For example, state maintenance-of-effort funds might be used to pay wages in CSE, while TANF funds might be used to pay for staff, support services, business services, and other functions.

XII. Conclusions

TANF offers an unusual opportunity for job creation that can at once benefit welfare recipients, taxpayers, the community, and employers. As a concept, wage-based community service employment has the potential to recreate, in more flexible form, many of the job training options previously available for the most disadvantaged workers, to tailor these to the needs and circumstances of participants and employers, and to serve wider social and economic purposes as well. Wage-based CSE is not yet a completely recognizable program, however. Much design and testing work will be required for stakeholders to learn the easiest and most effective approaches. Fortunately, there is considerable relevant experience for operating wage-based CSE. It requires reconfiguring functions that are already being performed in most localities, bringing functions together under a coherent management structure, and involving a broader group of interests and organizational resources than may have been involved in welfare-to-work efforts previously. There are also adequate organizational resources for the program in most localities. The most difficult task entails pulling these resources together in a way that will benefit all, which in turn requires thinking and acting outside the old organizational boundaries. As noted at the beginning of this paper, it is not too soon to start.

Notes:

1. TANF replaced the Aid to Families with Dependent Children (AFDC) program in 1996 as a result of PRWORA.

2. The U.S. Department of Labor has produced a guide in question-and-answer format that provides more details about these rules, called "How Workplace Laws Apply to Welfare Recipients" (Washington, D.C.: U.S. Department of Labor, May 22, 1997).

3. PRWORA set 60 months as the maximum for TANF cash assistance over the lifetime of a recipient, but it permitted states to establish a more stringent lifetime limit, which 19 have done. For example, California, Florida, and Virginia set 24-month limits, and Connecticut’s lifetime limit is 21 months.

4. Creating and operating enterprises that have both revenue-generating and social justice goals is not easy. For an examination of both the theory and practice, see Jed Emerson and Fay Twersky (eds.), New Social Entrepreneurs: The Success, Challenge and Lessons of Nonprofit Enterprise Creation (San Francisco: Roberts Foundation, 1996). For an illustration of the challenges of welfare-to-work efforts in depressed economies, see Jason DeParle, "Welfare Law Weighs Heavy in Delta, Where Jobs Are Few," New York Times, October 16, 1997, p. A1.

5. For analyses of whether the TANF goals, or similar employment levels for welfare recipients, can be achieved, see Greg J. Duncan, Kathleen Mullan Harris, and Johanne Boisjoly, "Time Limits and Welfare Reform: New Estimates of the Number and Characteristics of Affected Families," April 22, 1997; Suzanne L. Wagner, Toby Herr, Charles Chang, and Diana Brooks, Five Years of Welfare: Too Long? Too Short? Lessons from Project Match’s Longitudinal Tracking Data (Chicago: Project Match/Erikson Institute, June 1998); and James Riccio and Stephen Freedman with Kristin S. Harknett, Can They All Work: A Study of the Employment Potential of Welfare Recipients in a Welfare-to-Work Program (New York: MDRC, September 1995).
            For an example of an effective program that would fall short of TANF standards, see Susan Scrivener, Gayle Hamilton, Mary Farrell, Stephen Freedman, Daniel Friedlander, Marisa Mitchell, Jodi Nudelman, and Christine Schwartz, National Evaluation of Welfare-to-Work Strategies: Implementation, Participation Patterns, Costs, and Two-Year Impacts of the Portland (Oregon) Welfare-to-Work Program (Washington, D.C.: U.S. Department of Health and Human Services/U.S. Department of Education, 1998).
            For a description of GA programs, see Cori E. Uccello, Heather R. McCallum, and L. Jerome Gallagher, "State General Assistance Programs 1996" (Washington, D.C.: Urban Institute, November 26, 1996).

6. See Judith M. Gueron and Edward Pauly, From Welfare to Work (New York: Russell Sage Foundation, 1991).

7. For a discussion of the many routes out of welfare, see Toby Herr and Robert Halpern, Changing What Counts: Re-Thinking the Journey Out of Welfare (Project Match) (Evanston, Ill.: Northwestern University, April 1991); and Toby Herr, Suzanne Wagner, and Robert Halpern, Making the Shoe Fit: Creating a Work-Prep System for a Large and Diverse Welfare Population (Project Match for the U.S. Department of Housing and Urban Development, Office of Public and Indian Housing, December 1996).

8. See Susan M. Poglinco, Julian Brash, and Robert C. Granger, An Early Look at Community Service Jobs in the New Hope Demonstration (Working Paper) (New York: MDRC, July 1998).

9. See Gueron and Pauly, From Welfare to Work.

10. Reports of these circumstances in New York City’s work experience program (WEP) for welfare recipients are offered by participants in the New York City Parks Career Training (PACT) program, for example. (PACT was a special, more intensive alternative to WEP.) See Richard Williams and Sonia Ospina, PACT (Parks Career Training Program) Evaluation Project, First Round of Focus Groups (Willsop Consulting for the PACT Evaluation Group, August 17, 1994); and Bradford Petrie, Final Report of the Evaluation of the Parks Career Training Program (PACT): Implementation, Participation Patterns and Outcomes (PACT Evaluation Group, Inc., June 30, 1996). These documents also include PACT participants’ reports of how worksite supervisors and regular Parks Department employees treated them.

11. See PACT reports, op. cit., for examples of how participants feel about different types of supervisors and about meaningful work.

12. For a discussion of the criteria that employers use to screen applicants for jobs that do not require a college degree, see Harry J. Holzer, What Employers Want: Job Prospects for Less-Educated Workers (New York: Russell Sage Foundation, 1996), pp. 54-62. Research in four cities for this study showed that about 70 percent of noncollege jobs required general work experience, 60 percent required experience in the job category of the available position, and 73 percent required references.

13. See, for example, LaDonna A. Pavetti, Learning from the Voices of Mothers: Single Mothers’ Perceptions of the Trade-offs Between Welfare and Work (New York: MDRC, 1993).

14. Employment and training providers have never been good at predicting who can succeed, because the ability to stick with a program or a job depends heavily on situational factors. For a recent discussion of this phenomenon, see Herr, Wagner, and Halpern, Making the Shoe Fit.

15. Evaluations of such programs run during the 1980s found that the primary reason for selecting a particular worksite was location; close-by jobs were most often selected so that participants could either walk to them or minimize their travel time and work-related expenses. See, for example, the following reports from the Demonstration of State Work/Welfare Initiatives by MDRC: Janet Quint, Interim Findings from the Arkansas WIN Demonstration Program (1984); Barbara Goldman, Judith Gueron, Joseph Ball, and Marilyn Price, Preliminary Findings from the San Diego Job Search and Work Experience Demonstration (1984); and Joseph Ball, Interim Findings on the Community Work Experience Demonstrations (West Virginia) (1984).

16. See Poglinco, Brash, and Granger, An Early Look at Community Service Jobs in the New Hope Demonstration.

17. See Gueron and Pauly, From Welfare to Work.

18. Clifford M. Johnson and Ana Carricchi Lopez, Shattering the Myth of Failure: Promising Findings from Ten Public Job Creation Initiatives (Washington, D.C.: Center on Budget and Policy Priorities, November 12, 1997).

19. See Burt Barnow, The U.S. Experience with Public Service Employment Programs (Baltimore: Johns Hopkins University, 1994).

20. For a description of the program tested and its results, see Stephen H. Bell, John H. Enns, and Larry L. Orr, Overview of Evaluation Results: Evaluation of the AFDC Homemaker-Home Health Aide Demonstrations (Cambridge, Mass.: Abt Associates Inc., 1987).

21. See Ronald F. Ferguson and Philip L. Clay with Jason C. Snipes and Phoebe Roaf, YouthBuild in Developmental Perspective: A Formative Evaluation of the YouthBuild Demonstration Project (Cambridge, Mass.: Massachusetts Institute of Technology, September 1996).

22. See Petrie, Final Report of the Evaluation of the Parks Career Training Program (PACT), pp. 72-75.

23. After the Job Training Partnership Act (JTPA) system of publicly funded employment services for low-income and other disadvantaged groups had been operating for several years, the regulations governing subsidized on-the-job training were changed in order to assure that training was actually taking place. New documentation requirements for OJT were considered by many employer and local JTPA administrators to be onerous, and the use of OJT declined — although there was no evidence to indicate how much legitimate training on the job was driven out by paperwork versus how much subsidized placement without training was driven out by the prospect of closer scrutiny. The recommendation here is that the wage-based CSE program staff, rather than worksite sponsors, should complete most of the documentation work.

24. For a study of the last major devolution of federal government responsibility, see Richard P. Nathan, Fred C. Doolittle, and Associates, Reagan and the States (Princeton, N.J.: Princeton University Press, 1987).

25. Formerly, the American Public Welfare Association.

26. A key advantage of wage-based CSE over unpaid community service in exchange for TANF assistance is the possibility for more income for participants. Thus, it makes little sense to create program conditions that would yield less income for participants than TANF assistance, although TANF could be used to "top up" earnings from wage-based CSE in the same way that part-time work in the competitive labor market is supplemented by TANF if earnings do not reach the level to make a low-income person ineligible for assistance. Administrators in high-grant states will need to be particularly attentive to the income yield of wage-based CSE (including the EIC and considering how FICA and other work-related expenses are handled) compared to TANF assistance because, in the era of lifetime limits on cash assistance, combining work and welfare carries new costs for recipients.

27. For a review of the record of community economic development, generally, including job creation efforts, see Robert Halpern, Rebuilding the Inner City (New York: Columbia University Press, 1995). In reaction to the Clinton Administration’s creation of Empowerment (and Enterprise) Zones, Nicholas Lemann wrote a strongly critical review of federal urban economic development efforts in the January 9, 1994, issue of the New York Times Magazine. New York State’s experience has been regularly reviewed and critiqued by Frank Mauro, Director of the Albany-based Fiscal Policy Institute; see, for example, "State Government Targeting in Economic Development: The New York Experience," Publius: The Journal of Federalism, Vol. 19 (Spring 1989). Hugh O’Neill has also written about state government policies affecting jobs and income for disadvantaged individuals and communities in Creating Opportunity: Reducing Poverty Through Economic Development (Washington, D.C.: Council of State Planning Agencies, 1985). Industrial recruitment policies of the southern states have been studied and found ineffective for adding jobs by MDC, Inc.; see, for example, Coming Out of the Shadows: The Changing Face of Rural Development in the South (Chapel Hill, N.C.: MDC, Inc., May 1992). Philip Kasinitz and Jan Rosenberg point to inner-city employer hiring practices that specifically exclude neighborhood residents as the chief reason "Why Enterprise Zones Will Not Work: Lessons from a Brooklyn Neighborhood," published in City Journal (Autumn 1993).

28. For a review of the record of federal job creation programs, see Clifford M. Johnson, Enhancing Income Security Through Public Job Creation (Washington, D.C.: Center on Budget and Policy Priorities, April 1997). An article by Johnson for the 1997 Entrepreneurial Economy Review, published by the Corporation for Enterprise Development, argues for "using community jobs to leverage and support broader community-building efforts." The article is entitled "Toward a New Generation of Community Jobs Programs" and is dated July 25, 1997.

29. Community Development Corporations (CDCs), a major vehicle for neighborhood revitalization since the 1960s, might play a key organizing and administrative role in a neighborhood-targeted waged-based CSE. For research on the accomplishments, strengths and weaknesses of CDCs, see Avis C. Vidal, Rebuilding Communities: A National Study of Urban Community Development Corporations (New York: New School for Social Research, Community Development Research Center, Graduate School of Management and Urban Policy, 1992); Mercer L. Sullivan, More Than Housing: How Community Development Corporations Go About Changing Lives and Neighborhoods (New York: New School for Social Research, Community Development Research Center, 1993); and Xavier de Souza Briggs and Elizabeth J. Mueller with Mercer Sullivan, From Neighborhood to Community: Evidence of the Social Effects of Community Development (New York: New School for Social Research, Community Development Research Center, 1997).

30. In an essay entitled "Why Business Alone Won’t Redevelop the Inner City," Bennett Harrison and Amy Glasmeier offer "A Friendly Critique of Michael Porter’s Approach to Urban Revitalization" in Economic Development Quarterly, Vol. 11, No. 1, February 1997; they cite "a huge research literature" on the "trade flows" into and out of poor, minority, urban neighborhoods, which includes discussion of how money leaves these neighborhoods.

31. Poglinco, Brash, and Granger, An Early Look at Community Service Jobs in the New Hope Demonstration.

32. For a complete discussion of New Hope, its incentive structure, and how community service jobs fit into its array of services and its philosophy, see Thomas Brock, Fred Doolittle, Veronica Fellerath, and Michael Wiseman, Creating New Hope: Implementation of a Program to Reduce Poverty and Reform Welfare (New York: MDRC, October 1997).

33. Milwaukee County’s W-2 program for welfare recipients, which has eliminated all cash assistance for people determined to be job ready and has substituted community service jobs, transitional jobs, and trial jobs for those who cannot find unsubsidized employment, illustrates the central role of community service work in a welfare program that has taken extreme measures to close cases. After a dramatic decline of three-quarters in the county’s total welfare caseload between January 1987 and April 1998, 86 percent of the working W-2 participants in April 1998 were in community service jobs. Wisconsin Department of Workforce Development, "Statistical Report for the Quarter April-June 1998."

34. See Petrie, Final Report of the Evaluation of the Parks Career Training Program (PACT). As with the New Hope outcomes reported above, these results are not based on comparison with any type of control group, although New York City and a neighboring jurisdiction claimed to place 5 percent and 15 percent, respectively, of their regular CWEP participants in unsubsidized jobs.

35. See Petrie, Final Report of the Evaluation of the Parks Career Training Program (PACT).

36. For an extensive discussion of the advantages and disadvantages of different funding strategies, see Steve Savner and Mark Greenberg, The New Framework: Alternative State Funding Choices Under TANF (Washington, D.C.: CLASP, March 1997).

37. Herr, Wagner, and Halpern discuss flexibility of welfare-to-work programs and, by implication, of funding, in Making the Shoe Fit.

38. For details, see Mark Greenberg, "Welfare-to-Work Grants and Other TANF-Related Provisions in the Balanced Budget Act of 1997" (Washington, D.C.: CLASP, August 1997); and U.S. Department of Labor, Planning Guidance and Instructions for Submission of Annual State Plans, Fiscal Year 1998, Welfare-to-Work Formula Grants, October 1997.

39. In the 1980s, "social contract" in the welfare context referred to efforts toward self-sufficiency on the part of welfare recipients in exchange for publicly funded cash assistance. See, for example, Lawrence M. Mead, Beyond Entitlement: The Social Obligations of Citizenship (San Francisco: Free Press, 1986).

40. These are net costs — that is, the additional cost of providing the work program service above what control group members received, which was generally a minimal information and counseling contact at intake. See Gueron and Pauly, From Welfare to Work, pp. 168-174.

41. For detailed instructions on how to estimate costs of employment and training programs, see David Greenberg and Ute Appenzeller, A Guide to Estimating the Costs of Employment and Training Programs (New York: MDRC, June 1998).

42. David Osborne and Ted Gaebler provide a history of "entrepreneurial government" in the introduction to their book, Reinventing Government: How the Entrepreneurial Spirit Is Transforming the Public Sector from Schoolhouse to Statehouse, City Hall to the Pentagon (New York: Addison-Wesley, 1992).

43. The JTPA system of job training and placement of disadvantaged workers was superseded by the Workforce Investment Act (WIA) of 1998, although funding formulas under WIA are similar to the JTPA allocations.

44. Public funds had long been used to pay tuition for welfare recipients to attend for-profit vocational schools as well as nonprofit educational institutions under the Work Incentive (WIN) program, but these arrangements tended to be "self-initiated" educational plans made by welfare recipients rather than formal agreements between public agencies and training and education providers.

45. See, for example, David Firestone, "He Fights, Patiently, for Workfare Laborers," New York Times, January 16, 1998, p. B2. The right to bargain collectively has not been recognized by the U.S. Department of Labor as one of the standard worker protections covering welfare recipients participating in work activities under PRWORA.

46. See Maurice Emsellem and Steve Savner, The Fiscal and Legal Framework for Creating a Community Service Employment Program, November 1997.