Introduction

Much of the current effort to find new strategies for helping the poor is focused on finding ways to link income support more closely to work or work-related activities. The New Hope Project in Milwaukee, Wisconsin, offers an innovative approach to reducing poverty, reforming welfare, and addressing the economic insecurity of low-income workers. It seeks to increase employment and reduce poverty by creating better financial incentives to work and by changing labor market opportunities; it offers assistance that enables poor people to support themselves and their families through full-time employment. New Hope serves as a model program for planners involved in the design of welfare reform and antipoverty programs nationwide. It addresses many issues on the nation's social policy agenda, including the design and operation of the Earned Income Credit (EIC) for low-income workers, community service jobs for people who need employment, and access to health insurance and child care for working families.

Participation in the program is voluntary, and eligibility is based on income and a willingness to work at least 30 hours per week. Adults (defined as age 18 or over) are eligible regardless of whether or not they have children or are current or past recipients of public assistance. Persons meeting these criteria are eligible to receive these benefits or services:

• help in obtaining a job, including access to a time-limited, minimum-wage community service job (CSJ) if full-time employment is not otherwise available;

• a monthly earnings supplement that when combined with federal and state EICs brings most low-wage workers’ incomes above the poverty level;
  
• subsidized health insurance, which gradually phases out as earnings rise; and
  
• subsidized child care, which also gradually phases out as earnings rise.

New Hope staff are actively involved with participants — explaining the rules for accessing the various program components, providing information on health and child care services, reaching out to those not active in the program, and serving as coaches to support individuals’ employment efforts.

New Hope operates outside the existing public assistance system, though it is designed to be replicable as government policy should the demonstration findings be favorable. It is funded by a consortium of local, state, and national organizations interested in work-based antipoverty policy, as well as by the State of Wisconsin and the federal government. It was designed and is operated by a community-based nonprofit organization, the New Hope Project, and thus provides insights into the role nongovernmental agencies can play in income support.

One goal of the project is to provide credible information to policymakers on the implementation, effectiveness, and costs of the New Hope approach. In 1994, program designers initiated a demonstration of the program in two inner-city areas in Milwaukee. New Hope operated in two racially and ethnically diverse areas of the city (defined by two zip codes) that are economically depressed, but nevertheless contain working residents and households that do not fit the stereotypes of "dysfunctional" families. Geographic targeting of New Hope was intended to concentrate resources in two areas with high levels of poverty, thus allowing a more detailed analysis of program context than would be possible in a program that served a wide geographic area.

New Hope contracted with the Manpower Demonstration Research Corporation (MDRC) to conduct an independent evaluation of the program's context, implementation, impacts on key outcomes, and costs. Among the central questions in the evaluation are: How much will New Hope services actually be used, and do those with access to New Hope achieve better outcomes than those with access to the pre-existing service supply? In order to provide a reliable test of the difference the program made, applicants were randomly assigned in a lottery-like process to either a program group (with access to New Hope services) or a control group (with no access to New Hope services, but able to seek other services). The differences in the two groups’ outcomes over time (for example, their differences in employment rates or average earnings) are the observed impacts of the program.

This report examines the creation of the New Hope Project, the implementation of the demonstration, the labor market and neighborhood context of the experiment, and the use of program services by participants. It offers insights on program design, administrative and operational issues, and benefit use rates in New Hope. A future report will analyze program impacts and costs.

The early findings on implementation and program use, reported here, reveal that the New Hope package of benefits and services has considerable appeal for participants seeking to work and support themselves and their families. Even though this program may differ from reforms contemplated elsewhere, it has much to teach about the nature and appropriate responses to issues arising as programs change to supplement the payoff from work.

I. Findings in Brief

A. Demonstration Context

• New Hope was implemented in a strong labor market and a time of rapid change in the welfare system. In late 1995 at the point that recruitment for New Hope ended, the unemployment rate in the Milwaukee metropolitan area was low. However, much of the growth in jobs, especially those open to workers without a high school diploma was occurring in suburban locations difficult for residents of the New Hope neighborhoods to reach by public transit. Thus, while these strong labor market conditions increased the overall probability that those in New Hope could find an unsubsidized job and access program benefits, CSJs would still remain important for some participants. In addition, the public welfare system in Milwaukee and the State of Wisconsin was undergoing major reform. Within Aid to Families with Dependent Children (AFDC), program participation and work requirements increased over time and the caseload dropped substantially. At the same time, cash assistance under the county’s General Assistance program ended. These contextual factors do not invalidate the basic comparisons involved in the study of program impacts because they affect both those served within New Hope and those in the control group, but probably a more disadvantaged group applied for the program and fewer participants needed CSJs than would otherwise have been the case.
  
• Within this changing context, New Hope offered a distinct package of benefits and services with broader eligibility rules than normal in income support programs. For most single individuals and families without children, New Hope's benefits were not available under any other program. Even for families with children — the group typically served in public assistance programs — the package of benefits was unique. For these families, some of New Hope's benefits are available through other sources; subsidized health insurance and child care are available through public assistance programs and Medicaid, and earnings supplements are available through the federal and state EIC. However, paid CSJs are typically not offered. Furthermore, one premise of New Hope's design is that the combination of benefits is more than the sum of its parts because together they address the main barriers to the achievement of an income above poverty through work. Also, the assistance and "coaching" of New Hope project representatives can help participants take greater advantage of the services than they otherwise might.

B. Program Implementation

• Recruitment for the New Hope Demonstration occurred over a 16-month period beginning in July 1994 and produced a diverse sample for this research that in many ways reflected the characteristics of the eligible population in the neighborhood. Program applicants resembled in most ways the larger pool of neighborhood residents eligible for the program and interested in its services. Applicants included those traditionally served in public assistance programs (for example, unemployed parents with dependent children) and also low-income working parents and adults without dependent children. Recruitment proved a difficult challenge for New Hope staff. Key problems were finding ways to bring the program to the attention of potential applicants and explaining the geographic eligibility rules and program participation requirements. However, when people who met the program’s eligibility rules attended an orientation explaining the program, most found it an attractive option and applied to participate in the demonstration.
  
• The community-based organization operating New Hope successfully put in place the intended program services. Program services were fully implemented and available to program group members. A vital role is played in the New Hope program by the "project representatives," staff who explain program services, compute benefits, and monitor participation for their caseloads of approximately 75 participants each. Despite such efforts, participants had some difficulties understanding how the various parts of the New Hope offer worked.
  
• The random assignment impact research design was successfully implemented, providing a means to understand the net impact of New Hope on key outcomes. The goals of achieving a diverse and sizable sample were met; the background characteristics of the program and control groups are similar, allowing a comparison of the program and control groups’ levels of employment, earnings, public assistance receipt, family and child outcomes (where applicable), and other key measures. These findings, based on follow-up using administrative records and a survey, will be the subject of a later New Hope evaluation report.

C. Program Use

• At some point in the year following random assignment, approximately three-quarters of the applicants accepted into the New Hope program group worked full time and claimed a program benefit. Use of New Hope benefits is affected by the availability of and changes in other "safety net" programs, as described earlier in this summary. During the follow-up period for this report, earnings supplements were most frequently used (by 72 percent of the program group), followed by health insurance (38 percent), and child care (23 percent). Twenty-four percent took a CSJ for at least a day as a way to meet the New Hope requirement of employment. About 60 percent of these CSJ workers made a transition to a full-time, unsubsidized job at a later point in the follow-up period, which qualified them for New Hope benefits.
  
• People used the program in many different ways, with differences in use reflecting their different initial circumstances, their ability to find and retain a full-time job, and their desire to maintain contact with the program. After an initial start-up period (defined as the first three months after random assignment), 32 percent of the program group used benefits steadily or nearly so, 39 percent intermittently, and 29 percent not at all. Since most participants do not use services continuously, it appears that New Hope serves principally as a resource for those beginning employment and as a support and safety net for those who obtain a job. Later data collection will provide details about reasons for nonuse of program benefits.

II. The New Hope Program Design

A. The Program Model

Four principles underlie the New Hope program: (1) that people who are willing and able to work full time should be assured the opportunity to do so; (2) that people who work full time should not be poor; (3) that people who work more hours should take home more pay; and (4) for those eligible for public assistance, that full-time work should make people better off financially than they would be on welfare. These principles are realized by providing four benefits and services to participants who are willing to work an average of at least 30 hours per week: help in obtaining a job (including access to a CSJ if full-time employment is not otherwise available), an earnings supplement to bring low-wage workers’ income above the poverty level, subsidized health insurance, and subsidized child care. The major benefits and services offered by New Hope are summarized in Table ES.1.

The program is designed so that there will always be a financial incentive to increase work hours and earn higher wages. Because the New Hope earnings supplement and subsidies for health insurance and child care decline as earnings rise, a participant does not see a $1 increase in total income for each $1 increase in earnings. New Hope designers developed an earnings supplement that phased out at a slow enough rate so that participants always saw total income rise as they worked more or earned higher wages. In New Hope, people see at least a $.30 rise in total income for each $1 increase in earnings, compared with no increase in total income for some existing public assistance programs that reduce their grant $1 for each $1 earned.

New Hope is intended to be flexible. People in the program group may enter and exit voluntarily and use whichever benefits they need. They may also access public assistance alone or in combination with New Hope if they wish and are eligible. However, receipt of New Hope benefits generally makes people ineligible for welfare benefits because their total incomes become too high. Some people may use New Hope on an ongoing basis to boost their incomes and help them stay employed; others may use it as insurance for the times they need help. At all times staff try to provide full explanations to participants of program operation, benefits, and alternatives. In short, New Hope is a new antipoverty resource for individuals willing and able to work.

B. The New Hope Demonstration

During the demonstration, the New Hope Project is serving a diverse program group of 678 low-income people drawn from two areas of inner-city Milwaukee. The eligibility requirements are that applicants must live in the targeted service areas, be age 18 or over, be willing and able to work at least 30 hours per week, and have a household income at or below 150 percent of the federally defined poverty level. Single- and two-parent families and adults without children who meet income and geographic eligibility requirements are eligible to participate, and no past or current receipt of public assistance is required. Because of budgetary constraints, the New Hope offer is open to members of the program group for a period of three years from the date they became part of the demonstration. Such a time limit is not integral to the design of the program, and the New Hope demonstration was not intended to provide a test of the effects of time limits on public assistance.

The New Hope program is being evaluated to determine its effects on economic measures such as employment, income, public assistance use, access to and use of health insurance, and purchase of paid child care. In addition, the evaluation seeks to assess the consequences for participants’ sense of well-being as reflected in various other measures of material well-being, family stability, and progress in achieving personal goals. The evaluation will also focus on understanding outcomes for families with children.

III. Program Context

A. Labor Market Conditions

 New Hope was implemented during a period of strong economic growth and falling rates of unemployment in the Milwaukee metropolitan area. However, as in other older metropolitan areas, a “spatial mismatch” was evident: The greatest employment growth was occurring in the suburban fringe, not in the central city and not in the vicinity of the New Hope target areas. While many jobs are still available in the central city, the selection and wages offered are not generally as good as elsewhere in the labor market.

 B. Public Assistance Reforms

 Profound changes have also been occurring in the state and national welfare systems. General Assistance (a program of cash assistance largely for single adults and families not eligible for federally funded welfare) was recently eliminated in the State of Wisconsin, and the state’s AFDC program (now called Temporary Assistance for Needy Families) became more restrictive. In early 1996, the state began requiring applicants for AFDC to conduct a job search prior to the approval of their AFDC grant and linked payment of AFDC benefits to compliance with program participation requirements (with reductions in the benefit for hours of required activities or work missed). Both of these changes are elements of a major state welfare reform initiative — Wisconsin Works — which was implemented statewide in September 1997. Since New Hope operates entirely outside the public assistance system, any New Hope program group members who are also receiving public assistance are required to comply with relevant program requirements. Receipt of New Hope benefits normally raises a person’s income above the eligibility cutoff for cash assistance, but participants may still be receiving Food Stamps and Medicaid.

 By altering the prospects for persons relying on the traditional safety net provided by AFDC, these state changes have affected, and will continue to affect, program group members’ perceptions of the usefulness of the New Hope package (probably making it more appealing relative to welfare) and the alternatives available to members of the control group (making them more linked to work effort, like New Hope). These changes did not affect low-income program group members who were not receiving public assistance and were not contemplating accessing the affected programs.

 Despite these changes, New Hope’s package of benefits and services remains unique in Milwaukee and control group members cannot access any comparable program. No other provider offers paid CSJs and earnings supplements. Other New Hope benefits and services — job search assistance, health insurance, and child care assistance — may be available in some form through the welfare department (or in the future Wisconsin Works service providers) or other agencies. New Hope offers an alternative to services through the public assistance system and serves people who are ineligible for welfare.

C. Conditions in the New Hope Target Neighborhoods

The two areas targeted by New Hope have high unemployment and high poverty, and contain many families receiving welfare. They include many census tracts that have been identified in recent social science literature as exhibiting "ghetto poverty." Initially, program recruitment focused on smaller geographic areas that were based on census tracts. To facilitate recruitment by providing more easily identifiable target areas, they were expanded to include addresses in two entire zip codes: 53208 on the Northside of the city and 53204 on the Southside. The location of the target areas is illustrated in Figure ES.1. The majority of the population in the Northside area is African-American, while in the Southside area Hispanics predominate. In both areas there are more women than men, but the imbalance is somewhat greater on the Northside. Educational attainment is somewhat higher on the Northside; 66 percent have a high school diploma or a General Educational Development certificate (a GED), versus 57 percent on the Southside. Mobility is also substantial in both areas: One-third of Northsiders and one-quarter of Southsiders had lived at their current address less than a year when they applied to New Hope.

At approximately the end of New Hope recruitment, the circumstances of residents in the New Hope target areas were assessed with the New Hope Neighborhood Survey (NHNS), a general household survey of a random sample of more than 700 respondents from the two New Hope zip codes. Although substantial parts of both the Northside and Southside recruitment areas were economically depressed, the survey reveals that these neighborhoods contain many working residents and two-parent families. Nevertheless, one adult in four was jobless; among African-Americans the jobless rate was 47 percent. Almost 50 percent of the jobless residents reported that they were available for full-time work. About 70 percent of the jobs reported by employed residents of the targeted neighborhoods produced earnings in the range that made them eligible for the means-tested EIC. One-fourth of full-time workers and two-thirds of the part-time workers did not have health insurance. Few full- or part-time workers received assistance with child care.

The NHNS suggests that New Hope’s diagnoses of the problems confronting low-income workers and unemployed individuals in the target neighborhoods is relevant for a substantial portion of area residents. Many people appear to need jobs, child care, and health insurance. Using the NHNS, an estimate was constructed of the number of persons who fell within New Hope’s income eligibility rules and reported that New Hope would interest them "a great deal" if it were made available to them. Using this approach, one adult in four in the New Hope target area was judged a likely participant. Of these 12,400 people, 78 percent were jobless at the time of the survey and 59 percent live in households with children present.

The NHNS also identified some labor market difficulties facing residents that are not directly addressed by the New Hope program. Almost two-thirds of the 12,400 people referred to above lack a high school education. Eighteen percent of adults who reported being jobless but available for full-time work cited lack of transportation as a reason for not having a job. New Hope does not include skills training or transportation facilities; but project representatives are expected to refer participants to other agencies and programs to obtain these services if needed.

IV. Program Recruitment and Sample Characteristics

A. Recruitment Challenges

Recruitment of the sample was more challenging than staff had anticipated, partly because of the special circumstances of a research demonstration, which would not be present in an ongoing program. Achieving the sample goals required multiple recruitment tactics, a sustained campaign over many months, and expansion of the original target neighborhoods. Among the difficulties encountered were the constraint imposed by geographic targeting to two relatively small areas that could not be described easily; the resulting inefficiency of using many outreach tools such as newspapers, television, and radio that served the entire metropolitan area; residents’ unfamiliarity with New Hope; people "tuning out" new messages because of information overload; and the possible negative effects on word-of-mouth recruiting and willingness to enroll created by the research requirements, including the random assignment process.

Once contact was made, some people had trouble understanding or believing the New Hope offer; the arbitrary feel of the geographic restrictions, the unfamiliarity of the package of New Hope benefits, the complexity of the earnings supplements and copayment requirements, and the "too good to be true" nature of the offer all had to be overcome. Nevertheless, most of those who attended New Hope orientations and were eligible for the New Hope offer found it appealing and followed through with an application for the program.

In sum, the experience provides another illustration of the difficulty that new programs face in establishing themselves as "known quantities" within low-income communities. A telling measure of this difficulty comes from the NHNS: Even in the immediate aftermath of the recruitment campaign, 86 percent of eligible residents reported that they knew nothing about New Hope.

B. The New Hope Research Sample

The New Hope research sample (678 program and 679 control group members) was recruited over a 16-month period starting in July 1994. The recruitment effort led to a diverse sample, as the program operators desired. Table ES.2 summarizes the characteristics of applicants in the research sample.

Applicants included people who at random assignment were employed and unemployed; on welfare and not on welfare; living alone, with children, and/or with a spouse or partner; and from different racial or ethnic groups. Nearly everyone in the sample had work experience. However, all had low earnings (97 percent had earned less than $15,000 in the prior 12 months); and 71 percent had used some type of welfare or Medicaid in the previous 12 months. Forty-three percent of the sample lacked a high school diploma or GED.

Persons who applied to New Hope often indicated (in focus groups and conversations with project reps) that they were ready to make a positive change in their lives. In addition, applicants were often recruited from other service organizations in the community, implying that there may be a high level of participation in employment and social service programs other than New Hope by both program and control group members. This reinforces the importance of documenting participation in the program, comparing it with participation in other programs, and conducting an impact analysis of key program outcomes.

The research sample appears representative of the eligible residents in the target neighborhoods. Comparison of the characteristics of NHNS respondents interested in and eligible for New Hope services with actual program applicants in the research sample reveals few major differences, and most of these are attributable to specific strategic recruitment choices (for example, maintaining rough equality between the Northside and Southside samples or emphasizing inclusion of single individuals).

V. Program Operating Experience

All of the New Hope components — the earnings supplements, health insurance, child care assistance, and CSJs — were implemented and readily available to those assigned to the New Hope program group. There is no typical New Hope participant; in fact, the program is designed with an expectation that people will use the program in different ways. However, describing how the program works in general and for several hypothetical participants is useful in conveying how participants interacted with and used the program.

A. Experiencing New Hope

With few exceptions, participants access New Hope benefits and services by talking with the project representatives (project reps), who see their role as encouraging maximum use of these benefits and services to raise participants’ household income and improve their future economic prospects. Many participants seek only one or two of the New Hope benefits; the earnings supplement, for example, is used by virtually everyone active in the program. Others do not fully understand the various components of the program or how they can use them. Project reps try to make participants aware of their options and inquire regularly about changes in employment or family circumstances that might cause participants to need different benefits or services than they had in the past. Reps also serve as informal counselors and as "coaches" when people are searching for employment, providing leads on jobs and help in developing employment plans and résumés. In these roles, many of the project reps are able to draw on personal experience, having an "I have been there" credibility. For many participants, the help and encouragement offered by project reps is reported to be as helpful as the financial benefits offered by the program.

People working 30 hours or more per week are eligible for the earnings supplement and health insurance and child care. Those not working full time conduct an individual job search, with some assistance from project reps, to find qualifying employment. If they do not find full-time work after a search of eight weeks, they can interview for a CSJ that pays the minimum wage and that allows them to access other New Hope benefits. If they have been working and lose a job, a three-week job search is required prior to the offer of a CSJ. Staff have developed more than enough CSJ slots in various nearby nonprofit agencies for participants to choose from, but participants have to interview for the jobs, be selected by employers, and meet the attendance and other standards expected of regular employees. About 40 percent of CSJs are office support or data entry, 30 percent are construction and property maintenance, and the remainder are spread over a wide range of occupations.

Once participants are working and eligible to take advantage of New Hope’s financial incentives, the project reps’ role includes benefit processing. To qualify for financial benefits, New Hope participants have to provide proof of full-time employment by the fifth of each month. Reps review the pay stubs submitted to determine hours and earnings, and use worksheets and automated payment schedules to calculate the amount of benefits (earnings supplements and subsidies for health insurance and child care) that participants are to receive. Benefit processing is done on a monthly basis with payment made by the twentieth of the month following employment so that the amount of work and earnings will be quickly reflected in participants’ benefits.

The child care and health insurance assistance provided by New Hope is largely a financial transaction. Participants must find a qualifying child care provider they like; New Hope does not run its own child care center, nor do staff refer participants to specific providers. Payments can be provided to any state licensed or county certified provider, and the participant is required to pay a portion of the cost of child care through a copayment, adjusted based on family size and income. New Hope reimburses providers up to the same maximum level as the county provides for welfare recipients enrolled in work programs.

Of the benefits offered by New Hope, health insurance is mentioned as the most important by many participants and staff. While some participants are covered by employer health insurance or Medicaid, for those without coverage, the New Hope benefit is often the only affordable option. Participants working the required hours and not covered in another way can enroll in a health maintenance organization (HMO) that provides comprehensive services. Most choose the HMO that is used by the Milwaukee County Medicaid program. The participant copayments are set to reflect income and household size and are intended to fall within the range of the premiums that workers in many employer-sponsored plans pay.

Staff learned that it took continued effort to educate participants about the benefits and services available and to help participants understand how to use New Hope when their needs and circumstances changed. Despite these efforts, many participants had difficulty understanding how the benefits and services worked. Participants had the most difficulty understanding how earnings supplements were calculated, especially because of fluctuations in supplement checks. Differences in earnings from month to month often occurred because of differences in the number of pay days in a month or changes in hours worked. Former welfare recipients often were uncertain how New Hope supplements worked because they were used to relatively stable monthly welfare grants. Participants also had some difficulty understanding how health insurance and child care assistance would be affected if they lost a job or had a cutback in hours of work.

B. Illustrative Cases of New Hope Participants

Rather than creating a set sequence of services, New Hope designers created a collection of services and benefits that they believed would serve the needs of people in a variety of circumstances. The following three examples, two defined based on use of New Hope and one for a group often excluded from income support programs, illustrate the varying ways in which people use the program.

• Steadily employed full-time workers: About one-third of participants entered New Hope already working full time. Nearly three-quarters of these participants are women, about one-fifth were living with a spouse or boyfriend/girlfriend, and about three-fourths had children who lived with them. About one-fourth were receiving AFDC at entry into New Hope, and about one-fourth had earnings of $10,000 or more in the prior year.

• Unemployed persons without recent work experience: About one-fourth of the sample entered the program unemployed and with no earnings in the prior year. In terms of gender, age, race, parental status, and household composition these participants were quite similar to those who entered the program with a full-time job. However, only 83 percent had ever been employed and 66 percent had ever held a full-time job. Further, rates of receipt of public assistance were higher and education levels lower than for those working full time.

• "Single" men: About one-sixth of the sample is made up of men who are living with neither spouse nor other partner and without dependent children. Members of this group have traditionally been excluded from many public assistance programs, but are eligible for New Hope if they meet income and willingness-to-work tests.

C. The Use of New Hope Benefits

In program evaluations the use of program services is often of interest, but it is central to the New Hope story. In many other programs designed to help people find work (for example, training programs and job clubs), participation in the program is still one step removed from the outcome of central interest: employment. In New Hope, work is an eligibility rule for the program. For people to receive the New Hope benefits, they have to work full time. Hence, information on receipt of program benefits also conveys early information about the level of full-time work for those in the program group.

Table ES.3 provides summary data on benefit use for the portion of the New Hope program group for which 12 months of post-random assignment follow-up is available. Seventy-four percent of New Hope participants received at least one New Hope benefit at some point during the 12-month follow-up. Earnings supplements were used the most (by 72 percent of the program group), followed by health insurance (38 percent), CSJs (24 percent), and child care assistance (23 percent).

In interpreting these use rates, it is important to remember that New Hope is designed so that participants can access only those benefits that they want or need. Participants who are covered by employer health insurance, for example, do not need New Hope’s health insurance. Participants who had been receiving AFDC are encouraged to use transitional Medicaid and child care assistance before using New Hope’s benefits. About 30 percent of the sample lived in a household without children and therefore had no need for child care. It is also important to remember the labor market context in interpreting the CSJ use; the strong local economy meant that most participants found jobs in the private economy.

Once people moved beyond what might be considered a start-up period (the first three months after random assignment when unemployed applicants could find a job and qualify for benefits), approximately two-fifths of the program group used some type of New Hope benefit in a given month of follow-up. In this post-start-up period, about one-third of the entire program group used at least one New Hope benefit continuously or nearly continuously, about one-third used a benefit intermittently, and about one-third did not use any benefit.

Among subgroups:

• Those who were working at entry into the study, and especially those working full time, were more likely to access New Hope benefits, and used these benefits for more months on average.
  
• Of applicants with children (about 65 percent of whom were receiving AFDC at application to New Hope), those with access to a car and those with a high school credential were more likely than those without these characteristics to use benefits.

CSJs were intended to be the job of last resort for participants and tended to enroll lower-skilled and less-experienced individuals. Twenty-five percent of the participants who used CSJs moved directly into full-time, unsubsidized employment. The remaining 75 percent quit or left for personal reasons, were terminated by employers, or reached the CSJ time limit (six months in a placement and a total of 12 months overall) without finding unsubsidized work. But about half of those who left a CSJ without other employment found full-time unsubsidized work that qualified them for New Hope benefits at some later point in the 12 months of follow-up. Thus, about 62 percent of those working in a CSJ did make a transition to unsubsidized, full-time work during the 12-month follow-up.

A full explanation of why some in the program group did not use New Hope benefits and services will have to await completion of follow-up surveys with program group members. Among the known reasons, the two most common were that participants moved out of state or dropped out of the labor market to pursue schooling or become homemakers. In most instances, the reasons for nonparticipation are not clear. It could be that these individuals do not understand New Hope’s eligibility rules, decide to use the program only as "insurance" when they experience a job loss or other problem, have had negative experiences with the program, or have income exceeding program eligibility guidelines.

Whether these results are good or bad news for New Hope is hard to tell at this point. Complete information is not yet available on the employment behavior of the program group, nor is any information on the employment and service use of the control group outside New Hope (especially on child care and health assistance) ready to analyze. The follow-up survey currently in the field will yield information on why program group members did not use New Hope in months of nonuse.

The results presented in this report suggest the importance of recognizing that people do not use a program like New Hope in a simple way: Few of the program group members joined the program and immediately started participating, used the benefits continuously, and moved off the program permanently to "self-sufficiency." (Longer follow-up beyond the current 12 months will reveal the percentage leaving the program because their income has increased above program limits.) Instead, the use of benefits is likely to be much more complex and "nonlinear." Just as people go on and off welfare, get and lose jobs, and move into and out of poverty, their use of New Hope benefits will change to reflect these dynamic elements in their lives that affect their use of the New Hope benefits. Policymakers need to anticipate this pattern of use in work-based programs like New Hope, which fill the gap between earnings from available private market jobs and the poverty level and provide employee benefits not otherwise obtainable. Program designers and operators need to plan for multiple entries, exits, and spells of activity.

Assessment of the net effect of the New Hope offer on the likelihood of employment, movement to self-support, and movement out of poverty of program group households awaits accumulation of more data and comparison of outcomes between program and control groups. This comparison will be the subject of a later New Hope report.