Journal Issue: The Next Generation of Antipoverty Policies Volume 17 Number 2 Fall 2007
The second important authority structure for low-income men is the criminal justice system. Ex-offenders leaving prisons need work to rebuild their lives. Traditionally correction systems have sought mainly to incarcerate offenders. Work programs aimed at this group have not achieved much, but they could be revamped to promote successful reentry and reduce recidivism.
During the 1960s and 1970s, as more poor mothers went on welfare, more poor men committed criminal offenses, and crime rates soared. The nation responded by sending more offenders to prison for longer terms. Rates of incarceration went on increasing even during the 1990s, after crime rates had started to fall. More than 2 million people are now in prison or jail. The problem is most severe among blacks: probably 30 percent of young black men have criminal records. Crime also overlaps substantially with the child support problem: probably 70 percent of male offenders are also noncustodial parents.63
As more men enter prison, more also leave. Around 630,000 men now exit the prisons annually— four times more than in 1978. Few convicts learn meaningful skills while in prison, and few reliably reintegrate into the community upon their release. Recidivism runs high. Thirty percent of released men are arrested again for new offences within six months; two-thirds within three years.64
Ex-offenders must reconnect with families, handle various health problems, and find housing. Failure at any of these hurdles can drive them into homelessness or addiction, or back into crime. In the long run, however, whether they stay free depends more than anything else on whether they work steadily.65 Just as for other men, success—or failure—at work stands at the center of their lives.
Criminal justice has few successful model programs on which to draw. During the 1970s, vocational programs in prisons appeared to have no effect on recidivism, prompting the conclusion that “nothing works.” Later assessments have been more positive, but even the better programs reduce recidivism by only 8 to 17 percent, and only a minority of inmates receives remediation in prison.66 Prison-based rehabilitation appears to achieve little, in part because it takes place inside the walls, removed from the conditions ex-offenders face out in the society.
Work programs for convicts outside the walls would appear more promising, but experimental programs have not yet shown effects comparable to those of work programs in welfare. Ex-offenders were one of several groups served by the National Supported Work Demonstration run by the Manpower Demonstration Research Corporation (MDRC) during the late 1970s. This study placed disadvantaged job seekers in positions created in local nonprofit agencies. It improved employment for welfare mothers and former drug addicts, but not for exoffenders or youth. The ex-offenders did increase their work while they were in the program, but they also left quickly, after which the effect dissipated.67 During 1991–94 a program for convicts on work release in Washington State failed to reduce recidivism or costs, although it did help a minority of men transition from prison.68
The parole system oversees convicts who leave prison before their sentences end. Parole officers typically require clients to meet with them once or twice a month and to take drug tests, among other rules. These requirements look like the sort of oversight that has generated strong impacts in welfare work programs. Yet by itself parole does not reduce recidivism. Even intensive supervision serves mainly to detect more violations of parole conditions such as drug use.69 Some experts have concluded—adapting a phrase from welfare reform—that we must “end parole as we know it.”
In 2005 the Bush administration funded Ready4Work, a set of seventeen voluntary demonstration programs aimed at prisoner reentry. One goal was to involve faith-based groups. Because the programs are serviceoriented and do not enforce participation, they presume a motivated client.70 Perhaps they will do some good, but they are unlikely to produce significant change. Nor can any effect be proven, because no experimental evaluations are planned.
Improving Work Enforcement
Work programs for ex-offenders can be improved. The chief focus of parole supervision has been to detect parole violations. Changing clients’ work behavior has been secondary. To affect recidivism, a new program must combine parole with demands that clients participate in programs aimed at their problems. To promote employment, supervision must be targeted much more specifically on working, and it must be more immediate. The supervisor must monitor actual work or job search and must have some quick way to reward good behavior and penalize bad.71 The precedent is drug programs, where swift and certain, not severe, punishment is what promotes compliance.72
Besides better supervision, a second necessity is help in finding work. Low-paid jobs clearly are available, and most ex-offenders already find them on their own. But their work rates fall with time, and unemployment runs high.73 Most employers admit their reluctance to hire former convicts.74 The danger is that some reentering offenders will take too long to find work, become discouraged, and return to crime. So a reentry work program must ensure work for its clients in some way. Equally, it must deny to men who might resist taking menial positions the excuse that jobs are unavailable. Christopher Jencks has argued that if jobs could be guaranteed to the jobless adults of the ghetto, community pressure on them to go to work would become far more effective.75
But did not guaranteed jobs for ex-offenders fail in the past work programs just mentioned? Yes, but National Supported Work was voluntary. Those ex-offenders had finished their sentences and were no longer under correctional authority. In the program I am proposing they would be on parole and would have to work or return to prison. The difference from the Washington State work release program, which was mandatory, is that supervision would be far more work-focused.
A third element needed is orientation to the demands of working. Even if training is not generally effective, men who have lived behind bars need some instruction about the demands of the workplace. Fourth, they need some help dealing with other problems in their lives, such as health, housing, and relations with their families. A reentry work requirement should thus initially be part time, allowing time to address these other problems. In both New York City and Wisconsin, mandatory work assignments for welfare mothers have been less than full time, to accommodate remediation activities.
One program combining these four elements is the Criminal Justice Program run by America Works (AW) in New York City.76 America Works here applies to men the same privatesector approach to work placement that it has used successfully with welfare mothers. Exoffenders are given an intensive orientation, lasting up to six weeks, on getting a job and working, including interviewing, dress, and behavior. They are then placed by sales representatives in private firms that recruit lowskilled labor from AW. Once placed, AW “corporate representatives” visit the clients on the job, talk to the employers, and help to work out any problems that the new hires may have.
Thus, work is arranged and overseen, although jobs are found privately rather than created. Clients also receive preparation for work and help in working out difficulties. America Works is financed largely through incentive payments. In the evaluation in New York described below, AW receives $1,160 from New York State for each initial job placement, then $2,088 for each placement that lasts at least ninety days, then a final $464 for each that lasts six months or more, for a total of $3,712. In its first year, 2001, the Criminal Justice Program placed 78 percent of the clients who completed its orientation in jobs. Of these, 44 percent held their jobs for at least ninety days.
The program serves not only parolees who are referred to it but also other ex-offenders who choose the program themselves, food stamp recipients (who also face a work test), and men from New York City’s child support enforcement program. A version of the program serving only ex-offenders is now being evaluated experimentally by Public/Private Ventures. In this version, the orientation will be given to clients in prison, before they leave to come to the program in New York.
Center for Employment Opportunities
An alternative model is offered by the Center for Employment Opportunities (CEO), also in New York City.77 Parolees come to CEO from the state prison system. After receiving several days of preemployment instruction, they are assigned to work crews that CEO maintains through its Neighborhood Work Project (NWP). There they do maintenance and repairs for local government agencies. Their attendance, performance, and comportment are monitored daily, and they are also paid daily, which meets their need for immediate income. Pay is $6.75 an hour, the New York State minimum wage. Clients work full time, four days a week.
On the fifth day, they report to a Vocational Development Program (VDP), where they work with a “job coach” who instructs them on job interviewing and helps them straighten out personal problems that could interfere with working. After two weeks in NWP, they also see a job developer, who lines up interviews for them with private employers. Clients stay in NWP as long as is needed to get a regular job, with a limit of seventy-five days. After placement, they are followed up at thirty, sixty, ninety, and one hundred and eighty days. CEO’s job retention rate at six months has been about 40 percent, and it has recently begun tracking retention over a year.
CEO sells its programs to state parole officers in the city as a way to keep their parolees employed. It also serves youth returning from the state Shock Incarceration program (boot camp), as well as some offenders leaving city jails. It is funded mostly by the parole system, the agencies that hire its work crews, and other government agencies. It costs CEO $33,220 a year to provide a slot in its community work crews. Since an average of six clients will hold a slot in a year, the cost per client is only $5,537. Furthermore, these costs are largely defrayed by the income CEO earns from the agencies that employ its crews. The net cost is only $3,219 per slot, or $536 per client.78
The core program—NWP and VDP serving state parolees—is one of four now being assessed in MDRC’s Enhanced Services for the Hard-to-Employ evaluation. In addition, the Joyce Foundation has begun to evaluate a similar transitional work program for exoffenders at five sites in the Midwest.
Both CEO and AW arrange and oversee work, while providing work orientation and casework. But AW does not regard transitional jobs as necessary, whereas CEO does. America Works believes that only placing clients with regular employers can prepare them to work, that creating jobs in government is a waste of time and money. If clients fail, and some do, AW gets them further positions until they succeed. CEO, by contrast, sees a need for supported work. Ex-offenders must function for some period under conditions where serious work demands are made but standards are more lenient than in regular jobs, and supervisors accept a mentoring role.
Even if one accepts the need for transitional jobs, the CEO positions seem short, lasting at most seventy-five days. Positions in other work guarantee programs have lasted six months to a year or more, in part because more time was thought necessary to instill work discipline.79 Longer assignments might improve job retention after clients move on to private jobs. On the other hand, longer positions cost more, and many clients placed in public jobs for enforcement purposes leave them quickly. Average tenure in a government job is far less than the assignment. CEO finds that whether a client can succeed at work is usually settled well before seventyfive days.