CT FEATURE

By Pamela K. Lattimore

I

n 2003, the federal government funded all 50 states, the
District of Columbia and the U.S. Virgin Islands to develop and implement programs to facilitate the reentry of
individuals convicted of “serious and violent offenses”
who are returning to their communities following prison
stays. The Serious and Violent Offender Reentry Initiative
(SVORI) was a landmark undertaking not only because it
entailed an investment of more than $100 million in federal
grant funds but also because of its focus on more serious (as
opposed to less serious) inmates.
This second point is important because earlier federal
programs often focused on providing services to those who
were perceived to pose less of a threat to public safety (e.g.,
the Residential Substance Abuse Treatment programs for
first-time drug offenders) or increasing the likelihood and
terms of incarceration (e.g., the Violent Offender Incarceration/Truth in Sentencing program). Since the passage of
SVORI, Congress has passed the Prisoner Reentry Initiative,
which focuses on employment-based programming for
inmates, and the Marriage and Incarceration Act, which

focuses on family programming for adult male inmates. The
Second Chance Act was considered, but failed to pass, during
the 109th Congress.
After 20-plus years of focusing on deterrence and incapacitation, SVORI and these other federal initiatives are
important because they resurrect rehabilitation as a goal of
correctional policy. Underscoring these initiatives is a resurgence of belief that there are programs, services and
approaches that work in reforming the criminal proclivities
of inmates. And, indeed, there is a growing body of evidence
suggesting that reform is possible, albeit with relatively modest success. In fact, most of the successful programs identified by the 2006 Aos study, titled Evidence-Based Adult
Corrections Programs: What Works and What Does Not, are
associated with effect sizes of 10 percent to 20 percent, suggesting, for example, a 10 percent to 20 percent reduction in
rearrest rates (see the work of Steve Aos and colleagues at
www.wsipp.wa.gov). However, even such modest reductions, if broadly realized, imply substantial crime reductions
when applied to the millions of arrests each year.

Before looking further at current knowledge about what
works with reentry, it is important to address two areas that
are often acknowledged only implicitly, if at all. First, any discussion of inmate reentry should begin with an explicit
acknowledgement and description of the factors that many
individuals must overcome to be successful post-release. Consistent with this first point, policy-makers, practitioners and
researchers must also acknowledge that developing, implementing, assessing, revising and improving reentry programs
will require a long-term, sustained effort in contrast to the
sporadic patchwork of initiatives and programs that have
been implemented in the past.

Recognizing and Respecting
The Problem
The individuals incarcerated in U.S. juvenile and adult prisons have multiple deficiencies and problems likely to hinder
their ability to successfully transition from institution to a
crime-free life in the community. As detailed by many
researchers, these deficiencies include little education, few
job skills, little job experience likely to lead to good employment, substance and alcohol dependency, and other health
problems, including mental health problems. In addition, their
family and friends are often involved in crime and substance
abuse, and they disproportionately return to neighborhoods
with few economic opportunities and few, if any, positive role
models. Finally, each must cope with a criminal record that
can stand in the way of opportunities following release. While
policy-makers, practitioners and researchers are well aware of
these barriers to successful reentry, a failure to explicitly
acknowledge the length of this list may lead to unrealistic
expectations about the prospect of finding the magic bullet
that will “cure” recidivism.
What is less often acknowledged and discussed is that
these returning inmates have already been through (having
failed or been failed by) the other social and public institutions and agencies designed to produce good citizens. In particular, many of those who end up incarcerated did poorly in
the school systems that provide educational foundations for a
successful adulthood. Many offenders have histories of abuse
and neglect and may have been referred to, or in the custody
of, family and social services. Adult inmates often have histories of juvenile confinement and adult probation that failed to
provide the services, programming and support to reform and
rehabilitate. And finally, many inmates have received alcohol
and drug treatment outside the criminal justice system, but
may remain addicted to drugs and alcohol. It is worth recognizing that those held by the juvenile and criminal justice systems have fallen through most of society’s safety nets by the
time they end up behind bars. At this point, responsibility for
helping these offenders become productive citizens is handed
to underfunded agencies whose primary responsibility is public safety — not rehabilitation.
Judging by the ever-burgeoning prison population, which
continued to grow as crime rates and arrest rates for index
crimes declined — everyone has failed. According to the
Bureau of Justice Statistics, within three years, about twothirds of prison releases are arrested for new offenses, about
one-third are reincarcerated for new offenses and about half
are reincarcerated for either a new offense or parole violation.

What Has Gone Wrong?
Current methods have not worked for various reasons:

Fixing the problem on the cheap. BJS estimates that in
2001 state governments spent on average $22,650 per inmate
per year on operating costs. Although $20,000 per year
seems like a lot of money, it covers primarily security —
which protects the public from the inmates, the correctional
officers from the inmates and the inmates from each other
— and not programming. (Some states spend much more,
either because of higher costs in places like New York and
California or because they are, in fact, providing programs
and services.)
Vacillating on how the issue is dealt with. As policy and
politics reflect (or in some cases lead) public opinion, the
goal of criminal justice policy swings from rehabilitation to
deterrence to incapacitation. Thus, the war on crime and the
war on drugs followed a judgment that rehabilitation was
not attainable. These “wars” led to higher likelihoods and
longer terms of incarceration for some offenses, as well as
the imposition of numerous “collateral punishments” — for
example, losing access to Pell grants, public housing, children — that further burden returning inmates.
More recently, there has been a renewed focus on rehabilitation, albeit perhaps without an accompanying decline
in the efforts to be “tough on crime.” Thus, in fiscal year
2003, the federal government through the Department of Justice, provided more than $100 million to state and local government agencies under SVORI to develop programs for
returning inmates.
The SVORI programs were intended to identify and
address all of an offender’s needs and provide the appropriate services and programming. SVORI funds were intended
to supplement existing funding from correctional agencies,
other appropriate agencies, and faith-based and communitybased organizations. However, $100 million represents less
than $200 for each of the more than 600,000 individuals
released to parole each year. Further, the SVORI funds were
spread over three years.
Since SVORI was recently rolled out, Congress did not
reauthorize it, but instead chose to authorize the Prisoner
Reentry Initiative. This initiative is managed by a different
lead federal agency (the Department of Labor), excludes violent offenders and is focused largely on employment-related
issues.
The latest funded program is the “Marriage and Incarceration” initiative, which is managed by the Department of
Health and Human Services, not the Department of Labor
or Justice. This $30 million initiative is focused on
strengthening marriage and families among male correctional populations.
Most recently, the 109th Congress debated but failed to
pass the Second Chance Act. Although this legislation is
likely to be revived by the 110th Congress, if it passes it is
probable that the funding stream will head somewhere else
(e.g., to different populations, different services).
The federal government is trying to provide leadership
on this important issue. However, each time the funding
stream changes, state agencies must develop new programs
or retool existing programs to meet the objectives of the
new initiative. Given the complexity of the issues and the

difficulties faced by returning inmates, the current approach
to rehabilitation would be like the National Institutes of
Health changing its approach to cancer treatment every two
or three years.
Although this analogy may be a bit strong, it is clear that
the challenge of addressing the many needs of inmates
requires a concerted, sustained effort — one that is based
on a continuous improvement model of development. Such a
model should include technical assistance to state and local
governments to help them identify appropriate evidencebased practices, monitor implementation and measure outcomes to ensure effectiveness. This monitoring will help
agencies improve nascent programs, adjust them for their
populations and modify their approaches if they are detrimental. Such an effort will require many years of sustained,
applied work.
Having unrealistic expectations. Given the number and
seriousness of the problems inmates face in reentering the
community as productive citizens, the nature and level of
the response to these problems, and the lack of knowledge
about what to do to fix the problems, it is unrealistic to
expect to find a “silver bullet.” But a silver bullet is what one
looks for when new programs are sometimes implemented
with minimal funding to the rigor of scientific evaluation,
using stringent statistical standards to ensure that the new
program works.
Poorly designed evaluations that imply programs are
reducing recidivism by 50 percent or even 90 percent further
complicate efforts of agencies and evaluators to find what
works. These evaluations are usually based on comparing
outcomes for program completers with those of program
dropouts. These evaluations can illustrate only that people
who fail at one thing are more likely to fail at another when
compared with those who succeed, not whether the program was successful. In the meantime, however, many are
left believing that substantial reductions in recidivism are to
be expected.

What Works?
Aos and his colleagues at the Washington State Institute
for Public Policy have conducted one of the best reviews of
the current evidence of the effectiveness of adult correctional programs. Their review of 291 rigorous evaluations of
adult correctional programs found some approaches that
appear to be working in terms of reducing recidivism and
other programs that were less successful. According to Aos
and his colleagues, this is what works:
•
•
•
•
•
•

Adult drug courts;
In-prison therapeutic communities;
Cognitive-behavioral drug treatment in prison;
Community-based drug treatment;
Jail-based drug treatment;
General and specific cognitive-behavioral programs in
prison;
• Cognitive-behavioral treatment for sex offenders (in
prison or in the community);
• Treatment-oriented, intensive community supervision
programs;
• Correctional industry programs in prison;

• Basic adult education programs in prison;
• Employment training and job assistance in the community; and
• Vocational education programs in prison.
Approaches for which additional research and development are needed include:
• Case management in the community for drug
offenders;
• Therapeutic communities for mentally ill offenders;
• Faith-based programs;
• Domestic violence courts;
• A variety of approaches for managing sex offenders in
the community, including intensive supervision, mixed
treatment, medical treatment and faith-based
treatment;
• Regular parole supervision vs. no parole supervision;
• Day fines rather than traditional probation; and
• Work release.
Finally, Aos and his colleagues identified a number of programs that the evidence suggests are not effective. These
include some popular programs, such as boot camps and
electronic monitoring.
Thus, there is a good list of approaches that work or
appear promising. But, what does “work” mean? Generally, it
means reducing recidivism by 5 percent to 15 percent. Only
cognitive-behavioral treatment in the community for sex
offenders and treatment-oriented, intensive supervision
appear to reduce recidivism by greater amounts (31 percent
and 22 percent, respectively).
As a result, it is important that expectations are in line
with what has been proved so far. Are these levels the best
that can be done? Most likely these levels provide support to
continue focusing on these types of programs in an effort to
try to enhance and improve them.
A particular concern to many researchers is the issue of
“dosage.” How much treatment is needed to be effective?
Delivery of complete programs (e.g., drug counseling two or
three times a week for three or six months) is difficult for
offender populations for a variety of reasons. Within institutions, safety and operations drive activities. During lockdowns, programming stops; thus, there is no treatment, no
education and so forth. Further, population movements
between institutions traditionally are based on institutional
capacity, security levels and offender security classifications
— not on whether particular inmates are getting treatment
or participating in programs. For offenders being supervised
in the community, transportation, employment and family
responsibilities interfere with access to regular treatment
and programming.
Therefore, it is difficult to provide a set course of programming or treatment, making it difficult to determine
whether that amount of treatment is sufficient to be
effective. More attention to dosage (and to quality of services)
is needed. This is particularly important for programs that
have been shown to be effective or that appear promising.

Preview of the Findings From
The Multisite Evaluation of SVORI
Elsewhere in this magazine, Christy Visher and her colleagues present some implementation findings from the
SVORI evaluation (see pages 30-35). These results focus on
how the program directors have described their programs.
The SVORI evaluation team has also begun to analyze interview data collected from more than 2,500 individuals. These
data identify what the offenders reported needing and
receiving in terms of services, programs and other assistance. The evaluation team also is beginning to examine
interview data that indicate how SVORI participants are
doing on a range of intermediate outcomes (measured at
three, nine and 15 months following release) compared with
those who did not participate in SVORI programming.
Briefly, what has been found is:
SVORI programs have attempted to address the initiative’s goals and provide a wide range of services. Evaluators
have measured 28 categories of services prerelease and
post-release, as well as two additional services (supervision
and transportation) that could be provided post-release.
Most programs claim to be providing a wide range of
services across these 58 categories.
In addition, SVORI participants are getting more
services and programs than those who are not participating
in a SVORI program — a necessary but not sufficient condition for testing the effectiveness of SVORI. However, the levels
are far less than 100 percent and some of the nonparticipants

are also getting the same or similar services (albeit at a lower
rate), presenting a substantial evaluation challenge.
Finally, SVORI participants are doing better than average
on a vast array of outcomes — if sometimes only moderately. A total of 116 outcomes were measured at three
months. On average, SVORI participants were doing better
on 98 of these 116 measures. Not all of these differences
were statistically significant, but the odds of 98 of 116 outcomes being positive if SVORI were not having some type
of effect approaches zero.

Going Forward
The SVORI evaluation will provide considerable data and
information on both the challenges and successes of reentry
programming. This evaluation, like the initiative itself, is only
the beginning. As the criminal justice community moves forward, it must always keep in mind the substantial number of
problems that must be addressed and understand that these
problems will yield only to a sustained, long-term effort to
identify, implement and improve solutions.

Pamela K. Lattimore, Ph.D., is a principal scientist for RTI
International and is co-principal investigator for the NIJ-funded multisite evaluation of SVORI.