Parole Supervision: Pursuing the Balance

M
Introduction

y first real job was in a parole office – “field
supervision unit” as they called it. As a VISTA
Volunteer developing transition programs
for prisoners reentering the free world, I saw
firsthand the complexities of reentry and the
parole officer’s role in the process. I learned immensely from my parole
and treatment staff colleagues. I was constantly impressed with their
ability to both sanction and encourage parolees, particularly in the
face of high caseloads and limited time, tools and resources. It was also
apparent, however, in my office – as around the country – that parole

officers were driven by making their contacts and monitoring compliance
with the many conditions of release. The ultimate goal – preventing
reoffending, breaking substance abuse habits, and, in the end, changing
parolees’ lives for the better – was often more elusive.
This job was an important, inspiring work experience for me,
cementing my long-term interest in criminal justice policy and,
specifically, the issue of prisoner reentry. Over the past 15 years I
have changed hats, moving from practitioner to researcher. The study,
entitled, Does Parole Work? Analyzing the Impact of Postprison Supervision
on Rearrest Outcomes,1 describes a recent attempt to assess the impact of
parole supervision on recidivism. This article begins with an argument

by Amy L. Solomon

26

Perspectives	

Spring 2006

Parole Supervision: Pursuing the Balance

for why we should study supervision, followed by an overview of the
research. It concludes with some thoughts about policy opportunities
for the field, arguing that the current focus on prisoner reentry provides
a timely opportunity to “reinvent” parole.

Why Study Post-Prison Supervision?

states (Travis and Lawrence 2002). In some states, virtually all prisoners
are released to supervision; in others it’s less than half. Moreover,
different supervision practices are employed state to state. Some states
rely heavily on drug testing; others are focused on community-based
responses to parole violations. A few states are experimenting with
neighborhood supervision, others with Global Positioning Satellite
(GPS) tracking technologies. And many states are conducting routine
office visits as they always have. The bottom line is that parole practices
and policies vary substantially state to state and sometimes jurisdiction
to jurisdiction, providing a rich – if complex – research opportunity to
document which strategies work best.
Limited research exists on the topic of parole effectiveness.
Given the widespread use of parole and the diversity of practice, it is
remarkable how little attention has been paid to the impact of parole
on public safety. There have been a few studies comparing recidivism
outcomes of parolees and unsupervised ex-prisoners, but they tend to be
small, dated, or based in international settings (Ellis and Marshall 2000;
Gottsfredson and Mitchell-Herzfeld 1982; Jackson 1983; Nuttal et al.,
1977; Sacks and Logan 1980; Sacks and Logan 1979; Waller 1974).
Although these studies measure recidivism in different ways, most find
a small but statistically significant benefit from parole supervision in
terms of recidivism outcomes.
More generally, most of the larger, more rigorous correctional
studies and meta-analyses suggest that surveillance does little to
improve recidivism outcomes, unless it is coupled with treatment

There are many important reasons to study community supervision,
including:
Many people are on parole. Each year, over 650,000 individuals
are released from state and federal prisons across the country (Harrison
and Beck 2005). Most – about 80 percent – are released to supervision in
the community following their prison stay. Parolees spend an average of
26 months on post-prison supervision (Hughes et al., 2001), and at any
given time there are about 765,000 on parole (not to mention another
four million on probation) (Glaze and Palla 2005).
Failure rates are high. Less than half (46 percent) of all parolees
successfully complete parole without violating a condition of release,
absconding, or committing a new crime (Glaze and Palla 2005). As a
result, over 200,000 parolees return to prison each year (BJS 2000).
Nationally, parole violators account for about one third of all prison
admissions, and therefore account for a sizable fraction of many state’s
correctional budgets ( Jacobson 2005).
The way prisoners are released has changed substantially over
time. While the majority of prisoners used to be released by a parole
board, “discretionary release” has declined from about 55 percent of all
releases in 1980 to just 24 percent in 2000
Figure 1. Share of State Prisoners Nationwide Released Conditionally and
(Hughes et al., 2001). Mandatory releases
Unconditionally, 1980-2000
now account for about 40 percent of all
releases from prison, up from less than 60%
20 percent in 1980. Prisoners released
without supervision account for about
50%
one-fifth of all prison releases. (See
“Three Study Groups” below, for more
Conditional - Discretionary
discussion about each type of release.) It is 40%
unclear how this major shift in method of
release has impacted recidivism outcomes,
although the Bureau of Justice Statistics 30%
(BJS) studies indicate that more than
Conditional - Mandatory
half of discretionary parolees successfully 20%
complete their term of supervision
compared with one-third of mandatory
10%
Unconditional
parolees (Hughes et al., 2001).
Parole supervision is implemented
differently in each state. The use, 0%
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
duration and intensity of post-prison
supervision varies significantly across
Source: Hughes, Wilson and Beck, “Trends in State Parole, 1990-2000” and National Corrections Reporting Program

	           A m e r i c a n P r o b a t i o n a n d P a r o l e A s s o c i a t i o n 	

	
27

Parole Supervision: Pursuing the Balance

“Does Parole
Work?
A Word on the Title
There has been a good deal of

criticism of our study, much of it

centered on the title itself. Why?

Probably because the first half of

the title, “Does Parole Work?,” begs

a one-dimensional answer – YES or

NO – that grossly oversimplifies the

issue. I concede that point (see

“Limitations”). But I would also argue

that the question itself is not only a

fair question to ask, but a crucial one

– one that every state should be asking

itself. How can the criminal justice

community focus so much attention

on prisoner reentry and NOT demand

to know if post-prison supervision – the

biggest reentry intervention there is – is

contributing to public safety?

28

Perspectives	

interventions (Sherman et al., 1997; MacKenzie 1997; Petersilia 1998). Even intensive monitoring,
involving lower caseloads and more frequent contacts, does not produce reduced recidivism
(Petersilia and Turner 1993; Sherman et al., 1997). Taxman (2002) provides an excellent overview
of this literature.
In sum, there are large numbers of people on parole, high failure rates, substantial variation in
practice across states, and changes in release methods. At the same time, relatively little is known
about whether and how supervision increases public safety. To the study authors, this context begged
the question – the title of our study - Does Parole Work?

The Study
The study, Does Parole Work? Analyzing the Impact of Postprison Supervision on Rearrest
Outcomes, compares prisoners released to supervision, via discretionary and mandatory release, to
prisoners released without supervision. Using data from the BJS, we aimed to assess, at an aggregate
level, whether parole supervision “works” at reducing crime – as measured by rearrests – among
the parole population.
The study is organized around three key questions. First, do prisoners released with and without
supervision differ with respect to demographics, incarceration characteristics and criminal histories?
Second, do prisoners released with and without supervision recidivate at different rates? And third,
if so, for whom does supervision matter most?

Three Study Groups
The study tracks outcomes for three groups: Those released via discretionary release to
community supervision, those released via mandatory release to community supervision, and those
released unconditionally.
(1) Discretionary release involves a parole board decision to release a prisoner before he has
served his full sentence, serving the remainder of his sentence under community supervision. Parole
boards essentially screen prisoners and use their discretion to determine who is most “ready” to
return to the community. Parole boards may consider criminal histories, the incarceration offense,
institutional conduct, prisoner attitude and motivation, participation in prison programs and positive
connections to the community such as employment, housing arrangements and ties to family. In
this article, prisoners released by parole boards are referred to as discretionary parolees.
(2) Mandatory release occurs when a prisoner has served his original sentence, less any
accumulated good time credit, serving the balance of his sentence under supervision in the community.
Mandatory releasees have not received a determination of fitness to return to the community from
a parole board or other authority. This group is referred to as mandatory parolees.
Community supervision resulting from either discretionary or mandatory release is not
systematically different. In most states, conditions of supervision are similar for both types of
parolees, although discretionary parolees often spend more time on supervision.
(3) Unconditional releasees leave prison after serving their full term behind bars. These
individuals were not granted early release via a parole board in states retaining discretionary parole,
nor did they receive good time credit enabling mandatory early release. Therefore, unconditional
releasees exit prison without any conditions of release, community supervision or reporting
requirements.

Spring 2006

Parole Supervision: Pursuing the Balance

Data Sources and Methodology
Our study relies primarily2 on BJS data on 38,624 prisoners released in 1994 from prisons in
15 states. This sample is representative of the 272,111 prisoners released from those states in 1994
– two-thirds of all prisoners released nationwide in 1994. The states included in the BJS study are
Arizona, California, Delaware, Florida, Illinois, Maryland, Michigan, Minnesota, New Jersey, New
York, North Carolina, Ohio, Oregon, Texas and Virginia. Due to issues with the data, Delaware is
excluded from our analysis.
BJS tracked recidivism outcomes – rearrests, reconvictions and reincarcerations – for these
prisoners for three years after their release. We chose to use rearrest outcomes at two years post-release
instead of three to more closely mirror the average time on parole (26 months in 1999, Hughes et
al., 2001). The BJS findings resulted in their landmark report, “Recidivism of Prisoners Released in
1994,” by Patrick Langan and David Levin (Langan and Levin 2002).
We used descriptive analysis to address the first two questions – whom states release conditionally
versus unconditionally, and whether these groups recidivate at differential rates. To address the
remaining research question, for whom does supervision matter most, we utilized a combination
of multivariate regression and simulation analyses. For further information on our data sources
and methodology, please refer to the full study or a more detailed technical report, on file with the
authors.

An important note
for those interested
in replicating the
analysis: When coding the
original BJS data, we based many of
our decisions on protocols developed
by Richard Rosenfeld and Anne
Morrison Piehl, who were part of a
working group devoted to – and
resulting in a book on -- reentry and

Limitations
The BJS source data discussed above is by far the largest, most complete, most current dataset
that exists to address recidivism of prisoners. Because BJS had captured a variable indicating how
prisoners were released, and thus if they were supervised or not after release, it offered a rare research
opportunity to examine the different recidivism outcomes of prisoners released with and without
supervision. At the same time, the BJS recidivism study was not designed – and the data not collected
– to examine the impact of supervision on recidivism. Accordingly, it is not a perfect fit, resulting in
several limitations to our analysis.
Arguably the most problematic limitations are that the study could not address state-level
variation or identify, across states, what types of supervision strategies are most effective. Our reliance
on arrests as a measure of recidivism in lieu of actual offending is also less than ideal. Less challenging
are the critiques about old data and a universe of only 14 states. These issues are summarized below.
Our study does not address state-level variation. Ours was a multi-state analysis that described
a national-level story, when, as discussed above, the reality is that parole practices and outcomes vary
substantially across states, and even across jurisdictions within those states. The aggregate nature of
our analysis buries what are surely substantial differences at the state level relative to the outcomes
associated with parole.
The analysis could not address which types of parole strategies are more effective than
others. While the source data provided important information on the personal and criminal histories
of released prisoners, information on the nature of supervision – such as intensity of supervision,
length of supervision, reporting requirements and services received at the individual level – was not
available. Our data also did not include system-level data about risk assessment tools, contact standards,
caseload averages, case planning, case management strategies and neighborhood-based supervision
models. Without such information, we were unable to get inside the “black box” of supervision – to
consider how various types of supervision affect rearrest outcomes and assess what types of parole
strategies work better than others.

public safety (Travis and Visher 2005).
Allen Beck from BJS was also part of
this group, and early on he identified
problems with the codes for the
release type variable. Drs. Rosenfeld
and Piehl, in consultation with staff
at BJS, created “fixes” to account for
these coding errors and we followed
their example. Data from California,
Michigan and North Carolina in
particular had to be recoded. Details

are provided in the technical report.

	           A m e r i c a n P r o b a t i o n a n d P a r o l e A s s o c i a t i o n 	

	
29

Parole Supervision: Pursuing the Balance

The analysis uses rearrests as a proxy for actual reoffending.
The BJS collected three measures of recidivism – rearrest, reconviction
and reincarceration. We chose to focus on arrests as the closest proxy
to offender behavior because they involve the least amount of policy
interventions. Still, because rearrests reflect a combination of both
criminal activity and other decisions (e.g., to report a crime, to arrest
an individual, to revoke parole), those on supervision may be watched
more closely by law enforcement and parole officers. Thus criminal
activity committed by parolees may be more likely to be detected than
by unconditional releasees.
The study relies on data that are about ten years old. Our analysis
involves recidivism outcomes for individuals released from prison in
1994 and “tracked” for two years. With the emergence of prisoner
reentry as a major policy focus for the criminal justice community, the
corrections environment has certainly changed in the last ten years.
Yet it is not at all clear that parole supervision writ large has changed
dramatically in this time frame. While there are innovations occurring
in many parole agencies across the country, in most states these new

approaches are implemented more on the margins than the mainstream
of parole practice. Further, given the increasing demands on state
budgets, caseloads may be even higher and service resources lower than
was the case a decade ago. Accordingly, there is no reason to believe that
replicating the analysis using release data from, say, 2003, would yield
more favorable results.
The study includes data from only 14 states, e.g. the universe of
the BJS recidivism study minus Delaware. While true, taken together,
prisoners released from these 14 states accounted for about two-thirds
of all prisoners released in 1994 (Langan and Levin 2002). It is worth
noting that because California heavily influences national trends, we
re-analyzed the data including all states except California. These results
are reported in the “Findings” section of the article.

Findings

Do prisoners released with and without supervision have
different demographics, incarceration experiences, or criminal
histories?
As illustrated in Figure 2, there are
Figure 2. Characteristics of prisoners released in 1994, by supervision status at
statistical differences across groups, but
release
generally there are not large substantive
differences. The average age at release
Unconditional
Mandatory
Discretionary
for all three groups was 32 or 33 years
releasees
parolees
parolees
old, and the vast majority of releasees
was male. Just over half of unconditional
DEMOGRAPHICS
releasees and discretionary parolees were
Average age at release (years)
32.7
32.6
31.9
black, compared with about 40 percent of
mandatory parolees.
Male (%)
93
92
90
In terms of criminal histories, more
Black (%)
55
42
54
than 90 percent of each group had been
CRIMINAL HISTORY
arrested in the past. Unconditional
releasees and mandatory parolees, however,
Previously arrested (%)
93
94
92
had slightly higher average numbers of
Average number of prior arrests
9.6
9.5
7.5
prior arrests than discretionary parolees.
Previously arrested for violent offense (%)
67
63
55
We also looked at prior arrests for violent
crimes as an indicator of potential risk to
Prior incarcerations (prison or jail, %)
68
69
67
the community upon release. Larger shares
Average number of prior incarcerations
2.7
2.5
2.3
of prisoners released unconditionally
had previously been arrested for a violent
INCARCERATION CHARACTERISTICS
offense than had mandatory parolees, with
Incarcerated for violent offense (%)
27
21
23
discretionary parolees the least likely to
have been arrested for a violent offense
Incarcerated for drug offense (%)
30
31
34
in the past.
Incarcerated for property offense (%)
33
35
31
About two-thirds of each group had
Incarcerated for public order offense (%)
9
9
10
been confined to prison or jail in the past,
Average time served (months)
32.0
18.5
21.3
two to three times on average. In terms
Source: Urban Institute analysis of Bureau of Justice Statistics data.

30

Perspectives	

Spring 2006

Parole Supervision: Pursuing the Balance

of their most recent incarceration offense, about
one-fourth of each group had been incarcerated
for a violent offense, about one-third for a drug
offense, another one-third for a property offense,
and about ten percent for a public order offense.
Finally, unconditional releasees served substantially
more time behind bars, suggesting they may be more
disconnected from positive social networks than
their supervised counterparts.

Figure 3. First rearrest of prisioners who were rearrested at least once, by
supervision status at 1994 release
35
31%
30%

30

29%

Unconditional Releases
Mandatory Parolees
Discretionary Parolees

27%

25

25%
23%
21%

Do prisoners released with and without
supervision recidivate at different rates?
Sixty-two percent of unconditional releasees
were rearrested at least once over two years, compared
with 61 percent of mandatory parolees and 54
percent of discretionary parolees. Individuals in
each group had between two and two and a half
rearrests, on average, during the two-year period.
These findings mirror unpublished analysis by
BJS of a 1983 release cohort tracked for an earlier
recidivism study. BJS found that 62.3 percent of
conditional releasees were rearrested within three
years, compared to 64.8 percent of unconditional
releasees (Petersilia 2002).

20

19%

19% 19%
17%

16%

15

5%

5
0

10%

10%

10

Drug

First Rearrest Offenses
We examined offense types of those rearrested at least once. Similar
shares of all three groups were first rearrested for property offenses, while
a somewhat higher share of mandatory parolees were first rearrested for
drug offenses, and a slightly higher share of unconditional releasees were
first rearrested for violent crimes (Figure 3).
It is important to note that a small subset of public order offenders
were actually charged with technical violations as their new offense.
Most also had a concurrent charge for another offense, or were charged
with a new offense before or after the technical violation. In other
words, very few individuals (340 in the original sample) were rearrested
only for a technical violation, and arguably many of those involved an
underlying crime. In any case, since our study was published we have
re-analyzed the data, excluding all rearrests for technical violations.
Rearrest outcomes for unconditional and mandatory releasees barely
changed, if at all; rearrest rates for discretionary parolees went down
slightly (1.5 percentage points).
Comparing Similar Individuals
Because the three release groups were not identical on available
attributes, we conducted regression analysis to control for these

Property

Violent

Public Order

Other/Unknown

differences. The results indicated that when comparing two individuals
with similar demographics and criminal histories, their rearrest outcomes
— based exclusively on their supervision status — differed only slightly.
Specifically, when all other variables were controlled for, 61 percent of
both mandatory parolees and unconditional releasees were expected to
be rearrested at least once over two years, as compared to 57 percent of
discretionary parolees.
As noted earlier, we re-analyzed the data in order to determine the
extent to which California was influencing the results. The recidivism
findings change when California is excluded, but not dramatically:
The predicted probability of rearrest for unconditional releasees rises
to 63 percent, compared with 60 percent for mandatory parolees and
56 percent for discretionary parolees.
Interpreting the Differences
Mandatory parolees, who today account for the largest share of
released prisoners, fare no better with supervision than similar prisoners
released without supervision in terms of rearrest outcomes. While
discretionary parolees are somewhat less likely to be rearrested, this
difference is relatively small considering that parole boards are selecting
the “best risks” for release.
Clearly there is a value judgment being made here, in characterizing
a four percentage point difference as “relatively small,” differing “only
slightly.” In the criminal justice arena, where reductions in recidivism

	           A m e r i c a n P r o b a t i o n a n d P a r o l e A s s o c i a t i o n 	

	
31

Parole Supervision: Pursuing the Balance

Is Discretionary
Parole the Answer?
Whether one perceives a four
percentage point difference as
large or small, all can agree that
discretionary parolees were rearrested
at a lower rate than their mandatory
parole and unconditionally released
counterparts. That given, some would
contend that discretionary parole
should be re-expanded to more states
and prisoners (Rosenfeld 2005; Petersilia
2003). But discretionary release is
arguably a “solution” with a ceiling.

are stubbornly hard to come by, some will see this same difference and determine it evidence
that parole does work.
Because parole boards take into account factors such as a prisoner’s attitude and
motivation level, institutional conduct, preparedness for release and connections to the
community – important factors that our model could not control for -- I would expect this
group to be substantially, rather than marginally, less likely to recidivate. The suggestion
here is that the lower rearrest rates may be largely due to who is selected for discretionary
release rather than discretionary supervision itself, which is not systematically different
than mandatory supervision across states

For Whom Does Supervision Matter Most?
Certain prisoners appear to benefit more from supervision than others in terms of
rearrest outcomes. Specifically, females, individuals with few prior arrests, public order
offenders and technical violators were less likely to be rearrested if supervised than their
unsupervised counterparts (Figure 4). For example, the likelihood of rearrest for a female
parolee is 51%, as compared to 67 percent for a similar female released without supervision.
There is a similar pattern for public order offenses, although not as pronounced.
Those who had a combination of these characteristics – typically lower risk, lower
level offenders – yielded even greater benefits. It is possible that these individuals are more
responsive to the sanctions and services provided by supervision given their minimal prior
involvements with the justice system.
Conversely, supervision did not appear to improve rearrest outcomes for some of the
higher rate, more serious offenders – arguably those who warrant supervision most. The
Figure 4. Predicted probability of rearrest two years after release, by supervision status
at 1994 release
Unconditional
releasees (%)

Mandatory
parolees (%)

Discretionary
parolees (%)

OVERALL

61

61

57

Male

60

62

58

Female

67

51

51

Black

68

67

61

Non-black

54

56

53

Few prior arrests

53

49

44

Medium prior arrests

59

57

52

High prior arrests

68

70

66

Low release age

61

60

57

Medium release age

62

62

58

High release age

52

53

48

Violent offense

55

56

55

Property offense

68

67

62

Drug offense

56

61

54

Public order and other offense

65

57

55

New sentence

56

58

54

Revocation + new sentence

59

62

53

Revocation (technical)

71

68

63

By allowing parole boards to choose
the lower-risk, more-ready prisoners
for release, the implication is that
higher-risk, less-ready individuals stay
incarcerated longer. The unintended
consequence of this policy is that those
higher-risk, less-ready prisoners may
be released with little or no supervision
at the end of their sentence. In other
words, while discretionary release

Source: Urban Institute analysis of Bureau of Justice Statistics data. See “Methodology” section of original study for
definitions of characteristics shown in figure.

32

Perspectives	

Spring 2006

Parole Supervision: Pursuing the Balance

expected rearrest rates for those incarcerated for a violent offense is about 55 percent,
whether one is supervised or not. The likelihood of rearrest for a mandatory parolee with
high prior arrests is 70 percent, compared with 66 percent for discretionary parolees and
68 percent for unconditional releasees.
It is notable that technical violators released unconditionally have higher expected
rearrest rates than any other release group. The policy implication is that responding to
technical violations by reincarcerating violators for the remainder of their sentence does
not solve the problem. When these individuals are then re-released from prison without
supervision, they are highly likely to be rearrested – even more so than their counterparts
who are released to supervision.

How Does Supervision Affect the Largest Release
Groups?

enables to states choose lower-risk
candidates for release, it is still critical to
have an effective supervision component
for those higher risk individuals who may
never pass the “readiness” test and may
in fact warrant supervision most of all.
At the same time, there may be

Few prisoners have a combination of characteristics likely to yield either the highest
or lowest benefits from supervision. In fact, the public safety impact of supervision is
minimal and often nonexistent among the largest shares of the release cohort – males
convicted of property, drug and violent offenses who account for 80 percent of 1994
releases. As illustrated in Figure 5, supervision impacts rearrest outcomes differently
based on the incarcerating charge.
Specifically, supervision does not play much of a role among those incarcerated for
a violent offense (roughly one-fifth of the released population). Discretionary parole
does seem to benefit property offenders (roughly one-third of the released population),
although predicted rearrest rates for mandatory parolees are virtually the same as
Figure 5. Predicted probability of rearrest two years after release for largest release
groups, by supervision status at 1994 release

important lessons from the discretionary
release process that could be
transferred to post-release supervision.
For example, there may be ways other
than a parole board appearance to
stimulate good behavior and better
prepare a greater share of prisoners

PREDICTED PROBABILITY OF REARREST (%)
Percent of 1994
release cohort

Unconditional
releasees

Mandatory
parolees

Discretionary
parolees

Young males

11.7

68

67

62

Medium males

8.5

68

68

63

Older males

10.5

59

60

53

for release to the community. Jeremy
Travis (2005) introduces an innovative

PROPERTY OFFENDERS
twist on good time credits, suggesting
to transform it into something prisoners
must earn by participating in treatment

DRUG OFFENDERS
Young males

10.7

55

61

54

Medium males

8.0

55

62

55

Older males

9.7

45

54

45

VIOLENT OFFENDERS

and training and preparing for their
return to the community. This idea could
be implemented within the current

Young males

8.5

55

56

56

Medium males

5.4

55

58

57

Older males

7.4

45

49

47

Source: Urban Institute analysis of Bureau of Justice Statistics data. See “Methodology” section of full report for definitions of
age groupings.

legal framework and offers prisoners a
tangible, meaningful incentive to use the
time behind bars productively.

	           A m e r i c a n P r o b a t i o n a n d P a r o l e A s s o c i a t i o n 	

	
33

Parole Supervision: Pursuing the Balance

for unconditional releasees. On the other hand, the predicted rearrest
rates for drug offenders are the same for discretionary parolees and
unconditional releasees, while mandatory parolees actually have higher
rearrest rates than the other two groups

Discussion
Unanswered Questions
What these findings tell us is that the big picture warrants attention.
The analysis suggests that on balance, looking at a group of large states,
parole has not contributed substantially to reduced recidivism and
increased public safety. The public safety contributions of parole need
to be carefully examined and, more importantly, improved.
At the same time, the study does not conclude that parole can’t
work. In fact it may work quite well in certain states and jurisdiction.
But our study could not address how parole was practiced in various
states, nor its level of success in specific places. As discussed above,
parole practices and policies operate independently in each state, and
vary substantially from jurisdiction to jurisdiction. A rich research
opportunity exists to study various models and determine which
practices are associated with the best outcomes (Piehl 2005).
As discussed above, our study could not get inside the “black
box” of supervision. We did not have the data to address what types of
parole strategies work better than others. Specifically, we could not take
into account what impact various factors – such as the length, type or
intensity of supervision, assessment tools, access to programming and
treatment, caseload size and contact standards – had on recidivism
outcomes.
While the research did shed light on the types of individuals who
benefit most from supervision, the important next step is to figure out
why parole works for some better than others and how similar gains
could be realized for larger subsets of the parole population. Conversely,
the finding that the higher risk, more serious individuals benefit least
from parole supervision has confounded many experts who expect
higher risk individuals to be most impacted by supervision, as they are
by treatment. All of these issues warrant further study.
Supervision in Perspective
While the analysis is imperfect and the remaining questions
substantial, the implication that parole may not be particularly effective
at reducing reoffending should not come as a surprise to many in
the field. For years, community corrections leaders have alluded to a
“broken” system3 (Petersilia 2003:193) in need of “a major overhaul”
( Jacobson 2005: 148). According to Petersilia (2003:12), “No one
believes that the current prison and parole system is working.”
Little hard evidence exists as to why supervision may not be as
effective as it could be, but the realities of parole point to some clues.

34

Perspectives	

Spring 2006

To begin with, supervision in most cases is quite minimal. Parole
officers’ caseloads average 70 parolees apiece, translating to one or two
15 minute meetings a month (Petersilia 2003). While lower caseloads
do not ensure success (Taxman 2002), such high caseloads make it
virtually impossible (Rhine et al., 1991). Additionally, parole officers
are typically based in downtown offices far from the communities where
their parolees reside, and therefore lack the context and relationships
that neighborhood-based supervision – similar to community based
policing – could provide.
Supervision today is more surveillance-oriented than was once
the case, despite that research shows it takes a mix of treatment and
surveillance to change offender behavior (Petersilia 2003; Sherman
et al., 1997). Additionally, the response to parole failure is often a
failure itself. In many states responses to violations are inconsistent and
inappropriate to the seriousness of the infraction, therefore diminishing
any deterrent value and costing the public millions in reincarceration
costs ( Jacobson 2005).
Over the last decade several groups of practitioners and academics
have examined these issues, as well as the future of community
corrections. In the late 1990s, the Office of Justice Programs (Department
of Justice) held a two-day meeting of about 50 community corrections
practitioners to rethink community supervision and community safety.
They determined that the field was at a critical crossroads, facing both
“a moment of vulnerability” and “a moment of opportunity” (Dickey
and Smith 1998). Around the same time but over a longer period, a
group of a dozen prominent practitioners met under the auspices of the
“Reinventing Probation Council.” After three years of deliberation, they
issued an bold, candid, forward-thinking report, entitled Transforming
Probation Through Leadership: The Broken Windows Model (Reinventing
Probation Council 2000). It argued that probation should adopt a
community-centered, public safety-oriented approach similar to the
“Broken Windows” law enforcement model.
More recently, the Re-Entry Policy Council issued bi-partisan
consensus statements aimed at improving prisoner reentry. Some two
dozen recommendations address parole specifically, from the release
decision to responses to parole violations (Report of the Re-Entry
Policy Council 2005). Many of the ideas emanating from these groups
are also consistent with the sentiments of correctional experts who
were interviewed by Joan Petersilia (Petersilia 2002). According to
Petersilia, there is substantial agreement that a new supervision model
should be community-based, focus on the highest risk offenders, deliver
appropriate treatment as well as sanctions, and include an array of
intermediate sanctions in response to technical violations.
These policy discussions and interviews reveal a broad consensus
among seasoned practitioners and academics that community
supervision can and should work, but that change is necessary. Parole

Parole Supervision: Pursuing the Balance

has the potential to make a big contribution to public safety and successful reentry. The fact
that parole officers have the legal authority to set and enforce rules for a high risk population, to
coerce – and access – treatment and training for parolees, is largely under appreciated. To borrow
from the Reinventing Probation Council, “As a matter of social policy, [community supervision]
occupies the borderland between law enforcement and human services. As a justice system sanction,
[community supervision] is invested with wide-ranging leverage to influence the conduct of
offenders. Its strength lies in its authority and capacity to repair broken lives and hold offenders
accountable for the harm their actions have caused to victims and communities” (Reinventing
Probation Council 2000:3).
Policy Opportunities
This section outlines some broad opportunities for the paroling profession. None of these ideas
is particularly original. In fact, most echo recommendations of the groups discussed above. Some
of these proposals have research backing; others are testable and should be evaluated. While the
ideas themselves are straightforward, implementation would be complex and difficult, requiring
enormous change, especially in terms of organizational culture. Perhaps that is why more parole
agencies have not put into practice more of these approaches, despite the fundamental consensus
among many in the field. But the time to experiment with reinvention is now. If nothing else, the
Urban Institute study calls into question the efficacy of “business-as-usual.” And importantly, the
national policy interest in prisoner reentry affords a rare window of opportunity for parole to test
out new strategies in the name of improving prisoner reentry and reintegration outcomes.
Agency Level	
Starting at the top, parole should adopt a mission that puts public safety first (Reinventing
Probation Council 2000; Kleiman 2005). The field should be clear about its purpose and own the
recidivism problem, even if it is not responsible for all of it.
Parole agencies should operationalize this mission by setting – and being accountable for
– explicit public safety benchmarks (Reinventing Probation Council 2000; Kleiman 2005).
Following the lead of the policing profession, parole agencies should set performance goals that
aim to reduce reoffending rates by a specific amount. In the probation context, Beto, Corbett
and DiIulio (2000) suggest the goal that only 10 percent of all probationers commit a new crime
within three years. The Reinventing Probation Council argues that “embracing [such a] goal as a
benchmark against which to measure the performance of the field serves as a bold yet necessary
step in addressing the crisis afflicting probation” (Reinventing Probation Council 2000:6). While
that specific statistic may be unrealistic in the parole context – maybe the target is closer to 40 or 50
percent – defining success in such a way would be a sea change. It could both raise parole’s credibility
with the public and signal to line staff that controlling crime among parolees is possible.
Parole agencies should also take full advantage of what the research community has found
to be effective (Bogue et al., 2004; Burke 2004; Bureau of Justice Assistance 2004). As discussed
elsewhere in this journal, evidence-based practices represent a body of knowledge about programs
and interventions proven to reduce recidivism.4 Despite the empirical base, few agencies implement
these principles in their mainstream supervision practices.
Given the substantial treatment, health, housing, education and employment needs of
the parole population, it is also essential for parole to partner with other agencies – such as
community health care providers, housing authorities and workforce development boards – who
are now recognizing aspects of the reentry problem as their own (Report of the Re-Entry

Parolee Attitudes
Interviews with parolees for the Urban

Institute’s Returning Home study suggest

parolees are respectful of the their

parole officers, open to help, and

have high expectations about the

assistance they anticipate (La Vigne

and Kachnowski 2005; Visher et al., 2004;

Visher et al., 2003). About nine out of

ten parolees believe their parole officer

treats them with respect, is trustworthy

and acts professionally. The vast majority

express wanting help from their parole

officers finding jobs and expect their

parole officer to be helpful with their

transition. Unfortunately, they expect

more than is often delivered. Only about

half said their parole officer had actually

been helpful and thought supervision

would help them stay out of prison.

	           A m e r i c a n P r o b a t i o n a n d P a r o l e A s s o c i a t i o n 	

	
35

Parole Supervision: Pursuing the Balance

Policy Council 2005; Reinventing Probation Council 2000; Bureau
of Justice Assistance 2004; Petersilia 2003; Travis 2005; Burke 2004).
Collaborating with other agencies is a way to expand the capacity of
parole without necessarily having to develop and pay for it alone.

Supervision Strategies
In terms of supervision strategies, parole agencies should:
Align supervision resources with the risks, placing a premium
on the highest risk offenders, the highest risk places, and the highest
risk time for offending (Report of the Re-Entry Policy Council 2005;
Reinventing Probation Council 2000; Petersilia 2002; Petersilia 2003;
Travis 2005; Burke 2004). There is broad consensus – supported by
evidence-based principles – to focus resources on high risk populations.
The fact that the Urban Institute study indicated parole was least
effective with this population should only heighten concerns that the
highest risk parolees may not be receiving the right interventions in
the right dosage levels. High risk places are those neighborhoods with
the most returning offenders and/or the highest crime rates. And the
highest risk times are known to be the first days, weeks and months after
a prisoner is released (Travis 2005; Langan and Levin 2002). Focusing
both surveillance and treatment resources where the risks are highest
should ensure that the resources invested have the greatest impact.
Supervise parolees in their home neighborhoods (Report of the
Re-Entry Policy Council 2005; Reinventing Probation Council 2000;
Petersilia 2002; Petersilia 2003; Travis 2005). There is good reason to
end “fortress” parole that takes place in an office between 9 a.m. and
5 p.m. (Reinventing Probation Council 2000). Following the lead
of police, community-based parole officers would be responsible for
geographically-based caseloads, getting to know their neighborhood
resources and high-risk areas, and would thus be in a better position
to meaningfully assist and sanction parolees on their caseloads. By
supervising parolees where they live, fostering relationships with those
who know them best, parole officers could play an enhanced role in
making places safer.
Emphasize both surveillance and treatment (Report of the ReEntry Policy Council 2005; Petersilia 2002; Petersilia 2003; Taxman
2002). The research speaks clearly to the point that it takes a mix
of surveillance and treatment to reduce recidivism most effectively
(Sherman et al., 1997). Parolees should be assessed to identify risks and
needs, in accordance with evidence-based principles, and be provided
appropriate treatment, training and services. Even when parole can not
directly provide the services, they should access and connect parolees to
appropriate interventions and mandate their involvement.
Prioritize – and communicate – only rules and conditions
that can be realistically monitored and enforced (Kleiman 2005).
Conditions of release should be few, tied to positive expected outcomes
and tailored to individual risks and needs. Moreover, these rules and
the consequences for breaking them must be explicitly communicated
if they are to impact offender behavior. In other words, parolees need

36

Perspectives	

Spring 2006

to know the ground rules and expect them to be enforced if conditions
are to help deter reoffending (Kennedy 1998; Kleiman 1999; Harrell
et al., 1999; Taxman 2002).
Instill swift, certain, consistent, predictable responses to failures
(Burke 2004; Report of the Re-Entry Policy Council 2005; Reinventing
Probation Council 2000; Petersilia 2002; Kleiman 2005; Travis 2005).
The research literature suggests that to be effective, punishment should
be immediate and predictable, with clear, enforceable consequences
for violations (Burke 1997; Harrell et al., 2003; Taxman et al., 1999).
This ideal is far from actual practice in many states, where parolees
may violate conditions without being caught or may be caught several
times but receive nothing more than a warning, and then a seemingly
random violation results in their return to prison for the remainder
of their sentence. This recommendation is dependent both on parole
policy about responses to violations and, importantly, the availability
of intermediate sanctions in the community.
Introduce a range of incentives to induce and reward successes
(Travis 2005; Report of the Re-Entry Policy Council 2005; Burke 2004).
Research indicates that incentives and positive reinforcements may be
more effective than negative sanctions (Andrews et al., 1990). Concrete
incentives such as increasing curfew hours or reducing the number of
contacts could serve to motivate parolees to comply with conditions
and stay on the right track. Ultimately, parolees should be allowed to
earn their way off parole early by achieving certain milestones such as
keeping a job and staying sober (Travis 2005; Farabee 2005).

Looking Forward
In closing, there is a major opportunity to reform parole, or
“reinvent” it in the words of others (Corbett 1996; DiIulio 1997; Dickey
and Smith 1998; Rhine and Paparozzi 1999; Reinventing Probation
Council 2000; Lehman 2001; Petersilia 2002). While parole generally is
not producing large, visible reductions in crime among its caseload, it has
the potential to do so. In many ways, the situation is reminiscent of the
policing profession in the 1980s, when crime was high and confidence
in the police was low. The public expected police to catch and arrest
criminals – to react, but surely not prevent crime. Similarly, we count
on parole officers to catch parolees. Missed appointments, failed drug
tests and of course new crimes may all result in parole violations and
a return to prison. But few expect parole to actually deter and prevent
new crimes from occurring.
 Community policing has shown us what is possible: We now
expect police to help keep communities safe. In many ways parole has
advantages over their policing colleagues in the task at hand: Parole
officers know specifically who to watch – their caseloads – and they
have legal authority over them. Moreover they can set rules for these
individuals and implement a system of sanctions and incentives to help
coax good behavior. These are powerful tools that should be strategically
employed, not minimized.
At the same time, supervision should not be expected to singlehandedly reform former prisoners. More broadly, parole agencies must

Parole Supervision: Pursuing the Balance

work together with their prison and community-based colleagues to
prepare inmates for release, help parolees navigate those first critical
hours and days of freedom, and connect those motivated to jobs,
treatment, healthcare, housing and a supportive network of family and
friends. Supervision is only part of the reentry solution – but a very
important part.
Given the national momentum on the topic of prisoner reentry, there
is a real opportunity – if not obligation – to think big and expect more
from parole. There is no better time than now to improve supervision
and make it deliver on its potential to reduce crime, particularly among
the highest-risk individuals who warrant it the most.

Endnotes

1
Does Parole Work? Analyzing the Impact of Postprison Supervision on Rearrest Outcomes is
available in full at http://www.urban.org/url.cfm?ID=311156. For a copy of the technical report,
please contact the authors directly.
2
We also used data from the Census Bureau and National Corrections Reporting Program
3
The full statement from Joe Lehman, former Commissioner of the Washington State
Department of Corrections, was, “We have a broken parole system. Part of the problem is that parole
can’t do it alone, and we have misled the public in thinking that we can – hence the frustration, and
the cries to abolish parole. We don’t need to abolish parole, but a new model is sorely needed.”
4
See National Institute of Correction website for a series of papers that discuss evidencebased practices and principles in the community corrections setting (http://www.nicic.org/
Library/019342).

References

Andrews, D.A., Zinger, I., Hoge, R.D., Bonta, J., Gendreau, P., and Cullen, F.T. (1990). Does
Correctional Treatment Work? A Clinically Relevant and Psychologically Informed Meta-Analysis.
Criminology 28, no. 3: 375.
Beto, D.R., R.P. Corbett and J.J. Dilulio (2000). Getting Serious About Probation and Crime
Problem. Corrections Management Quarterly. (4) 2: 1-8.
Bogue, B., N. Campbell and E. Clawson (2004). Implementing Evidence-Based Principles in
Community Corrections: Collaboration for Systemic Change in the Criminal Justice System. Washington,
DC: National Institute of Corrections.
Bureau of Justice Statistics (2000). Conditional Release Violators Returned to State or Federal
Jurisdiction. Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, National
Prisoner Statistics Data Series (NPS 1) August 2000.
Burke, P. (1997). Policy-Driven Responses to Probation and Parole Violations. Silver Spring, MD:
Center for Effective Public Policy.
Burke, P. (2004). Parole Violations Revisited: A Handbook on Strengthening Parole Practices
for Public Safety and Successful Offender Transition. Washington, DC: National Institute of
Corrections.
Corbett, R.P. (1996). When Community Corrections Means Business: Introducing “Reinventing”
Themes to Probation and Parole, NCJ 168868. Federal Probation, 60, 1, (March 1996), 36-42.
1/1996.
Dickey, W. and M. Smith (1998). Dangerous Opportunity: Five Futures for Community
Corrections. Washington, DC: U.S. Department of Justice, Office of Justice Programs. http://www.
ojp.usdoj.gov/probation/.
DiIulio, J.J. Jr. (1997). Reinventing Parole and Probation. The Brookings Review. Vol.15 No.2,
pp 40-42.
Ellis, T., and P. Marshall (2000). Does Parole Work? A Post-Release Comparison of Reconviction
Rates for Paroled and Non-Paroled Prisoners. Australian and New Zealand Journal of Criminology.
Vol.33, n. 3, pp 300-317.
Farabee, D. (2005). Rethinking Rehabilitation. Washington, DC: The AEI Press.
Glaze, L.E., and S. Palla (2005). Probation and Parole in the United States, 2004. Washington,
DC: Bureau of Justice Statistics Bulletin.
Gottsfredson, M.R., and S.D. Mitchell-Herzfeld (1982). Another Look at the Effectiveness of
Parole Supervision. Journal of Research and Delinquency, Vol. 19 (2) 277-299.
Harrell, A. V., S. Cavanagh, and J. Roman (1999). Final Report: Findings from the Evaluation of
the DC Superior Court Drug Intervention Program. Washington DC: The Urban Institute, 1999.
Harrell, A.V., O. Mitchell, J. Merrill, and D. Marlowe (2003). Evaluation of Breaking the Cycle.
Washington DC: The Urban Institute, 2003.
Harrison, P. A. Beck (2005). Prison and Jail Inmates at Midyear 2004. Bureau of Justice Statistics
Bulletin. Washington, DC: U.S. Department of Justice
Hughes, T., D. Wilson, and A. Beck (2001). Trends in State Parole, 1990–2000. Bureau of Justice

Statistics Special Report. Washington, DC: U.S. Department of Justice, Office of Justice Programs.
Jackson, P. (1983). Some Effects of Parole Supervision on Recidivism. British Journal of
Criminology. Vol.23, n.1, pp 17-34.
Jacobson, M. (2005). Downsizing Prisons. New York and London: New York University
Press, 2005.
Kennedy, D. (1998). Pulling Levers: Getting Deterrence Right. National Institute of Justice
Journal, no. 236.
Kleiman, M. A. R. (1999). Controlling Drug Use and Crime Among Drug-Involved Offenders:
Testing, Sanctions, and Treatment. In Phillip H. Heymann and William N. Brownsberger, eds. Drug
Addiction and Drug Policy. Cambridge: Harvard University Press.
Kleiman, M.A.R. (2005). When Brute Force Fails: Strategic Thinking for Crime Control. Report
submitted to National Institute of Justice, March 31, 2005. mimeo.
Langan, P., and D. Levin (2002). Recidivism of Prisoners Released in 1994. Bureau of Justice
Statistics Special Report. Washington, DC: U.S. Department of Justice, Office of Justice Programs.
La Vigne, N.G., and V. Kachnowski (2005). Texas Prisoners’ Reflections on Returning Home.
Washington, DC: The Urban Institute.
Lehman, J.D. (2001). Reinventing Community Corrections in Washington State. Corrections
Management Quarterly, 5 (3), 41 - 45. 
MacKenzie, D.L. (1997). Evidence-Based Corrections: Identifying What Works. Crime and
Delinquency, 46: 457-471.
Nuttal, C.P., and Associates (1977). Parole in England and Whales. Home office Research
Studies No. 38. London, England: Her Majesty’s Stationary Office.
Piehl, A., and S. LoBuglio. 2005. Does Supervision Matter? In Prisoner Reentry and Crime in
America, edited by Jeremy Travis and Christy Visher. New York: Cambridge University Press.
Petersilia, J. (1998). A Decade of Experimenting with Intermediate Sanctions: What Have We
Learned? Federal Probation 62(2): 3–9.
Petersilia, J. (2002). Reforming Probation and Parole. Lanham, MD: American Correctional
Association, 2002.
Petersilia, J., and S. Turner. (1993). Intensive Probation and Parole. In Crime and Justice: A Review
of Research, vol.17, edited by Michael Tonry. Chicago: University of Chicago Press.
Petersilia, J. (2003). When Prisoners Come Home: Parole and Prisoner Reentry. New York:
Oxford University Press.
Reinventing Probation Council (2000). Transforming Probation Through Leadership: The
“Broken Windows” Model. New York: Manhattan Institute.
Council of State Governments (2005). Report of the Re-Entry Policy Council: Charting the Safe
and Successful Return of Prisoners to the Community. New York, NY. Council of State Governments.
www.reentrypolicy.org.
Rhine, E.E. and M. Paparozzi (1999). Reinventing Probation and Parole: A Matter of
Consequence. Corrections Management Quarterly. (3) 2: 47-52.
Rhine, E.E., W.R. Smith, and R.W. Jackson (1991). Paroling Authorities: Recent History and
Current Practice. Laurel, Maryland: American Correctional Association.
Rosenfeld, R., J. Wallman, and R. Fornango. (2005). The Contribution of Ex-Prisoners to Crime
Rates. In Prisoner Reentry and Crime in America, edited by Jeremy Travis and Christy Visher. New
York: Cambridge University Press.
Sacks, H., and C. Logan (1980). Parole: Crime Prevention or Crime Postponement? Storrs:
University of Connecticut School of Law Press.
Sacks, H., and C. Logan (1979). Does Parole Make a Difference? Storrs: University of
Connecticut School of Law Press.
Sherman, L.W., D. Gottfredson, D. MacKenzie, J. Eck, P. Reuter, and S. Bushway (1997).
Preventing Crime: What Works, What Doesn’t, What’s Promising. Research in Brief. Washington, DC:
National Institute of Justice.
Taxman, F., D. Soule, and A. Gelb (1999). Graduated Sanctions: Stepping Into Accountable
Systems and Offenders. Prison Journal 79 (2): 182-205.
Taxman, F. (2002). Supervision—Exploring the Dimensions of Effectiveness. Federal Probation
66 (2): 14-27.
Travis, J. (2005). But They All Come Back: Facing the Challenges of Prisoner Reentry. Washington,
DC: Urban Institute Press.
Travis, J. and C. Visher, eds. (2005). Prisoner Reentry and Crime in America. New York:
Cambridge University Press.
Travis, J., and S. Lawrence (2002). Beyond the Prison Gates: The State of Parole in America.
Washington, DC: The Urban Institute.
Visher, C., V. Kachnowski, N. La Vigne, and J. Travis (2004). Baltimore Prisoners’ Experiences
Returning Home. Washington, DC: The Urban Institute.
Visher, C., N. LaVigne, and J. Farrell (2003). Illinois Prisoners’ Reflections on Returning Home.
Washington, D.C.: The Urban Institute.
Waller, I. (1974). Men Released from Prison. Toronto, Ontario: University of Toronto Press.

Amy L. Solomon is a Senior Research Associate at the Urban Institute’s Justice Policy
Center. The author is grateful to Vera Kachnowski and Avinash Bhati, the study’s co-authors;
Christy Visher, who provided critical feedback on both the study and this article; and the
JEHT Foundation which funded the research. The views expressed in this article are those of
the author and should not be attributed to the Urban Institute, its trustees, or its funders.

	           A m e r i c a n P r o b a t i o n a n d P a r o l e A s s o c i a t i o n 	

	
37