Penny-Wise &
Pound-Foolish
Assaultive offender
programming and
Michigan’s prison costs

A rreport
eport b
y
by
Citiz
ens Alliance on Prisons and Public Spending
Citizens
and
American F
riends Service Committee
ogr
am
Friends
Committee,, Criminal Justice Pr
Progr
ogram
April 2005

Penny-Wise &
Pound-Foolish
Assaultive offender
programming and
Michigan’s prison costs

April 2005

A rreport
eport b
y
by
Citiz
ens Alliance on Prisons and Public Spending
Citizens
and
American F
riends Service Committee
ogr
am
Friends
Committee,, Criminal Justice Pr
Progr
ogram

Additional copies of this report may be purchased for $4, including shipping. The full report is also available on the CAPPS website: www.capps-mi.org

Citizens Alliance on Prisons and Public Spending
115 W. Allegan St., Suite 850
Lansing, MI 48933
Tel: 517-482-7753
Fax: 517-482-7754
www.capps-mi.org

American Friends Service Committee
Criminal Justice Program
1414 Hill St.
Ann Arbor, MI 48104
Tel: 734-761-8283
Fax: 734-761-6022
www.prisoneradvocacy.org

Contents
Acknowledgments
Executive Summary
Introduction ...................................................................................................................................................... 1
I. An overview of the Assaultive Offender Program .......................................................................................... 3
A. What the program is like ................................................................................................................................. 3
B. How prisoners are admitted............................................................................................................................. 3
II. The problems ............................................................................................................................................... 5
A. Prisoners do not have sufficient access to AOP to complete it before being considered for parole
or to have a beneficial impact on institutional behavior early in their incarceration. ........................................ 5
B. Parole board procedures may compound the waiting list delays. .................................................................... 14
C. AOP is not offered to prisoners housed at security levels III, IV and V ......................................................... 17
D. Parole has been delayed by the application of AOP eligibility criteria. ........................................................... 20
E. The parole board denies parole based on its disagreement with a positive assessment
by the AOP therapist. ................................................................................................................................... 22
III. Where are they today? An analysis of the 2004 waiting lists one year later ................................................ 24
A. Discharged .................................................................................................................................................... 24
B. Paroled .......................................................................................................................................................... 26
C. Still in prison, still waiting ............................................................................................................................ 26
D. Still in prison, no longer on the AOP waiting list ............................................................................ 29
IV. Recommendations ...................................................................................................................................... 30
V. Conclusion .................................................................................................................................................. 31
Appendix ......................................................................................................................................................... 33

Acknowledgments

For many years, the Criminal Justice Program of the American Friends Service Committee (AFSC) has
received letters from prisoners whose paroles were delayed by their inability to gain access to assaultive
offender therapy. As a result, AFSC has long advocated changes in the way treatment programs are provided. Since the Citizens Alliance on Prisons and Public Spending (CAPPS) began examining the impact
of parole denials on the size and cost of Michigan’s prison system, it has also received such letters. This
report grew out of that correspondence.
The principal researcher for the report was Charlene Lowrie of CAPPS. Ms. Lowrie and CAPPS Executive
Director Barbara Levine wrote the text. Penny Ryder and Natalie Holbrook of AFSC provided substantial
source materials and analytical insights as well as editorial support. Gail Light of CAPPS designed and
produced the report.
CAPPS and AFSC wish to thank the Michigan Department of Corrections for its cooperation in providing
the data upon which this report was built. Clayton (Tony) Straseske, Mental Health Services Manager, was
particularly generous with his time, providing background information and answering many questions.
Assistance provided by Richard Russell, John Rubitschun, Heidi Washington, Russ Marlan, Teri Cline and
Sergio Cacciani is also appreciated.
We would also like to thank the many prisoners who shared their circumstances with us. While only a
handful of these stories are actually told in the report, we hope these examples will help improve the
delivery of assaultive offender programming to thousands of Michigan prisoners in the future.

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Executive Summary
The Assaultive Offender Program (AOP) is a group therapy process for Michigan prisoners. It is designed
to change assaultive attitudes and behaviors and thus reduce the individual’s risk to the public upon parole.
AOP takes 44-52 weeks to complete. Because of long waiting lists, many prisoners cannot complete AOP
before they become eligible for parole. The parole board then often denies release.
•

There are 2,053 prisoners currently enrolled in AOP.

•

There are 2,256 prisoners on waiting lists who have already passed or are within one year of reaching
their first parole dates.

•

These waiting lists grew by 19 percent from February 2004-February 2005.

Each time the parole board denies parole for a year so a Level I or II prisoner can complete AOP, it costs the
taxpayers 14 times as much as it would have cost to have provided AOP in a timely manner.
•

The net cost of keeping a prisoner at a Level I or II facility for one year (after
subtracting the cost of parole supervision) is about $20,000. The cost at higher
security levels is substantially more.

•

The per prisoner cost of providing AOP is about $1,400.

•

Although the MDOC does not track this information, it appears that the parole board denies release to as many as 600
prisoners a year solely to permit completion of AOP, at a cost
for further incarceration of about $12 million.

Some prisoners serve their maximum sentences without having the
opportunity to take AOP. These prisoners are discharged to the
community without parole supervision or supportive services.

$20,000
Incarceration
(net) at Level
I or II

$1,400
AOP

•

Many of the prisoners who “max out” had behavioral problems
while incarcerated. Of those on the 2004 waiting lists, it appears that 67 were at security Level IV and 49 were at Level V
when they were discharged.

•

Prisoners at higher security levels are housed under highly restrictive conditions. They are the least prepared to live on their own and pose a greater risk upon
release.

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

The frequent failure to provide AOP before parole consideration and even discharge has several causes:
•

There are not enough therapists available to conduct AOP groups.

•

The MDOC does not maintain one statewide AOP waiting list. Instead, facilities have their own lists
which change constantly as prisoners are transferred.

•

The availability of AOP at Level I and II prisons and camps varies widely.

•

AOP groups are not conducted at Level III, IV and V prisons or at two camps.

•

Prisoners nearing their release dates are commonly housed at facilities where the treatment is not even
provided.

•

Program eligibility criteria have been changed over time so that prisoners who have served many years
sometimes reach their first parole interview without having been assessed for AOP.

Parole board practices may compound the problem.
•

Prisoners are frequently denied parole for “failing” to complete a treatment program they could not
access.

•

Prisoners enrolled in AOP were often continued in prison for a full year when they were only a few
months from completing treatment, a practice the parole board is working to correct.

•

There are no criteria for deciding when AOP need not be required or when it would be appropriate to
make completion of a comparable program a condition of parole.

Even successful completion of AOP does not guarantee parole since the board routinely disagrees with the
assessments of therapists.
•

In 2004, of the prisoners considered for parole who had completed AOP, more than half were denied
release.

•

When the board concludes that a prisoner who has completed AOP with a positive assessment remains
a risk to release, it is not required to explain its difference of opinion with the treating therapist.

•

No process exists for reviewing the board’s findings or appealing its decisions.

•

Currently, nearly 1,500 prisoners have served their minimum sentences and completed AOP
but have been denied parole.

Numerous steps could be taken to increase the availability of assaultive offender treatment to prisoners and
the movement of prisoners to parole. Such steps would include:
•

hiring more therapists to conduct AOP groups

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

•

establishing a statewide waiting list and transferring prisoners as needed so they can enter groups in
time to finish before they become parole eligible

•

adopting practical AOP admissions criteria and applying those criteria consistently

•

providing AOP at all custody levels

•

allowing the parole board to defer decisions for up to three months to allow a prisoner to complete
AOP

•

prohibiting the parole board from denying release solely because a prisoner has not completed a
treatment program the prisoner could not access

•

creating a review process for cases in which the board disagrees with the conclusions of
MDOC therapists

•

identifying criteria for when AOP participation can reasonably be waived altogether and when it can
safely be completed in the community as a condition of parole.

Taking such steps would require a short-term investment of resources and a commitment by the MDOC to
give the same priority to treatment as it does to institutional management. It would also require a thorough
analysis of the relationship between program completion and parole decision-making. The fact that the
Department is currently exploring some of these options is a hopeful sign. In the long-term, increased
public safety and decreased incarceration costs would be well worth the effort.

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Introduction
Nearly 23,000 Michigan prisoners are serving sentences for assaultive offenses, accounting for 46 percent
of the entire prison population. 1 Their crimes range from resisting arrest, domestic violence and felonious assault to robbery and murder. Since these offenders may pose a particular risk to public safety upon
release, programming to change the attitudes and behaviors that lead to assaultive crimes is essential.
Although the Michigan Department of Corrections (MDOC) has provided treatment programming for
assaultive offenders for over 30 years, many prisoners are unable to complete treatment before they become eligible for release. This creates a variety of serious and costly problems.
The Assaultive Offender Program (AOP) is intensive group psychotherapy that takes about one year to
complete. As of January 1, 2005, there were 2,053 2 assaultive offenders participating in AOP at 46 of the
52 prisons and camps throughout the state.
In addition to those currently in AOP, more than 11,000 prisoners are on waiting lists. While many of
these prisoners are years away from finishing their minimum sentences, those waiting include 2,256 prisoners who have already passed or are within one year of their earliest release dates.3 Most are unlikely to
complete AOP before they are considered for parole.
The Assaultive Offender program is not required by law. The parole board retains absolute discretion to
decide whether a prisoner who has served his or her minimum sentence should be released. The board
paroles prisoners who have not completed AOP and denies parole to prisoners who have completed the
program successfully.
The overall parole grant rate for assaultive offenders is low. In 2004, of 6,796 prisoners referred for AOP,
the board granted parole to only 39 percent. Nonetheless, AOP participation improves a prisoner’s chances.
While MDOC data is incomplete, at least 2,960 prisoners in the 2004 group had finished AOP. 4 Their
parole grant rate was 47 percent. Conversely, the board commonly denies parole to prisoners who have
been unable to gain entrance to treatment groups.5

1. CAPPS, The high cost of denying parole: an analysis of prisoners eligible for release, Lansing: November 2003, pg. 39. This
number does not include sex offenders.
2. MDOC Bureau of Health Care, Psychotherapy Program Summary, Report No. HC-341, January 1, 2005.
3. MDOC Bureau of Health Care, Psychotherapy Program Waiting List, Report No. HC-331, February 1, 2005. Earliest
release date is the point at which the judicially imposed minimum term has been served and the parole board obtains the
authority to grant release.
4. Data provided by MDOC Public Information Office, March 25, 2005.
5. The Department of Corrections has no statistics showing the number of times the parole board has denied parole
expressly because the prisoner had not completed AOP or the number of referrals the board has made for AOP assessment.

CAPPS/AFSC

1

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

The average cost for each prisoner who is not paroled because assaultive offender programming was not
delivered in a timely fashion is roughly $22,000.6 Even though the connection between AOP completion
and parole is far from automatic, the dollar amount wasted by AOP waiting lists is huge. If the board
denies parole to just 600 prisoners a year (35 percent of those on waiting lists at security Levels I-IV) solely
to permit completion of AOP, the cost is more than $12 million. The $1,400 per prisoner cost to provide
AOP, which must be paid in any
event, has merely been delayed
If the board denies parole to just 600 prisoners a
while the cost to further incarcerate these individuals has
year (35 percent of those on waiting lists at
been added.

security Levels I-IV) solely to permit completion of

Expense is not the only conseAOP, the cost is more than $12 million.
quence of delaying or denying
access to assaultive offender programming. The program is not
even offered at higher security facilities. The theory is that prisoners at these facilities are less likely to be
granted parole and that scarce treatment resources should be directed to those who will be returning to the
community soonest. However, many prisoners are classified to higher custody precisely because they have
engaged in assaultive behavior in prison. Those who have not only committed assaultive crimes but
continue to behave in an assaultive fashion are presumably most in need of therapeutic programming. If
successful, treatment would even allow them to reduce their custody levels, thereby reducing the cost of
their incarceration and increasing their chances of obtaining release. When these prisoners are continually
kept at higher custody, they eventually serve their maximum sentences and must be released into the
community without parole supervision and without the benefit of the assaultive offender program. By not
giving the same priority to treatment as it does to institutional management, the MDOC loses an opportunity to reduce the risk to public safety.
Other prisoners are classified to higher custody for reasons unrelated to assaultive conduct in prison.
These people are caught in a classic catch-22. They cannot enter AOP because it is not offered at their
security level and they cannot get paroled because they have not completed AOP.
Problems with the delivery of assaultive offender programming have a significant impact on prisoners, the
public and the state budget. This report analyzes the causes of these problems, illustrates their impact and
proposes solutions. To place the problems in context, it is necessary to briefly consider the nature of AOP
and how eligibility for the program is determined.

6. Figures provided by the MDOC indicate that the current average cost of keeping a prisoner in a Level I or II prison for
one year is $21,794. (This amount includes medical but not mental health care.) The cost rises substantially as security
levels increase. The cost of keeping a prisoner on parole for one year is $1,977. When the parole cost is offset against the
cost of incarceration it yields a net average cost of $19,817 for keeping a Level I or II prisoner for one
additional year.

2

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

I. An overview of the Assaultive Offender Program
A. What the program is like 7
The Assaultive Offender Program is designed to address the factors that underlie assaultive behavior. The
program enables prisoners to recognize and understand their own behavior patterns and develop new skills
to manage their conduct.
Treatment is provided through four major components. One therapist leads each group, averaging 9-10
prisoners. Groups meet for 1½ to 2 hours a week over the course of 44-52 sessions.
During Orientation, group members are expected to develop an understanding of “program content, goals,
philosophy and standards for completion.” They should also acquire a general understanding of the thought
patterns and “target behaviors” that need changing.
During Case Disclosure, members investigate their personal history and family background. Material for
both “psychotherapeutic and relapse prevention interventions” comes from information about the individual’s
“personality, psychological characteristics, social functioning, beliefs and life experiences.” “Understanding and identifying abuse and violence” is emphasized. Members are “challenged to make a searching
evaluation of their background to identify factors contributing to their distorted thinking and mistaken
beliefs and to relate these issues to their current behaviors.”
During Offense Precursors, group members are assisted in understanding the elements that “. . . make up
the pre-assault pattern for offending. Behaviors, thinking errors and emotional functioning are discussed
and related to their own past circumstances and assault patterns. The all-important issue of intimacy is
addressed and related to their interpersonal difficulties.”
Self -Maintenance is the last segment. Development of a written relapse prevention plan is the final step.
Emphasis is placed on identifying and understanding individual risk factors that have “surfaced during
earlier group work, understanding and visualizing the lapse-relapse process, developing coping strategies to
manage internal and external risks and the incorporation of a plan for lifestyle changes.” Group members
are expected to incorporate into their relapse prevention plans “an outline of their community support
network and life goals.”

B. How prisoners are admitted
People who may need AOP are identified at the Reception and Guidance Center when they first enter
prison. AOP is automatically recommended for everyone whose offense appears on a list of crimes desig-

7. This summary is drawn from the MDOC Assaultive Offender Program materials provided to participating prisoners.

CAPPS/AFSC

3

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

nated as assaultive.8 In addition, assaultive behavior exhibited in connection with an unlisted offense or in
the prisoner’s history may also generate a recommendation. Subsequent to Reception Center screening,
prison staff and members of the parole board may recommend AOP participation. Prisoners can also seek
admission on their own initiative.
A recommendation for AOP does not automatically result in placement in a group. The recommendation
simply secures the prisoner a place on a waiting list. Determining whether someone will be admitted to
AOP occurs at the prison facility through an assessment interview conducted by the therapist. Reasons for
denying admission have historically included the prisoner’s failure to sufficiently express: 1) interest in
AOP, 2) responsibility for his/her assaultive behavior, 3) the need to change behavior, gain insight and/or
acquire new skills because of the past assaultive behavior or potential for future assaultive behavior, and 4)
appropriate goals and objectives relating to assaultiveness.9 In 2004, of 2,864 assessments, more than 31
percent resulted in admission to AOP being denied.10
Each facility maintains its own waiting list, even if AOP is not offered there. Multi-level facilities maintain
a list for each custody level. The waiting list consists of every prisoner at the facility who has been recommended for AOP. The list is ordered according to the prisoners’ earliest release dates. The order changes
constantly as prisoners are transferred in and out of the facility. Thus the person who is first on the list
could suddenly become number five because four prisoners with shorter minimum sentences have just
arrived.
The assessment takes place when a new group begins and the prisoner is next in line on the facility’s AOP
waiting list. If admission is approved, the person’s name is removed from the waiting list, an AOP admission
report is issued and he or she begins treatment. If not admitted, the person’s name is removed from the
waiting list and an AOP non-admission report is issued explaining the basis for the decision to deny entry.
A prisoner may seek reconsideration, in writing, but must explain to the therapist’s satisfaction what has
changed since the prior assessment to warrant admission.
Prisoner custody levels run from I (minimum) through V (maximum). Although the Psychological Services Unit has stated the intent to give the highest priority in delivery of AOP to custody Levels I through
IV, 11 in reality, the Department has focused AOP only on prisoners housed in Levels I and II. Moreover, of
the 10 camps, all classified Level I, two do not have AOP available.12 Prisoners who are on a waiting list at
8. See Appendix: MDOC, Operating Procedure 04.06.180-D, 1/13/03, Attachment A.
9. MDOC Assaultive Offender Program, Program Statement, June 3, 2002, pg. 5, “Procedural Guidelines for Assessment
for Assaultive Offender Program”, items 5(a)-(e).
11. Program statement, note 9 supra, pg. 3, “Highest priority in terms of AOP participation will be offered to waiting list
prisoners classified to the following security level designations: Levels I through IV. A regional psychological services director
may authorize exceptions to this and extend AOP participation to other security levels (V and VI) when deemed
appropriate.” [Italics in the original.]
12. The Michigan Youth Correctional Facility, a privately-operated Level V prison, also provides AOP. According to the
January 1, 2005 Bureau of Health Care Program Summary Report No. HC 341, 54 prisoners there were
enrolled in AOP, 5 had been terminated and 18 were on the waiting list, 4 of whom were within 1 year of
their earliest release date.

4

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

an institution where AOP is not offered must await a transfer to another institution where he or she will be
placed on that facility’s waiting list. Prisoners housed at Level III and above will appear on a waiting list but
will only be considered for admission to AOP when their security level has been reduced to I or II.

II. The problems
The insufficient number of groups at Level I and II facilities, the lack of an integrated system for managing
waiting lists and the lack of programming at Level III, IV and V facilities are among the primary problems
with the MDOC’s current delivery of assaultive offender programming. The relationship between AOP
participation and parole decision-making creates further difficulties.

A. Prisoners do not have sufficient access to AOP to complete it before
being considered for parole or to have a beneficial impact on institutional behavior early in their incarceration.
1. AOP waiting lists at many facilities are too long, with too many
individuals past or within a year of their earliest release dates.
As of February 1, 2005, there were 1,190 prisoners on facility waiting lists for AOP who had already served
past their earliest release dates. By definition, these prisoners were seen by the parole board before having
had the opportunity to complete AOP. The problem is not only bad, it is getting worse.
As Figure 1. shows, the total number of
these wait-listed prisoners increased by ten
percent from February 4, 2004. The lists
grew longer at every security classification
except Level III.13 Particularly disturbing
is that the number at Level I facilities grew
by 45 percent. Since these are minimum
security prisoners who have served their
minimum sentences, they should have the
best chance of being paroled.
As of February 1, 2005, there were 1,066
prisoners on facility waiting lists for AOP
who were within a year of their earliest release date. Since AOP takes nearly a year
to complete, very few of these prisoners

Figure 1. Prisoners Past Their Earliest Release Date on
2004 and 2005 AOP Waiting Lists
391

400

371
321

350
300

254
225

250
200

175

161

172

150
70

68

100

50

50
0

1

I

II

III

IV

V

VI

12

1

Unspecified

Security Level
2004

2005

13. The drop in the Level VI waiting list occurred because Level VI was abolished in 2004.

CAPPS/AFSC

5

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

have any hope of finishing treatment before they finish their minimum sentences, even if they are admitted
to AOP. As Figure 2. shows, the total
number of these wait-listed prisoners inFigure 2. Prisoners Within 1 Year of Earliest Release Date
creased by 31 percent from February 4,
on 2004 and 2005 AOP Waiting Lists
2004. Again, the greatest increase was
583
at Level I, where the list of prisoners 600
within a year of their earliest release dates
500
rose by 41 percent.
413
Taken together, the number of prisoners who are within a year of or are past
their earliest release dates and are still
awaiting entry to AOP grew by 19 per
cent in just the year from February 4,
2004 to February 1, 2005.

400

300

231
200

200

101
70

100

36

78

79
57

15

12

0

6

0
The access problem is compounded by
I
II
III
IV
V
VI
Unspecified
the fact that AOP is not even offered at
Security Level
Levels III-V. Therefore, prisoners at
2004 2005
those levels have no prospect of completing it unless they are transferred to
lower security facilities. (See discussion below, Section C.) For just those prisoners at Level I and II facilities as of January 2005, the length of the waiting lists and the number of people participating in AOP
groups appear in Table 1 (See page 7):14

Two problems are apparent from this table. First, the current level of program delivery is inadequate to
meet the overall need. This is a resource problem, pure and simple. Every facility is in need of additional
AOP groups. If more therapists were available, more groups could be run simultaneously. If the backlog
were eliminated, so too would be the problem of people being denied parole because they had not completed AOP.
The total of 7,371 Level I and II prisoners on AOP waiting lists includes people who are many years from
completing their minimum sentences. To understand the immediate problem in delivering AOP to paroleeligible prisoners, it is most useful to compare the individual facility lists of those who are within a year of
their earliest release dates or have already passed those dates with the number currently enrolled in AOP
groups.
Sixteen facilities have too few prisoners on their waiting lists even to establish a therapy group. Twenty-two
have waiting lists of 20 prisoners or more. Conversely, twenty facilities have fewer than 20 prisoners

14. MDOC Psychological Services Unit, Psychotherapy Program Summary, Report No. HC-341, January 1, 2005 and
MDOC Psychotherapy Program Waiting List, Report No. HC-331, February 1, 2005.

6

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Table 1. Level I and II Prisoners on AOP Waiting Lists
as of Feb. 1, 2005 and Enrolled as of Jan. 1, 2005

CAPPS/AFSC

Facility

Total

Baraga Maximum Corr Facility - Baraga
Bellamy Creek Corr Facility - Ionia
Boyer Road Corr Facility - Carson City
Brooks Corr Facility - Muskegon
Camp Branch - Coldwater
Camp Brighton - Pinckney
Camp Cusino - Shingleton
Camp Kitwin - Painesdale
Camp Koehler - Kincheloe
Camp Lehman - Grayling
Camp Manistique - Manistique
Camp Ottawa -Iron River
Camp Sauble - Freesoil
Camp Tuscola - Caro
Carson City Corr Facility - Carson City
Chippewa Corr Facility - Kincheloe
Cooper Street Corr Facility - Jackson
Deerfield Corr Facility - Ionia
Egeler Reception & Guidance Center - Jackson
Florence Crane Corr Facility - Coldwater
G. Robert Cotton Corr Facility - Jackson
Gus Harrison Corr Facility - Adrian
Hiawatha Corr Facility - Kincheloe
Huron Valley Complex – Women’s - Ypsilanti
Ionia Maximum Corr Facilty Ionia
Kinross Corr Facility - Kincheloe
Lakeland Corr Facility - Coldwater
Macomb Corr Facility - New Haven
Marquette Branch Prison - Marquette
Mid-Michigan Corr Facility - St. Louis
Mid-Michigan Youth Corr Facility - Baldwin
Mound Corr Facility - Detroit
Muskegon Corr Facility - Muskegon
Newberry Corr Facility - Newberry
Ojibway Corr Facility - Marenisco
Parnall Corr Facility - Jackson
Parr Highway Corr Facility - Adrian
Pine River Corr Faciltiy - St. Louis
Pugsley Corr Facility - Kingsley
Richard A. Hanlon Corr Facility - Ionia
Riverside Corr Facility - Ionia
Ryan Corr Facility - Detroit
SAI - Boot Camp - Men - Chelsea
SAI - Boot Camp - Women - Chelsea
Saginaw Corr Facility - Freeland
Scott Corr Facility - Plymouth
Southern Michigan Corr Facility - Jackson
St. Louis Corr Facility - St.Louis
Straits Corr Facility - Kincheloe
Thumb Corr Facility - Lapeer
West Shoreline Corr Facility - Muskegon
Totals

64
275
199
170
29
10
79
74
73
158
48
72
49
102
183
32
103
63
239
57
273
151
124
138
44
324
215
273
146
24
14
280
310
258
287
152
91
56
127
342
196
373
10
1
234
151
237
3
173
246
39
7,371

Within 1 Year
or Past ERD
16
34
24
14
24
4
19
5
28
51
10
10
22
83
3
13
18
2
48
16
22
1
56
51
0
52
13
57
72
5
3
17
16
43
65
29
20
8
63
14
4
34
10
1
14
8
4
0
73
25
15
1,239

Currently
Enrolled
32
72
11
68
26
57
27
0
34
12
0
19
16
15
58
26
82
63
0
32
140
44
37
24
13
21
39
15
18
56
54
24
60
40
19
110
15
50
33
125
64
19
0
0
18
149
72
15
50
42
37
2,053

7

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

currently enrolled in AOP while 17 facilities have 50 or more AOP participants. Those with smaller
current enrollments tend to have longer waiting lists; those with larger enrollments tend to have shorter
waiting lists.
The disparities have a strong geographic component. The facilities with the longest waiting lists and fewest
AOP groups tend to be in the Upper Peninsula and Northern Lower Peninsula. Prisons in the Jackson and
Ionia areas have the highest enrollments and smallest backlogs.
The obvious question is why this maldistribution occurs. Is it a result of the number of therapists available
or the priorities set by administrators at particular facilities? In any event, the immediate answer seems
clear. Funnel more resources to facilities with long waiting lists and place more prisoners needing AOP at
the facilities with the shortest lists, so their chances of gaining entry will at least be improved.
Currently, the MDOC employs about 100 therapists who have multiple duties. Those working at the 46
facilities that actually offer AOP are expected to conduct five psychotherapy groups each year, which may
be any combination of assaultive offender and sex offender groups. If each therapist conducts three AOP
groups per year, each with 10 participants, the estimated cost per prisoner of assaultive offender therapy is
$1,404.15
One way to cure the backlog is to employ more full-time therapists to do nothing but conduct AOP
groups. They could be permanent MDOC employees, contractual employees hired only for as long as
necessary to clear the backlog or a combination of both. Whatever their status, they could each be expected to conduct seven or eight groups a year, resulting in AOP completion for 70-80 prisoners.16 Another alternative is to contract with community-based service providers who would enter the prisons to
provide AOP. In geographic areas where this option is available, presumably it would be less expensive. In
some areas, graduate students working under supervision might also be used to increase AOP availability.
Cheaper still, in appropriate cases, would be permitting prisoners to complete AOP in the community as
a condition of parole. At roughly $1,400 per prisoner, even the most expensive option is far more costeffective than housing people who have been denied parole.
The MDOC has said that physical space limitations at overcrowded prisons also prevent conducting more
treatment groups. The pressure that double-bunking places on common areas, such as classrooms, gyms,
visiting rooms and other potential meeting places leaves nowhere to hold more AOP groups. Presumably,
this problem could be remedied by using portable classrooms or similar modular units designed to house
treatment programs in a setting that provides adequate privacy. Again, the purchase price would be far
out-weighed by the cost of keeping hundreds of prisoners for an additional year, especially if such units
15. Based on pay range information provided by the MDOC, this assumes a pay rate of $27/hr and fringe benefits equaling
50 percent for a total cost of $84,240 per year. It further assumes each therapist spends half of his or her time on providing
AOP to 30 prisoners and AOP related activities, including admission assessments, monthly progress notes and the
preparation of various reports.
16. Adding one therapist solely to deliver AOP at 51 facilities and two therapists at each of the remaining two facilities
would have allowed the prisoners who were within a year of their release dates to have completed the program before the
parole board gained jurisdiction. At an estimated cost of $4.6 million for 55 additional therapists, the savings in housing
from the improved likelihood of parole would be considerable. If groups at Levels IV and V were kept to a
maximum of seven prisoners for institutional management purposes, five more therapists would be needed,
raising the total cost to about $5 million.

8

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

could be built by prisoners themselves, perhaps through the Prison Build Project.
If the waiting lists for people within a year of their earliest release dates were eliminated, the MDOC could,
at a minimum, provide AOP before people reached the one-year mark from then on. Ideally, it could
begin providing AOP to prisoners much earlier in their incarceration, closer to the time of the offense,
when the prisoner might benefit most
and when assaultive behavior in
prison could also be affected.
The second problem is the inefficient
distribution of the resources that do
exist. Facilities at the same security
level have very different waiting lists.
Someone whose earliest release date
might put him at the top of the list at
Camp Manistique could be 30th at the
Straits Correctional Facility.

Because there is not a single, system-wide
waiting list, people are not really being
admitted to AOP according to their earliest
release dates.

Because there is not a single, system-wide waiting list, people are not really being admitted to AOP according to their earliest release dates. Rather, their admission depends on the length of the list at the facility
where they are housed. Except for transfers from camps that do not provide AOP to camps that do, there
is little effort to place prisoners at facilities based on AOP availability. Prisoner placements are dictated by
many concerns, including sheer bed availability. This further complicates the AOP admissions process,
since a prisoner’s name may keep moving up or down on the waiting list as other prisoners recommended
for AOP come and go. Thus AOP access, and ultimately parole, is indiscriminately delayed. For example:
Dan O’Bryan, #429329, a bricklayer with a family and no prior record, was sentenced to concurrent terms of 1-5 year years for attempted home invasion and 1-15 years for manslaughter. O’Bryan
had confronted his sister’s ex-boyfriend, who had been harassing and stalking her. He pushed his
way into the other man’s apartment and the two fought briefly, throwing punches and wrestling
each other to the floor. O’Bryan then left the ex-boyfriend lying on the floor, unaware that he was
asthmatic. The victim subsequently died of asphyxiation.
The Reception Center automatically recommended AOP based on O’Bryan’s offense. Between December 8, 2002 and April 21, 2003, O’Bryan wrote eight memos to prison staff seeking to enter
AOP. In them, he emphasized his first release date was October 9, 2003. He was told repeatedly that
he was on the waiting list. Finally, five months before his earliest release date, on April 28, 2003,
O’Bryan began the program. Completion was scheduled for April 2004.

CAPPS/AFSC

9

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

O’Bryan had a high parole guidelines score, indicating a low risk for re-offending, He received two
positive points for “situational crime, unlikely to recur.” The detective who investigated the case
wrote to the board in support of parole, saying O’Bryan presented no risk to the public. Nonetheless,
on September 10, 2003, the parole board continued O’Bryan’s imprisonment for an additional 12
months. Its reasons were: “Offender’s involvement in AOP designed to address his assaultiveness
may provide him with the requisite insight to demonstrate that he better understands his aggressive
tendencies; provide tools to control his temper and reduce his risk.” 17
Since the research for this report was begun in November 2004, the MDOC has focused substantial
attention on the backlog of Level I and II prisoners awaiting entry to AOP. It is beginning a comprehensive screening of the Level I and II waiting lists to better assess needs and match resources appropriately for
prisoners within 24 months of their earliest release dates. It is exploring various ways to reduce the backlog, including using more clinical social workers to conduct groups, paying current therapists overtime to
conduct additional groups, moving more prisoners to facilities where AOP is available and developing
alternative shorter programs for selected offenders.18

2. Admission to AOP does not guarantee that programming will not be
interrupted due to transfer for administrative reasons.
Under current Department of Corrections policy, security classification takes precedence over program
classification. Many kinds of decisions can fall under the broad umbrella of “security” due to the nature of
a prison setting. As a result, prisoners who have not engaged in inappropriate behavior may be transferred
even though they have started AOP.19
The MDOC’s security classification policy does not offer sufficient guidance to prevent AOP program
interruptions.20 On the one hand, custody reductions will not be delayed if the same or similar program is
offered at the less secure facility. This assumes the transferred prisoner will be readmitted to the interrupted
program upon arrival at the new facility. While this is true for most programming, it is not true for AOP.
On the other hand, lateral transfers and transfers to reduced custody will be delayed if the offender is
involved in AOP as a result of a recommendation by the parole board. In those cases, of course, parole has
already been denied.

17. A full profile is available at www.capps-mi.org/pdfdocs/profiles/obryan.pdf
18. Information provided by MDOC at a meeting on with CAPPS on March 15, 2005.
19. MDOC Policy Directive 05.01.130, “Prisoner Security Classification”, Attachment B, effective March 1, 2004,
governs prisoner security classification and permits a custody “waiver” in order to enable the prisoner’s access to medical
and program needs or to facilitate a writ. It also allows a waiver for administrative needs based on a lack of bedspace.
20. Id., paragraphs W and X.

10

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

MDOC policy does provide discretion to wardens to allow prisoners to avoid transfer if they are in a
program, but the language gives such leeway that the transfer seldom is delayed.21 However, the MDOC’s
routine use of transfers to accommodate bedspace considerations alone can preclude keeping the prisoner
in AOP at that location, regardless of the time and resources already invested.
Given the breadth of discretion these directives confer and the availability of language to support almost
any decision, it is not surprising that transfer decisions made at the facility level affect AOP participation
in inconsistent ways.
Moreover, prisoner records are not always properly flagged to reflect involvement in AOP, resulting in
transfers that could have been prevented.
Transfer decisions are further complicated by the structure of the group therapy process. A prisoner who is
transferred during the course of AOP involvement is not placed in an ongoing group at the receiving
facility. Instead, the transferred prisoner’s name goes on the new facility’s waiting list. Programming that
is interrupted in this manner can lead to undue delay in AOP completion and, ultimately, parole denial.
For example:
Kenneth Swanchara, #375263, pled guilty to aggravated stalking and possession of a short-barreled
shotgun for harassing a former girlfriend and threatening to harm her, her new boyfriend and
himself. He was sentenced to serve 1 ½ - 5 years in prison. The Reception Center recommended
that Swanchara be assessed for AOP. Although he was admitted to the program within two months
of his arrival in prison, three weeks later he was transferred to an Upper Peninsula camp where
AOP was not offered. He remained there until his first parole interview. The parole board then
denied release because of Swanchara’s “failure” to complete the program. 22
In another example,
Charles McDonald, #420696, pled guilty to armed robbery and was sentenced to a term of 1 year,
9 months to 15 years on July 24, 2002. His earliest release date was January 25, 2004.
The Reception Center recommended that McDonald complete GED and AOP programming, even
though GED was not statutorily required in McDonald’s case because his minimum term was less
than two years. McDonald was admitted to AOP after he was screened on August 27, 2003. On
November 4, 2003, he was notified that the therapist was on extended leave and not expected to
return until the middle of December. On January 20, 2004, McDonald’s AOP group was officially
cancelled when the therapist transferred to another facility.
McDonald had attended four sessions and had two excused absences when his group was cancelled.
The therapist concluded that Mr. McDonald was a “good candidate for continued participation in
21. Id., paragraph “X”, states that reductions in custody and lateral transfers of offenders involved in therapy programs
will be made only when necessary and only with the approval of the warden or designee if there has been a considerable
investment of time and resources, the offender’s completion of the program is imminent and space at the current
placement allows such delay. There are no criteria to guide each facility when assessing each of these factors.
22. A full profile is available at www.capps-mi.org/pdfdocs/profiles/swanchara.pdf

CAPPS/AFSC

11

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

AOP.” McDonald was informed that no existing AOP groups at his facility could absorb him at that
time. The notice also said that he would be placed in a new group in May 2004.
However, McDonald’s custody level was reduced and he was moved to several other locations, including an Upper Peninsula camp that provided GED but not AOP programming. Although he
has since been moved from that location to one that does have AOP, he remains on an AOP waiting
list.
These problems, created by current MDOC policies, illustrate an unnecessary waste of resources. The
prisoner that is transferred will be required to repeat segments of AOP that he has already completed; a
new therapist will have to repeat the work already performed by another therapist; and the prisoner that
was next in line on the waiting list will still have to wait until the next group begins because he cannot take
the place of the transferred prisoner in an on-going group.
The MDOC advises that it has greatly reduced the disruption of AOP participation by administrative
transfers and that there is a strong commitment to coordinating placements, within security levels, with
programming needs. Nonetheless, policy revisions designed to place and keep the prisoner at a facility
where required programming is provided in time to permit completion before the prisoner’s earliest release
date would be desirable to ensure that problems are avoided.

3. Because two of the 10 camps do not offer assaultive offender programming, minimum security prisoners are often denied timely access
to AOP while they wait for transfer to a camp that does offer it.
There is no central oversight within the Department of Corrections to make sure that someone waiting for
AOP is placed in a location where it is offered. Ironically, the problem is particularly acute for low security
prisoners close to their earliest release dates. The same characteristics that make these prisoners prime
candidates for parole also make them prime candidates for placement at a minimum security camp. There
are five camps for male prisoners in the Upper Peninsula, three of which provide AOP. Since two Upper
Peninsula camps housing male prisoners do not provide AOP, prisoners who rise to the top of the AOP
waiting list at these facilities must be transferred to a camp with an opening. This problem is compounded
by an MDOC policy that requires prisoners placed in a Region I facility (Upper Peninsula/Northern
Lower Peninsula) to stay in Region I for a year before being allowed to transfer downstate. Thus, a prisoner
who has spent less than a year housed in Region I can only enter AOP if an opening occurs within the
Region. 23
Some prisoners find themselves still waiting to be transferred in order to receive AOP when they are
interviewed by the parole board. For example:
23. The availability of AOP to camp residents may be further diminished by the imminent closing of Camps Sauble and
Tuscola, both of which provided the program.

12

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Robert Smith, #342868, was convicted of assault with intent to commit great bodily harm less
than murder for stabbing a man with whom he had argued. Smith was sentenced in November
2003 to serve 1½ - 10 years in prison. Although the Reception Center had recommended AOP,
Smith was housed in a Level I camp in the Upper Peninsula that did not provide the program.
After repeated requests, he was finally screened at the camp in June 2004 and determined to be
eligible for AOP. The therapist advised: “When his name is at the top of the waiting list he will
be considered for transfer to a camp where the program is offered.” In the meantime, Smith has
been denied parole despite his favorable parole guidelines score. Smith was number four on the
facility waiting list as of February 1, 2005. He became eligible for parole on February 22nd.
In all of these cases, the MDOC failed to place prisoners with minimum sentences of 18 months or less in
facilities where they could complete AOP programming before their earliest parole dates. O’Bryan was at
a facility where he could not get to the top of the waiting list in time. McDonald’s AOP involvement was
terminated when the therapist transferred to another facility. He was then moved to a location that did not
offer AOP. Smith was at a camp where the program did not exist and, even when finally moved to a
location that offered AOP, has not been able to gain admission. Swanchara was actually taken out of AOP
after he had started and moved to a camp that did not provide it. Where these prisoners stood relative to
other prisoners with later release dates is unknown. The lack of any centralized planning that makes
treatment completion a priority from the time they leave Reception is evident.
An integrated statewide waiting list would resolve many of these problems. Central office psychological
services staff could work in cooperation with the central office classification staff to place prisoners at
locations where AOP is offered when their names reach the top of the centralized list. Although it will not
change the fact that newly incarcerated individuals will be added as they enter, affecting placement on the
list to some degree, a statewide list would also remove the fluctuation in waiting list placement currently
caused by the high transfer rates within the system. In addition, transfers of prisoners who have already
begun AOP could be prevented, unless necessitated by medical reasons or reclassification to a high security
level.
The Department of Corrections’ current plan to analyze an integrated AOP waiting list for level I and II
prisoners is an important starting point. To reduce the number of prisoners denied parole in order to
complete AOP, the Department must not only implement this list but must enforce a policy of promptly
transferring people to whatever facility at their security level has an AOP group opening. This will require
a commitment to making treatment completion a high priority in prisoner transfer decisions. 24 The
Department’s stated intent to increase coordination of AOP planning among institutional administrators,
treatment managers and the parole board is a hopeful sign that this commitment is growing.

24. Ideally, a statewide list would govern all required programming, including AOP, sex offender therapy, substance abuse
treatment and GED classes to insure the prisoner’s optimal placement for all needed programs. Such coordination will be
necessary if the MDOC’s ambitious Michigan Prisoner Re-entry Initiative is to succeed, since that process envisions the
systematic implementation of programming from the prisoner’s arrival at the Reception Center.

CAPPS/AFSC

13

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

B. Parole board procedures may compound the waiting list delays.
The lack of centralized planning extends to a lack of coordination between the delivery of treatment
services and the parole decision-making process. Parole board procedures have often failed to accommodate the inability of prisoners to complete AOP causing further unnecessary and costly delays in moving
people to parole.

1. The parole board often conducts parole interviews months before the
prisoner’s earliest release date, when not enough relevant information
is available, and sometimes fails to recognize the prisoner’s inability to
control program access.
By statute, a parole board interview must be conducted at least one month before the prisoner’s earliest
release date.25 However, in what might be considered an excess of efficiency, the board sometimes conducts interviews as much as five, six or seven months before the prisoner becomes parole-eligible. The
board explains this practice by the need to manage the logistics of arranging interviews at numerous locations. Nonetheless, it raises questions in any case where the prisoner is serving a minimum sentence of 18
months or less, especially if the person enters prison with any credit for time served in jail, since there is
little time for the person to build a prison record for the board to review.
These early interviews are also problematic for any prisoner who is waitlisted for AOP. The farther from
the release date that the interview occurs, the less likely the prisoner will have been admitted to AOP, much
less to have completed it. This means a parole decision will be made before the prisoner has an opportunity
to demonstrate any progress in the program or to demonstrate that he or she might reasonably be released
without the program or with a requirement to complete it in the community. It is not unusual for the
parole board to then deny parole based on its conclusion that the prisoner “failed” to complete AOP, even
when the failure was the MDOC’s for not providing timely access. For example:
Robert Smith, described at page 13, first became eligible for release in February 2005. However,
the parole board interviewed him seven months earlier, in July 2004. At that point, Smith had
received no major misconduct reports, had excellent block and work reports and was awaiting
transfer to a facility where AOP was available. The board continued his incarceration for an additional twelve months, until February 2006, stating: “Prisoner who is serving for assaultive offense
has failed to involve himself in therapeutic programming which is designed to allow him to gain
greater insight into his assaultive behavior. Unwilling to parole at this time.”

25. MCL 791.235(1).

14

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Similarly,
Charles McDonald, described at page 11, has twice been denied parole, despite a parole guidelines
score indicating a high probability of release. Although he had begun AOP in a group that was
cancelled and been sent to a camp that did not have the program, the parole board concluded that
McDonald “failed” to complete therapy.
If parole board interviews were scheduled closer to the prisoner’s first release date, the board would have a
stronger basis for deciding whether the prisoner could safely be released without completing AOP or with
a condition that the program be completed in the community. At a minimum, when non-completion is
a factor in the parole decision, the board should familiarize itself with the prisoner’s waitlist status and not
appear to count the prisoner’s inability to gain entry to AOP as a negative.

2. The parole board has ordered longer continuances than are necessary for AOP completion.
The parole board can and does release people who have not had the opportunity to complete AOP when
it is persuaded that doing so does not create an undue risk to the community. However, when the parole
board denies release because someone has not completed AOP, it has typically ordered at least a 12-month
continuance, regardless of the circumstances. In cases where the prisoner is nearing the end of treatment,
months may be wasted between completing AOP and the next scheduled parole decision, known as the
“next official date”. For example:
In December 1989, Tracy Nelson, #205729, was sentenced to two concurrent terms of 18-30 years
for kidnapping and assault with intent to commit murder. When he became eligible for parole on
May 24, 2004, he had completed all recommended programming as well as numerous classes
designed to enhance his skills and ready him for employment outside prison. His parole guidelines
score indicated a high probability of parole.
Although MDOC staff had not recommended it, after 10 years in prison Nelson asked to be evaluated for AOP. He was screened and placed on the waiting list at Macomb Correctional Facility in
September 2000. Nelson was still waiting to begin the program when he was transferred, first to
Chippewa where AOP was not available, then to Camp Koehler where he could not gain admission. He was transferred again in June 2002, to Camp Tuscola. With his earliest release date
approaching, Nelson again sought placement in AOP. He was finally enrolled in August 2003.
Meanwhile, Nelson was assigned to public works at the Frankenmuth Chamber of Commerce and
Convention & Visitors Bureau, leaving camp to work in town five days a week with no Corrections
supervision. He worked there from May 2003 to October 2004, and received a strong letter of
recommendation from his supervisors.
By the time of his parole interview in January 2004, Nelson had been misconduct-free for 9½
years. He had been working outside prison for eight months and was scheduled to complete AOP
within a few weeks of his potential May release date. In addition, his wife had arranged an anger
CAPPS/AFSC

15

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

management course in the community should the parole board desire that Nelson participate in it.
Instead of simply deferring its decision until he had completed AOP, the board continued Nelson’s
imprisonment for 12 months - until May 2005, stating: “Prisoner presently involved in therapeutic
programming designed to allow prisoner to gain insight into his assaultive behavior. Prisoner is
deemed an unwarranted risk to public safety at this time.”
In July 2004, with ten months remaining on his 12-month continuance, Nelson completed AOP
and received an excellent termination report. He immediately sent the board this report and a letter
requesting that it re-consider its decision to continue him in prison for a year. The board replied: “A
copy of your psychological termination report has been placed in your central office file. A few
months prior to your next parole board interview, the board will review your psychological report
and take it into consideration when making their [sic] decision. Your case will be reviewed again
prior to your May 2005 official date.” However, when he was interviewed in January 2005, he was
granted release in February, seven months after successfully completing AOP.
Similarly,
Dan O’Bryan, described at page 9, was continued for 12 months even though he had completed five
months of AOP before reaching his earliest release date. As a result, he served five months after
successfully completing AOP and double the one and a half-year minimum term imposed by the
court.
For people who have finished their minimum terms, there is no reason why the parole decision cannot be
made as soon as an AOP termination report is available. With a waiver from the prisoner, the initial
interview could be deferred until AOP is completed, thereby increasing the likelihood parole will be granted.
Alternatively, continuances can be for only as long as necessary to permit program completion. Whatever
the mechanism, if parole is being denied because AOP has not been completed, it should only be denied
for as long as that problem exists.
Recent parole board data indicates that deferred decisions have become increasingly common and that the
deferral periods are growing longer. In 2002, 462 decisions were deferred for completion of therapy within
90 days; in 2003, 616 such decisions were deferred. In 2004, the deferral period was extended up to six
months and 1,116 decisions were deferred. In 2005, the time limit for deferrals was extended to eight
months and the parole board anticipates deferring more than 1,800 decisions until therapy termination
reports are available.26 Eighty percent will involve AOP; the rest will be for sex offender therapy. This
change in parole board practices will help prevent releases from being delayed for additional months beyond AOP completion. However, the need for this change and the large number of cases involved further
demonstrates the extent to which parole decisions are affected by late admissions to AOP.

26. Parole board memorandum regarding program completion delays, provided March 15, 2005. These deferrals do not, of
course, extend to parole-eligible prisoners who are still on AOP waiting lists.

16

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

3. The parole board fails to order community-based AOP programming as
a parole condition when that alternative would be reasonable.
Release without AOP and parole denial are not the only possibilities. The board routinely makes participation in various treatment programs a condition of parole. It is unclear why this alternative is not used
more often for people who have been unable to complete AOP. Dan O’Bryan, for example, was told by an
Ann Arbor counseling center that he would be a likely candidate for participation in its anger management
program while on parole. Presumably the MDOC could contract with community-based service providers
to provide the full AOP program. Instead of requiring taxpayers to pay for an additional year of incarceration, a low risk offender like O’Bryan could be back in the community, working to support his family and
paying for his own participation in group therapy.

C. AOP is not offered to prisoners housed at security
levels III, IV and V.
While over 70 percent of Michigan prisoners are housed in prisons classified at security levels I or II, nearly
8,400 are at higher custody levels.27 Higher security levels are different in several ways. They have higher
staff to prisoner ratios and are thus more expensive to operate. Prisoners at Level V cannot be doublebunked. The annual cost per Level I prisoner is $21,013 while the cost per prisoner at Level V is $31,397.
Movement within the prison is more controlled as custody levels increase, activities are more limited and
privileges, like visits and the possession of personal property, are fewer. Another critical difference at
higher custody levels is the unavailability of AOP. Although the program statement stresses that priority is
given to the delivery of AOP to Levels I through IV, 28 the reality is that prisoners at Levels III, IV and V are
denied access to AOP.

1. Prisoners who may need AOP the most are discharged to the
community without therapy after serving their maximum sentences.
Prisoners are commonly classified at higher security levels based on institutional misconduct, which may
include an accumulation of non-assaultive misconducts, verbal altercations, or physical fights and assaults
on staff or other prisoners. These prisoners may already have been referred for AOP, based on the crimes for

27. MDOC “Prison Population & Capacity”, Senate Judiciary Committee hearing, February 1, 2005. This figure includes
1,084 at Level III, 5,502 at Level IV and 1,806 at Level V. It does not include prisoners in reception, segregation or other
special use beds.
28. See footnote 9, supra. Another group of prisoners who are now routinely excluded from treatment programs, even when
housed at Level II, are parole-eligible lifers. Operating Procedure 04.06.180-D, Attachment A.

CAPPS/AFSC

17

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

which they were sentenced, or the need for AOP may have become evident through their institutional
behavior. Whatever the case, these prisoners, who may well need AOP the most, do not receive it because
the Department of Corrections does not provide AOP at security levels III, IV and V. If they do not work
their way back down to facilities where AOP is offered, they may ultimately be kept to finish their maximum sentences and then be discharged from custody. When this happens, they are returned to the community without the benefit of therapy designed to control their assaultive behavior and without supervision by a parole officer.
Currently, there are 4,078 prisoners on the AOP waiting lists at Levels III, IV and V. There are 1,188 at
Level 5 alone. In 2004, more than 1,800 prisoners were discharged after serving their maximum sentences.
While it is not known how many did not receive needed assaultive offender therapy, or how many were at
Levels III, IV and V when they were released, the prospects are alarming. In the name of protecting society
from potentially assaultive individuals by denying them parole, over time the MDOC is allowing hundreds
of higher risk prisoners to return to the community without supervision, supportive services or treatment.
Part of the solution is obviously to increase the resources available so that AOP can be delivered at higher
custody levels. There is nothing about their security classification that requires excluding these prisoners
from treatment programming, other than the policy decision that they are low priority for receiving it.
While some particularly disruptive individuals might have to be excluded on a case-by-case basis, even
prisoners at Level V (excluding those in segregation) are permitted to engage in group programming. AOP
should be no exception.
Another practice worth considering is not having currently assaultive offenders “max out.” By granting
parole, under intensive supervision, for
a year before a prisoner is scheduled to
discharge from his or her maximum sentence, the MDOC would retain a role in
. . . simply turning loose someone with
helping the person adjust to the commupoor control over aggressive impulses,
nity as well as the opportunity to act
who has never had any treatment and who
quickly if behavioral problems arise. Perhaps small community corrections cenhas been functioning for years in the
ters with programming geared for pristightly structured environment of a maxioners who would otherwise max out
should be developed.
mum security prison, is as frightening for
In any case, simply turning loose somethe prisoner as it is for the community.
one with poor control over aggressive impulses, who has never had any treatment
and who has been functioning for years in the tightly structured environment of a maximum security
prison, is as frightening for the prisoner as it is for the community. Keeping people with ongoing assaultive
behavior in prison as long as legally possible may intuitively seem like the safest course. However, the

18

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

public may be better protected by a decompression period when the prisoner can re-enter the community
with pre-arranged connections to mental health agencies, close monitoring by a parole officer and assistance with the inevitable readjustment difficulties that cause frustration and anger. That is, the very purposes parole supervision are supposed to serve for prisoners generally may be even more important for the
very people the parole board currently declines to parole at all.

2. Prisoners housed in a Level III, IV or V facility for reasons other than
misconduct are denied access to AOP and then cannot get paroled.
In addition to those housed in higher custody facilities due to their behavior, prisoners may be housed in
these locations due to a lack of lower custody beds, for placement in protective custody or for other administrative purposes. Custody increases for administrative reasons are particularly easy to initiate at multilevel prisons. Since central office approval is not required, wardens may allow movement of Level I or II
prisoners to Level III, IV or V within their own facilities. 29 This leeway is available in 24 prisons, creating
a large opportunity for custody increases solely to meet bedspace needs.
Prisoners who are housed at higher level facilities have no opportunity to enter AOP, regardless of the
reason they are housed there. Although they might otherwise be good candidates for parole, they face the
prospect of being repeatedly denied release because they cannot obtain required programming. In addition
to the obvious unfairness, these parole denials are particularly costly to taxpayers because they increase the
length of stay at more expensive higher security facilities.

3. An opportunity to transfer some prisoners to lower, less expensive
custody levels is lost by withholding AOP from prisoners at Levels III,
IV and V.
Arguably, prisoners whose assaultive behavior within the prison setting results in higher custody have the
greatest need for intervention to change
poor attitudes and behaviors. If AOP
were offered at the higher level facilities
Using treatment programming to reduce
and had the desired effect on at least
the risk of assaults within the prison system
some prisoners, the changed behavior
would permit their custody levels to be
is a desirable way to protect staff and
reduced. Using treatment programming

other prisoners.

29. MDOC Policy Directive 05.01.130, “Prisoner Security Classification”, paragraph “M”, “2”. An exception exists under
paragraph “R”, stating: “Female prisoners shall not be classified to Level V unless prior approval is obtained from the CFA
Deputy Director or designee.”

CAPPS/AFSC

19

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

to reduce the risk of assaults within the prison system is a desirable way to protect staff and other prisoners.
It is also cost-effective. Each reduction in the custody of a prisoner from Level V to Level IV saves $4,146
per year. Each reduction from Level IV to Level II saves $4,676. AOP is not a magic bullet and may well
not be sufficient to change the behavior of some of the most difficult prisoners. But at least attempting to
provide AOP at higher security levels may shed light on how treatment can be provided more effectively to
prisoners who are hard to manage. Moreover, given the relative costs of AOP programming and of custody
at higher security levels, even a partial success rate would make the investment in treatment worthwhile.

D. Parole has been delayed by the application of AOP eligibility criteria.
The MDOC’s own eligibility criteria for AOP participation have actually delayed parole in some cases.
These problems are relatively easy to correct.

1. Prisoners incarcerated before the designated offense list was changed
have been denied parole because they were never referred for AOP.
In 2002, the MDOC’s Psychological Services Unit expanded the list of assaultive crimes that automatically
require referral for AOP assessment. The change was applied to new prisoners at the Reception Center but
the files of prisoners already committed were not flagged so that they could be added to AOP waiting lists.
The end result for some prisoners was parole denial for not having completed AOP and a parole board
referral for assessment. For example:
Derrick Gooley, #236951, was sentenced to a term of 10-22 years for unarmed robbery in 1994.
When an elderly woman who employed him as a home health aide fired him for arriving late,
Gooley, who had a substantial drug habit, pulled two rings from her finger and fled. He turned
himself in to police the next morning.
In December 2000, Gooley tried to enter AOP but was denied because his offense did not qualify
him for the program and he had shown no assaultive behavior in prison during the previous five
years. In June 2001, his request to enter AOP was again denied. The AOP offense list was changed
in June 2002 and unarmed robbery was added. In March 2003, because he was nearing his earliest
release date, Gooley again asked for program admission. Failing to recognize that the eligibility
criteria had changed, the therapist concluded: “a third screening will not change anything … you do
not need one more assessment just because you’re going to the parole board.”
Gooley’s earliest release date was February 1, 2004. When the parole board interviewed him five
months before that, it questioned his “failure” to enter AOP and denied parole for another 12

20

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

months. Because the board recommended that he be screened for the program, Gooley was immediately admitted. He completed AOP on September 1, 2004 and was paroled on November 13th,
nearly a year past his earliest release date, at an additional cost of more than $17,000. 30
The change in the designated offense list even affected those prisoners who had been paroled without AOP
but had been returned for a technical parole violation that did not involve any assaultive behavior. In these
cases, when the underlying offense was one of those added to the list, AOP became a requirement. For
example:
In 1997, William Cousino, #255186, began serving sentences of 2-20 and 1-10 for home invasion
and assault with intent to commit great bodily harm. He had assaulted a male friend of his wife’s
shortly after he and his wife had separated. He had no prior adult or juvenile record. He was not
referred for AOP because his offense was not on the list.
Cousino completed boot camp and was paroled in 1997, but his parole was revoked twice for
technical violations. The first revocation was based on substance abuse. Cousino was paroled again
after 11 months. The second revocation, in October 2001, was for driving on a suspended license in
a car with expired plates. That time, the parole board issued a 12-month continuance, followed a
year later by another 12-month continuance. During the second continuance, the AOP referral
offense list was changed to include Cousino’s original offense. It took another six months before
Cousino, who was working in the community on a gate pass, gained entry to the program. In the
interim, the parole board issued a third 12-month continuance, specifically so he could complete
AOP. He was finally paroled again in June 2004. 31
A thorough audit of prisoner records involving the added offenses would ensure that oversights are corrected before parole is denied to permit program completion. Long-term prisoners who may not have been
referred to AOP for other reasons no longer in effect should also be reviewed. The cost of such an audit
would surely be less than the cost of multiple year-long parole denials.
In addition, criteria should be adopted for determining whether to waive AOP participation or permit
completion of a similar program in the community for prisoners whose belated assessments resulted from
changes in eligibility criteria. Requiring Gooley and Cousino to complete AOP after they were paroleeligible, when neither had engaged in assaultive conduct for years, is a questionable use of two scarce
resources – openings in AOP and expensive prison beds.

30. A full profile is available at www.capps-mi.org/pdfdocs/profiles/gooley.pdf
31. A full profile available at www.capps-mi.org/pdfdocs/profiles/cousino.pdf

CAPPS/AFSC

21

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

2. The subjective application of motivational criteria often denies programming to the very individuals that the program was designed specifically to rehabilitate, i.e. those who did not recognize the problem
behavior or their need to change.
Historically, the criteria for admission to AOP included: 1) accepting responsibility for assaultive behavior
by acknowledging having engaged in it and that it was illegal; 2) admitting the need to change behavior,
gain insight and/or acquire new skills because of past assaultive behavior or potential for future assaultive
behavior; and 3) establishing appropriate goals and objectives relating to the individual’s assaultiveness.
These criteria effectively eliminate participation in AOP by any offender who denies or minimizes guilt to
any degree or is unable to demonstrate understanding and interest to the therapist’s satisfaction.
Many people convicted of assaultive offenses continue to feel strongly that they acted with some degree of
justification. In some cases, these feelings may find support in the official record. In others, the person
may be engaged in massive denial about the nature of his or her own behavior. Either way, prohibiting
participation in AOP would seem to be counterproductive. People in the first group, who may be genuinely less dangerous to the public, have their chances of obtaining parole reduced by their “refusal” to meet
AOP entry criteria. People in the second group, who may need AOP the most, are denied entry because
they have not received enough therapy to recognize that they need it.
The application of these criteria raises a second concern. Since literally 100 different therapists at the
facility level are making subjective judgments about whether prisoners are sufficiently motivated to enter
AOP, the potential for inconsistent application is high.
The Psychological Services Unit acknowledges these concerns. However, no formal changes have yet been
implemented.

E. The parole board denies parole based on its disagreement with a
positive assessment by the AOP therapist.
All the time and effort prisoners put in to entering and completing AOP is no guarantee of release on
parole. It is not uncommon for the parole board to deny release to prisoners who have finished AOP and
received excellent reports from a therapist. In such cases, the board often simply asserts that it remains
unpersuaded the prisoner is a good risk for release. No critique of the therapist’s assessment is given. No
specific facts in support of the board’s own conclusion are offered. When the parole board disagrees with
therapy reports regarding the prisoner’s progress, repeated parole denials ensue. For example:
In November 1999, Nicholas Seymour, #312131, was involved in a fist fight with an acquaintance
that resulted in the victim’s death from a head injury. Seymour pled guilty to voluntary manslaughter and was sentenced to 3 – 15 years.

22

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Seymour has an excellent institutional record. He earned outstanding reports from work supervisors
and completed courses in substance abuse prevention and anger management.
Despite his repeated inquiries, Seymour was not placed in AOP until April 2002, just seven months
before his earliest release date.
When he was first interviewed by the parole board in September 2002, Seymour scored favorably on
the board’s own parole guidelines, was housed in minimum security and had a gate pass to work
outside the facility. However, the parole board continued him for 12 additional months, stating:
“The offender’s involvement in AOP designed to address his anger and assaultiveness is incomplete.
His meaningful insight into the causes of this assaultive crime is also incomplete. Departure warranted. Still view as risk.”
By the time of his 2003 parole interview, Seymour had completed AOP. The therapist provided a
strongly positive completion report supporting parole, concluding that Seymour’s risk of re-offending
had been greatly reduced. Nonetheless, the parole board denied parole, stating: “During Parole
Board interview prisoner continues to minimize his level of responsibility for this offense. Despite
completion of recommended therapy, the Parole Board is not assured that prisoner’s risk of re-offending has been diminished.”
Seymour asked for reconsideration, saying that the decision to deny parole was based on inaccurate
information and that he had been cheated out of a fair hearing because he was able to talk for
“maybe 20 seconds” before the interviewer cut him off. The board responded with a form letter
saying there was no basis for changing its decision.
In July 2004, the parole board notified Seymour that it was denying release for the third time,
stating: “Despite completion of recommended therapy, the P.B. is not assured that prisoner’s risk of
re-offending has been diminished.” It justified the additional 12-month continuance on the basis of
“the very brutal crime that occurred in front of victim’s child” and called Seymour “an unwarranted
risk to public safety.” With this decision, the board effectively doubled the three-year minimum
sentence imposed by the judge. His next re-consideration date is in November 2005.
Cases like this are troubling. Seymour took advantage of everything the MDOC had to offer, including
nearly 12 months of group therapy. The detailed assessment by the therapist who spent at least 100 hours
with him seems virtually disconnected from the terse conclusions reached by the parole board after its very
brief interviews. We cannot tell from the board’s 2003 statement what Seymour did to “minimize his level
of responsibility.” The reference in the 2004 statement to the facts of the crime suggests that the parole
board may really be disagreeing more with the Court’s minimum sentence than with the therapist’s assessment of Seymour’s current risk.
We do know that Nicholas Seymour has now served three years more than the judge, bound by statutory
sentencing guidelines, determined was appropriate for punishment – at an additional cost to taxpayers of
over $60,000. We also know that he is far from unique. There are currently 1,484 prisoners
who have served their minimum sentences and completed AOP but have been denied parole.
CAPPS/AFSC

23

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Since prisoners have no right to appeal the denial of parole, there is no
administrative or judicial review of
In cases where the parole board disagrees
decisions like this. Yet in cases where
directly with a therapist’s findings, some form
the parole board disagrees directly
with a therapist’s findings, some form
of review seems warranted.
of review seems warranted. A review
committee should assess the record,
interview the prisoner and the therapist, and issue a report with its findings to be made part of the prisoner’s
medical file. The committee could be empowered to recommend reconsideration by the entire parole
board when warranted. Members of such a committee could include the Department of Corrections’
Mental Health Services Manager, the Deputy Director of the Field Operations Administration and a psychiatrist who has direct prisoner contact under the Corrections Mental Health Services contract.

III. Where are they today? An analysis of the 2004
waiting lists one year later
To better understand the impact of delays in providing assaultive offender programming, a random sample
was drawn from the February 4, 2004 AOP facility waiting lists. Every tenth name was selected from the
list of 1,348 prisoners at security levels I-IV. 32 Of the 134 people selected, two were in the special
alternative incarceration (SAI) program, a 90-day boot camp. People who complete SAI are paroled automatically. Since these prisoners would not have the opportunity to complete AOP but would be paroled
anyway, they were dropped from the sample.
The February 1, 2005 waiting list was reviewed in an effort to determine what had occurred to each of the
sample prisoners during the intervening year. These 132 prisoners fell into one of four categories: discharged upon completing the maximum sentence, paroled, still in prison and still on AOP waiting list and
still in prison but no longer on AOP waiting list.

A. Discharged
Ten members of the sample (7.5 percent) served their maximum sentences and were discharged. They
were serving terms ranging from 1 year, 1 month – 2 years for attempted felonious assault to 7-15 years for

32. Level 5 and 6 prisoners were excluded from this sample because people are virtually never paroled from these custody
levels in any event. A separate sample of the Level 5 prisoners was examined to determine how many have been discharged
after serving their maximum.

24

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

assault with intent to commit armed robbery.
Three of the ten were lower security prisoners serving relatively short sentences. However, the other seven
were all housed at Level 4 as of February 2004 and all but one had already been denied parole. Available
MDOC reports do not indicate whether any of the discharged prisoners ever completed AOP. If the seven
who were housed at Level 4 did not have their custody levels reduced before serving their maximum
sentences, they would not have had the opportunity to enter AOP. Given their discharge dates, four
would not have had time to complete the program in any event.
Thus it appears that nearly eight percent of the prisoners at Levels I-IV who are on waiting lists for AOP
when they are within a year of completing their minimum sentences may ultimately complete their maximum sentences and be discharged from custody without therapy and without community support or
supervision. Extrapolating from the sample to the entire 2004 waiting list, this would involve more than
100 prisoners. Five percent, or 67 of the 2004 total, were classified at Level IV, suggesting that in addition
to committing assaultive crimes they had behavioral difficulties while incarcerated.
In February 2004, 449 prisoners who were within a year of or had passed their earliest release dates were on
AOP waiting lists at Level V facilities. A random sample of these cases, 44 in all, was also examined.
Though some had completed their minimum terms many years earlier, 38 of these prisoners were still
incarcerated. One, who was housed at the Michigan Youth Facility for a robbery he had committed at the
age of 16, was paroled after serving his four-year minimum.
The remaining five Level V prisoners present a disturbing prospect. They were discharged from custody
after serving their maximum sentences. These men, between the ages of 34 and 45, had apparently never
participated in assaultive offender programming. None had a conviction for a homicide or sex offense and
none had a minimum sentence longer than 4 years, but all had served at least 11 years. This indicates they
were classified to maximum security because they posed management problems for institutional staff,
perhaps because of mental health problems. While at Level V, their movements and interactions with
other people were severely restricted. Yet these prisoners were released directly into the community without any supervision or support. They included:
•

Prisoner A. Sentenced to 4-12 years for armed robbery and attempted felonious assault; served 11
years.

•

Prisoner B. Sentenced to 3-15 years for unarmed robbery; served 15 years.

•

Prisoner C. Sentenced to 4-20 years for breaking and entering a building; served 20 years.

•

Prisoner D. Sentenced to four consecutive terms of 3-10 years for larceny from a person, 2 ½ – 4 years
for assaulting a prison employee, 2-4 years for malicious destruction of property and 1-4 years for
assaulting a prison employee; served 22 years.

CAPPS/AFSC

25

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

•

Prisoner E. Sentenced to terms of 2-4 years for child abuse and 1 year, 4 mos.- 4 years for attempted
larceny in a building and consecutive terms of 1 ½ to 5 years for prison escape and 6 mos.- 5 years for
being a prisoner in possession of a weapon; served 13 years.

These five prisoners constitute 11 percent of the Level V sample. In February 2005, there were 470
prisoners on the Level 5 waiting lists who were within a year of or had passed their earliest release dates.
Based on the sample, we can anticipate that, within 12 months, 52 of them will be discharged from
custody after serving their maximum terms.

B. Paroled
Thirty-three of the prisoners waitlisted for AOP in February 2004, or 25 percent, had been paroled by
February 2005. Nineteen were paroled at their earliest release dates; 14 were past their earliest dates.
Again, it cannot be determined from available MDOC reports whether these prisoners ever gained entry to
the program. Thus, it is not possible to identify any patterns in parole board decision making regarding the
release of prisoners without AOP. However, it is clear from their release dates that at least 22 of the parolees
would not have had time to complete AOP. In addition, five people had been housed at Level III or IV
facilities that did not provide AOP. Thus, at a minimum, two-thirds of those paroled had not completed
AOP and it may be that none of them had.
When the sample percentage is applied to all 1,348 Level I - IV prisoners on waiting lists in February 2004,
it appears that at least 222, and perhaps as many as 333, were paroled without taking AOP. While nine of
these parolees had parole conditions that said “Treatment program (mental health/other) as approved by
agent”, none were expressly required to complete assaultive offender programming of any kind.
If the parole board is correct and these individuals could be safely released without treatment, then the
question becomes why they were on waiting lists for AOP at all. If assessment criteria that looked beyond
the name of the offense are applied at the reception center, people who do not need AOP could be screened
out there. That would reduce the size of the waiting list, permit a more accurate prediction of resource
needs and make it easier to manage inter-facility transfers.

C. Still in prison, still waiting
Thirty-eight prisoners – 29 percent of the Level I - IV sample – were still on a waiting list for AOP in
February 2005. All but two are now past their earliest release dates. (Two were prosecuted for offenses
committed in prison – one for escape and one for possessing a weapon – and thus their earliest release dates
changed.)

26

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Table 2. Prisoners Still on AOP Waiting Lists
2004
Custody
Level

2005
Custody
Level

1

I

I

2

I

I

3

I

4

I

5

Prisoner

CAPPS/AFSC

2004 Position
on Facility
Waiting
List
COT
2

2005 Position
on Facility
Waiting
List
OCF

1

WCF

21

WHV

11

I

CTU

17

CTU

19

I

WCF

31

CBI

3

I

I

CDW

7

CDW

3

6

I

I

CTU

27

CTU

30

7

I

II

CLE

3

MRF

5

8

I

I

CKT

7

CKO

9

9

I

I

KTF

51

KTF

9

10

II

II

RRF

13

RRF

12

11

II

I

WCF

4

WHV

13

12

II

II

OCF

44

OCF

15

13

II

II

JMF

20

OCF

18

14

II

II

MRF

15

MRF

21

15

II

II

RRF

3

RRF

4

16

II

II

OCF

34

OCF

8

17

III

III

URF

24

URF

16

18

III

III

URF

34

URF

22

19

III

IV

URF

54

URF

25

20

III

IV

IBC

14

LRF

3

21

III

III

URF

44

URF

30

22

III

III

URF

4

URF

4

23

IV

IV

HVM

25

HVM

1

24

IV

IV

NRF

4

NRF

1

25

IV

V

URF

24

URF

1

26

IV

IV

IBC

21

IBC

14

27

IV

IV

DRF

15

ARF

15

28

IV

IV

HVM

35

URF

15

29

IV

IV

URF

4

JCF

2

30

IV

IV

HVM

15

HVM

23

31

IV

IV

IBC

41

IBC

25

32

IV

IV

ARF

14

HVM

28

33

IV

IV

ARF

4

ARF

3

34

IV

IV

IBC

81

IBC

48

35

IV

I

IBC

61

OTF

5

36

IV

IV

MRF

10

LRF

5

37

IV

II

TCF

2

TCF

6

38

IV

IV

IBC

31

MRF

7

27

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Table 2 (on page 27) shows the security level and waiting list position of each of these prisoners in both
2004 and 2005. Abbreviations identify the facility.33
Table 2 reveals that in February 2004, 23 prisoners who were within a year of or had passed their earliest
release dates were housed at facilities that did not provide AOP: six were at Level III, 16 were at Level IV
and one was at Camp Kitwen (Level I). A year later, 20 of these prisoners were still at facilities without
access to AOP: four were at Level III, 15 were at Level IV and one was at Level V. The remaining 18
prisoners were simply sitting, awaiting their turns, at facilities that have AOP groups.
In 2004, the positions of these prisoners on the waiting lists at their respective facilities ranged from
second to 81st. By 2005, the range had been narrowed from first to 48th.
Most noteworthy is the way in which individual’s positions on waiting lists change. Twenty-six people
moved up. Of these, 17 were at the same facility during both years and nine had been transferred.
Ten actually lost ground. Of these, five were at the same facility and same security level in both years.
That is, they were on the very same waiting list but their positions got worse, presumably because other
prisoners whose earliest release dates were closer had been transferred in. Three had been transferred to
different prisons, one because of a custody increase.
Two were at the same facility but their custody levels had decreased. Ironically, these prisoners will have to
wait longer to enter AOP, thereby decreasing their chances of obtaining parole, because they improved
their behavior, thereby increasing their chances of obtaining parole.
Finally, two sample members were stuck in the same position. One remained in fourth place on the URF
list; one was 15th on the DRF list in 2004 and 15th on the ARF list in 2005.
The arbitrariness of this scheme and the frustration it causes prisoners who are attempting to enter AOP is
apparent. And, of course, improving one’s position on the waiting list of a Level III or IV institution that
does not provide AOP anyway is essentially meaningless.
Six sample members are now within 18 months of serving their maximum sentences and being discharged.
Two are housed at Level III and two are housed at Level IV. Thus, unless they are transferred to reduced
custody soon and placed into AOP groups almost immediately, they will join the ranks of prisoners who
are discharged without completing AOP.
When one extrapolates from the sample to the entire group, it appears that nearly 400 prisoners on the
AOP waiting lists at Level I, II, III and IV prisons will still be on those waiting lists a full year later. Some
people will have moved closer to the top. One hundred will actually have moved lower down, most
through no fault of their own. The one thing that is certain is that virtually all of them will be past their
earliest release dates and will have been denied parole. Sixty will serve their maximum sentences and reenter the community without therapy or supervision.

28

33. The entire population of women prisoners at Western Wayne Correctional (WCF) was moved in
December 2004 to Huron Valley Complex -Women’s (WHV). These are the same institution for waiting list
purposes.

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

D. Still in prison, no longer on the AOP waiting list
The last group in the sample consists of 51 people, 38.6 percent of the total, who are still in prison but
whose names no longer appear on any facility’s AOP waiting list. Available MDOC reports give no indication of why these prisoners are no longer on waiting lists. At least five explanations are possible. These
prisoners may:
•

Be currently enrolled in AOP

•

Have entered and completed AOP

•

Have been admitted to AOP and subsequently terminated without completing the program

•

Have been assessed for AOP when an opening occurred and been denied admission

•

Have been inadvertently omitted from a waiting list after transferring to a different facility.

What we do know is that 40 of the de-listed prisoners are currently housed at security levels I and II, that
is, facilities where AOP groups exist. Ten of them had been tenth or higher on a waiting list in 2004 and
were at the same facility, at the same custody level, in 2005. Those ten, at least, may well have gained access
to an AOP group.
On the other hand, some had been as far down on Level I waiting lists as numbers 38, 41 and 50, so it is
unlikely they had the opportunity to enter AOP in the intervening year. And nine are currently at Level III
or IV facilities, where AOP groups are not conducted. Thus some other explanation for why these prisoners are no longer waitlisted must exist.
The one thing that is known for sure about these 51 prisoners is that all but one are past their earliest release
dates and are being denied parole. Some, including 10 who are currently housed at Level I or II, are more
than two years past their first release dates.
The sample data translates into nearly 500 people who were within a year of or past their earliest release
dates in 2004 and who have simply disappeared from waiting lists. To the extent that some of these
prisoners have been able to access AOP, that is, of course, to the good. But then the question arises as to
why they have not been paroled. A close examination of this group could help answer questions about how
many people are denied admission to AOP, and why; how many start AOP but fail to complete it, and why;
how many people complete AOP but are denied parole nonetheless, and why; and how many people fall
off waiting lists because administrative errors are made when they are transferred between facilities.
At a minimum, a tracking system that reveals when and why each prisoner has been removed from an AOP
waiting list would seem desirable.

CAPPS/AFSC

29

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

IV. Recommendations
1. Increase the availability of assaultive offender programming at each MDOC facility by hiring more
therapists solely to conduct AOP groups. These therapists could be full-time state employees, contractual employees hired to eliminate the waiting list backlog, or qualified graduate students in clinical
psychology or social work.
2. Maintain a statewide roster of prisoners awaiting entrance to AOP. As openings in groups occur, fill
them with the next wait-listed prisoners classified at the appropriate security level, transferring people
between facilities as necessary. Similar statewide rosters should be maintained for sex offender therapy
groups and GED classes.
3. Prohibit transfers of prisoners who have already started participating in required treatment and educational programs unless necessitated by specified security reasons or medical emergencies.
4. Provide AOP at security levels III - V.
5. Prohibit the denial of parole solely for failure to complete a particular treatment or educational program
if the MDOC did not make the program available to the prisoner sufficiently in advance of the prisoner’s
earliest release date to permit completion or if the prisoner was ineligible to participate in the program
under the department’s own criteria. Permit parole decisions to be deferred for up to three months to
allow for completion of a program in which the prisoner is already enrolled.
6. Contract with service providers in the community to offer AOP to probationers and parolees. Establish
criteria to determine when prisoners should be permitted to complete AOP in the community as a
condition of parole and when the requirement for AOP should be waived altogether.
7. Establish an administrative review process for cases in which the parole board denies parole to a prisoner
with a favorable AOP termination report, based on the board’s belief the prisoner is still a risk to the
community. The review committee should be comprised of the MDOC’s Deputy Director of Field
Operations Administration, the Director of the Psychological Services Unit and a DCH psychiatrist
who provides mental health treatment to prisoners. The committee should assess the AOP record and
the parole board file. It should have the authority to interview the prisoner. The committee should
make written findings and recommend reconsideration by the entire parole board when warranted.
Findings should be copied to the prisoner and the prisoner’s medical record.
8. Conduct all future assessments for AOP at the Reception Center to ensure consistent decision-making
about who should be admitted to the program.
9. Eliminate “motivational criteria”, such as acknowledging illegal behavior and admitting the need to
change, as a basis for admission to AOP. Reinstate on waiting lists all prisoners who have already been
denied admission to AOP based on these motivational criteria.
10. Audit the files of all prisoners serving sentences for offenses currently designated as assaultive to ensure
that appropriate AOP recommendations have been made.

30

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

V. Conclusion
Providing treatment to prisoners whose attitudes and behavior threaten the safety of others is a critical
function of the Michigan Department of Corrections. Not only does successful treatment reduce the risk
to the public when these prisoners are released, it enhances the safety of institutional staff and other
prisoners.
Assaultive offender programming is also a wise financial investment. AOP completion can give the parole
board assurance that prisoners who have served their minimum terms are safe to release, thereby increasing parole rates and decreasing the cost of incarceration. Successful AOP completion could also permit
reductions in custody level for prisoners who have learned to control their conduct, thereby reducing the
need for expensive maximum-security beds.
Delivering a year-long group therapy program at prisons and camps all over the state is not an easy task.
Sufficient resources must be dedicated, qualified staff must be retained and the appropriate prisoners must
be placed in the program at the appropriate times. The current enrollment of more than 2,000 prisoners
in AOP at several dozen Level I and II facilities is not an insignificant accomplishment.
Nonetheless, a close look at how AOP programming is being delivered raises numerous important issues.
The insufficient number of groups being conducted and the woefully inefficient system of matching
prisoners to treatment openings has left more than 2,200 Level I and II prisoners who are less than a year
from their first parole dates still waiting to begin AOP. The parole board estimates that 1,440 prisoners
enrolled in AOP will have their release decisions deferred, many for six, seven or eight months, until
therapy termination reports are available. Continuing to incarcerate parole-eligible prisoners because
they have not been able to access an MDOC treatment program unnecessarily bloats the size of our prison
system and costs taxpayers millions of dollars.
Although we know that hundreds of prisoners are denied parole each year because they have been unable
to complete AOP, the relationship between AOP and parole is far more complicated than that. The
available data indicates that hundreds of prisoners who had been referred for AOP assessment are paroled
each year without completing the program. This suggests that many people are being referred to AOP
unnecessarily and that waiting lists need not be so long.
On the other hand, 1,500 prisoners who have completed AOP are still being denied parole. Despite
positive assessments from MDOC therapists, the parole board is concluding that these prisoners pose a
current threat to public safety. Without a process for reviewing parole board decisions, it is impossible to
reconcile these apparently conflicting assessments, or to determine the extent to which parole-eligible
prisoners remain incarcerated unnecessarily, again at great cost to taxpayers.
Still another important issue arises from the fact that AOP is not offered to Level III, IV and V prisoners.
As a result, people are being denied the opportunity to learn to control assaultive tendencies and to work
their way down to lower custody levels. Even more importantly, maximum security prisoners are being
discharged after serving their maximum sentences without having received treatment and without parole
supervision in the community.
CAPPS/AFSC

31

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Taking the steps necessary to improve the delivery of assaultive offender programming to the right
Michigan prisoners will require a change in MDOC priorities and a shift in the allocation of resources.
Developing more consistency between the assessments made by treatment personnel and parole board
members will require closer examination of how each group performs its function. However, making
such efforts will pay off with increased public safety and decreased prison spending. Not taking these
steps would be truly penny-wise and pound-foolish.

32

CAPPS/AFSC

Penny-Wise & Pound-Foolish:
Assaultive offender programming and Michigan’s prison costs

Appendix
MDOC, Operating Procedure 04.06. 180-D, 1/13/03, Attachment A
CRITERIA FOR ASSAULTIVE OFFENDER PROGRAM ASSESSMENT
If the prisoner is currently serving a sentence for one of the following offenses she/he is recommended for Assaultive
Offender Program assessment unless serving a life sentence.
750.136 Cruelty to children
750.136(A) Torturing of children
750.136(B) Child abuse
750.136(B2) Child abuse, 1st degree
750.136(B3) Child abuse, 2nd degree
750.136(B3B) Child abuse, 2nd degree-can cause harm
750.136(B3C) Child abuse, 2nd degree-cruelty
750.136(B4) Child abuse, 3rd degree
750.136(B5) Child abuse, 4th degree
750.197(C) Assault of jail/prison employee, etc.
750.204 Sending explosives intent to injure
750.207 Placing explosives causing injury
750.210 Possession of bomb(s) with intent
750.234(A) Firearm discharge from motor vehicle
750.234(B) Firearm discharge from dwelling
750.234(C) Firearm discharge at law enforcement
750.317 Homicide, murder 2nd degree
750.321 Manslaughter
750.324 Negligent homicide
750.325 Manslaughter, motor vehicle
750.327 Death due to explosives
750.328 Death due to explosives, intent destroy
750.349 Kidnapping
750.349(A) Prisoner taking another as hostage
750.349(S) Kidnapping/soliciting
750.350 Kidnapping; child under I4
750.397 Mayhem
750.411(H) Stalking
750.411 (I) Aggravated stalking

CAPPS/AFSC

750.506(A) Assaults while lawfully imprisoned
750.517 Entering train for robbery
750.529 Armed robbery; aggravated assault
750.529(A) Car jacking
750.530 Robbery, unarmed
750.531 Robbery, bank, safe or vault
750.531(A) Robbery, bank
750.72 Burning dwelling house
750.72A Arson dwelling house
750.72B Arson dwelling house curtilage
750.81 Assault & battery; domestic assault
750.81 (3) Domestic violence, 2rid offense
750.81 (4) Domestic violence, 3rd offense
750.81(A) Assault & inflict serious injury
750.81 (A2) Domestic violence - aggravated
750.81(A3) Domestic violence - aggravated 2rid
750.82 Felonious assault
760.82(2) - Assault with dangerous weapon
750.83 Assault/Intent to commit murder
750.84 Assault/Intent great bodily harm
750.86 Assault/Intent to maim
750.87 Assault/Intent to commit felony
750.88 Assault/intent to rob unarmed
750.89 Assault/Intent to rob armed
750.91 Attempted murder
752.191 Felonious driving

33