Washington State
Institute for
Public Policy
110 Fifth Avenue Southeast, Suite 214

•

PO Box 40999

•

Olympia, WA 98504-0999 •

(360) 586-2677

•

www.wsipp.wa.gov

January 2007

THE DANGEROUS MENTALLY ILL OFFENDER PROGRAM:
COST EFFECTIVENESS 2.5 YEARS AFTER PARTICIPANTS’ PRISON RELEASE
Reductions in criminal recidivism attributed to the
“Dangerous Mentally Ill Offender” (DMIO) program
in the previous Washington State Institute for Public
Policy (Institute) evaluation are sustained at the 2.5year mark. The subsequent reduction in felonies
associated with the program is valued, by taxpayers
and crime victims, at approximately $820 per
participant minus program costs; this represents a
return of about $1.03 for every public dollar spent on
the program. Approximately 165 clients are enrolled
in the program in a given month.

Summary
Washington State’s DMIO program, enabled by
the 1999 Legislature, identifies mentally ill
prisoners who pose a threat to public safety and
provides them services and treatment up to five
years after their release from prison. Our analysis
of 100 DMIO participants 2.5 years after release
from prison indicates that the program:
9 Reduces new offense rates by 38%.
9 Reduces felony recidivism rates 45%.

In 1999, legislation was passed to better identify and
provide additional mental health treatment for
mentally ill offenders released from prison, who
pose a threat to public safety, and agree to
participate in the program.1 A “Dangerous Mentally
Ill Offender” is defined by the legislation as a person
with a mental disorder who has been determined to
be dangerous to self or others. Through interagency
collaboration and state-funded mental health
treatment and support services, the legislation
intends to promote the safe transition of these
individuals to the community.
The original legislation directed the Institute and the
Washington Institute for Mental Illness Research
and Training to evaluate the program. The 2005
evaluation compared 1.5-year outcomes of DMIO
participants admitted to the program during its first
two years with outcomes of a similar comparison
group of offenders from the Community Transition
Study (CTS) who were released in 1996 and 1997.2
Subsequently, the legislature budgeted funds for the
Institute to continue the evaluation. The DMIO
program is intended to serve participants up to five
years after prison release; this analysis re-examines
criminal recidivism outcomes 2.5 years post-release.
A detailed report on program costs and
implementation is also available.3
1

Chapter 214, Laws of 1999.
D. Lovell, G. Gagliardi, & P. Phipps. (2005). Washington’s
Dangerous Mentally Ill Offender Law: Was community safety
increased? Olympia: Washington State Institute for Public Policy.
3
D. Lovell. (in press). Washington’s Dangerous Mentally Ill
Offender Law: Community safety, costs, and program
development. Olympia: Washington State Institute for Public
Policy.
2

9 Has yet to demonstrate a statistically
discernable effect on violent felonies.
Using methods developed by the Institute for
previous crime studies, these recidivism outcomes
were used to estimate the total economic impact
of the program for both taxpayers and victims of
crime. The state spends $24,280 per DMIO
participant. For taxpayers and victims, the
DMIO program generates:
9 $25,100 in benefits per participant.
9 $1.03 for every dollar spent.

Previous Findings
The 2005 report demonstrated that the DMIO
program significantly reduced recidivism after 1.5
years.4 Overall, the program appeared to be
accomplishing its other principal objectives such as
improved delivery of social services and improved
living situations. The benefit-cost analysis in that
report indicated that the reductions in DMIO
recidivism generated financial benefits to taxpayers
that were less than program costs. This report
provides an improved estimate of costs and benefits
based on newly acquired information on per-person
program expenditures and a re-estimate of the total
economic benefits to taxpayers and crime victims.

4

Lovell et al. (2005).

Exhibit 1

DMIO and Comparison Groups

Significant Pre-Existing Differences Between
DMIO Participants and Comparison Group

To evaluate the program, it is necessary to compare
DMIO participants with a similar group of offenders
(comparison group) who were released without the
interagency coordination and supplemental funding for
services created for the DMIO program. Consistent
with the 2005 study, the selected comparison group
comprised mentally ill offenders, released from
Washington State prisons in 1996 and 1997, who were
part of the Community Transitions Study (CTS). The
DMIO study group consists of 113 participants
released by the end of 2002, the same individuals
followed in the previous study.

Pre-Release Variables
Female*
≤ age 25 at release**
Time served (days)*
Previous felonies**
Previous violent
felonies*
Previous sex felonies*
Previous drug felonies*
Misdemeanor assaults*
Misdemeanor offenses**
Index violent offense*
Felony risk probability*

One DMIO participant died immediately after release
and was excluded from the analysis. Other
participants were omitted because they were
committed to a state hospital over the follow-up
period, transferred out-of-state, or committed as
sexual predators. The outcomes of the remaining 100
participants were examined. Their outcomes were
compared with the outcomes of 287 individuals in the
comparison group. Subjects in both groups were
examined over an equal length of time, 2.5 years after
release.5

Means/Proportions
Comparison
DMIO
33%
9%
19%
9%
799.9
1362.4
3.98
3.33
.72

1.47

.16
.97
.66
3.16
37%
.41

.32
.32
1.37
3.29
82%
.29

* Statistically significant, p<.01
** Statistically significant, p<.05

The observed differences between the recidivism
rates of DMIO participants and the comparison
group could potentially be caused by factors other
than DMIO participation. To account for this
possibility, we conducted a logistic regression
analysis that included, as explanatory variables, all
pre-existing differences detailed in Exhibit 1 (in
addition to DMIO treatment).

Key Methodological Issues
Due to ethical and political concerns about denial of
service and public safety, a random assignment
research design was not used for this study. Rather,
we used a quasi-experimental approach that relied on
a comparison group from the CTS. Multivariate
statistical controls (logistic regression) were also used
to account for observed differences between the two
groups.

There are several limitations to the research design
adopted for this study:

Pre-existing differences were found between DMIO
participants and comparison subjects and are reported
in Exhibit 1. More comparison group subjects were
female or 25 years old or younger upon release; most
importantly, they also had more previous felonies,
though fewer violent ones. As a result of these and
other differences, the comparison group was found to
be at substantially higher risk of felony recidivism than
the DMIO group (41 percent versus 29 percent
respectively).6
5

This study relied on databases maintained by the
Administrative Office of the Courts; DOC; the Department of
Social and Health Services Mental Health Division, Division of
Alcohol and Substance Abuse, and the Research and Data
Analysis Division; and Department of Health.
6
G. Gagliardi, D. Lovell, P. Peterson, & R. Jemelka. (2004).
Forecasting Recidivism in Mentally Ill Offenders Released From
Prison. Law and Human Behavior 928(2): 133-155.
2

•

Individuals in the comparison group were
released from prison more than four years
before DMIO participants were released.
During the intervening period, changes in
factors such as interagency coordination and
community supervision could account for
some effects attributed to the DMIO program.

•

Statistical controls account for significant
observable differences between the study
groups. Possible unobserved differences,
however, such as motivation and participant
selection, may bias the estimate of program
effects. Consequently, for the benefit-cost
analysis, we discount the estimated effect size
to arrive at a more conservative estimate of
the economic outcomes.

•

This analysis of DMIO participants’ criminal
recidivism should still be considered
preliminary. This report describes recidivism
outcomes observed over the first 2.5 years of
a program that is available to participants for
up to five years.

Other Recidivism Measures. Similar analyses were
conducted for two other measures: “any new offense,”
a composite of misdemeanor and felony recidivism,
and for violent felony recidivism. Relative to
comparison subjects, DMIO-treated subjects were
about 63 percent as likely to commit any new offense.9
The DMIO treatment effect size was -.47, larger than
that for felony recidivism.

Criminal Recidivism Rates After 2.5 Years
Significant Reductions in Recidivism. We define
recidivism, in all Institute reports, as a reconviction in
a Washington court for any offense during the followup period. We examined three categories of
recidivism: felony, any new offense (including all
felonies and misdemeanors), and violent felony
recidivism. The statistically significant differences in
criminal recidivism associated with the DMIO
program are shown in Exhibit 2.

The unadjusted rates for violent felonies were higher in
the DMIO group; but after controlling for other factors
that affect recidivism, the impact of the DMIO program
was not significantly different from zero.10

After controlling for other characteristics that may
influence recidivism, participation in the DMIO
program significantly reduced the likelihood of felony
recidivism (22 versus 40 percent) and recidivism for
any new offense (40 versus 64 percent).7

Exhibit 3

1.5- and 2.5-Year Criminal Recidivism
Unadjusted Rates for DMIO and Comparison Groups
Recidivism
Type

Exhibit 2

Reductions in Recidivism Rates
Attributable to the DMIO Program*
(2.5-year follow-up, statistically adjusted)
75%

Felony
DMIO

Any Offense

(Statistically Adjusted)

N

DMIO
Rate

Effect
Size

1.5-year follow-up

Non-DMIO

97

34%

15

15%

-.34*

152

53%

31

31%

-.41*

14

5%

10

10%

NS

122

43%

22

22%

-.39*

185

65%

40

40%

-.47*

25

9%

13

13%

NS

Violent Felony

64%

50%

2.5-year follow-up
40%

Felony

40%

25%

Any Offense
22%

Violent Felony

*Statistically adjusted effect sizes significant at p<.05

0%
*p<.05

Comparison
N
Rate

Felony

Any New Offense

Program Effects Sustained Over Time. The
impact of DMIO participation on recidivism has not
weakened since the 2005 evaluation. At the 1.5-year
follow-up, 15 percent of DMIO participants had
committed new felonies since release, compared
with 34 percent in the comparison group (Exhibit 3).11
The statistically-adjusted effect size associated with
DMIO participation at that time was -.34 for new
felonies. A year later, the felony recidivism rates had
increased to 22 and 43 percent, respectively, with an
adjusted effect size of DMIO participation at -39.
The adjusted effect size of DMIO on “any new
offenses” followed a pattern similar to felony
recidivism: -.41 at 1.5 years and -.47 at 2.5 years.
The impact of DMIO on violent felony recidivism was
not statistically significant in either follow-up period.

Felony Recidivism. After controlling for the preexisting differences listed in Exhibit 1—many of
which are well-known recidivism risk factors—the
DMIO program was found to substantially reduce
felony recidivism relative to the comparison group.8
After adjusting for these pre-existing differences,
DMIO participants were about 55 percent as likely to
be convicted of a new felony as the comparison
group. That is, the comparison subjects were about
1.8 times more likely to be reconvicted of a felony.
The DMIO program had an effect size of -.39 in
reducing the likelihood of a future felony conviction.
7

Here, the actual recidivism rates of DMIO participants are
compared to “statistically adjusted” rates obtained by calculating
odds ratios based on the logistic regression coefficients
associated with DMIO participation and applying them to the
actual DMIO recidivism rates.
8
Beta=-.85; p=.016; ROC=.82.

9

Beta=-.94; p=.003; ROC=.82.
p=.22
11
Because they are unadjusted, these rates are not equal to
those in Exhibit 2 for the comparison group.
10

3

Violent Felony Offenses in Detail. Our analysis
did not detect a statistically significant link
between DMIO participation and violent felonies.
This finding may be due to the lack of statistical
power resulting from the low number of offenses
committed by the study groups. Future analyses
with larger samples and longer follow-up will
increase the confidence and precision of this
analysis. The DMIO program was created in part
to reduce serious violent crime. Therefore, we
display new violent felonies in detail (Exhibit 4).
New violent felonies were relatively infrequent for
both the comparison and DMIO groups, and very
few DMIO offenses are classified as “most serious
violent felonies” as defined in RCW 9.9A.030. At
this time, however, these data do not provide a
statistically meaningful comparison of the DMIO
and comparison groups regarding violent felonies.

Program Costs and Recidivism Savings
Benefit-Cost Analysis. The Institute has developed
methods of economic analysis to assess program
benefits in terms of reduced costs to taxpayers for law
enforcement, adjudication, and corrections, and for the
victims of crime. To calculate benefits, the reductions
in recidivism attributable to the DMIO program are
applied to the life-time distribution of criminal offenses
expected from those released from prison. Per-person
program costs were estimated based on a review of
provider billing records.
Program Costs. The state compensates Regional
Support Networks (RSNs) and other providers who
contract with the Department of Social and Health
Services (DSHS) to provide additional support
services for DMIO program participants. The
program funds up to $10,000 per DMIO participant
per year, for a maximum of five years. The specific
funding formula established by DSHS-Mental Health
Division is as follows:

Exhibit 4

Violent Felony Recidivism in the Comparison and
DMIO Groups: 2.5-Year Follow-up
Comparison
N=287

DMIO
N=100

Total

1

0

1

1

0

1

0

1

1

1

0

1

5

0

5

0

1

1

1

0

1

2

0

2

5

5

10

1

0

1

2

0

2

1

1

2

Harassment

1

4

5

Firearm
Possession

3

1

4

Unspecified

1

0

1

Total Violent
Felonies

25

13

38

Violent Felony
Murder
in the 1st Degree†
Rape of a Child
in the 1st Degree†
Rape
in the 2nd Degree†
Robbery
in the 1st Degree†
Robbery
in the 2nd Degree†
Extortion
in the 1st Degree†
Assault
in the 1st Degree†
Assault
in the 2nd Degree
Assault
in the 3rd Degree
Vehicular
Homicide
Custodial
Assault
Violation of Protection
Order

†

y

Providers of special services during the
three months just before and just after prison
release are reimbursed $6,000 to engage
the participant.

y

After the first three months, providers are
reimbursed $700 per month for special
DMIO service for Medicaid-eligible
participants and $900 per month for nonMedicaid-eligible participants.

Per-person program costs over the 2.5 year followup period are estimated at $24,100 per participant
(in 2006 dollars). This estimate is based on a
detailed review of billing records for agencies serving
131 DMIO participants released between July 1,
2002, and December 30, 2003.12 The appendix
provides more detail on costs.
Cost Savings of Reduced Recidivism. Does the
value of the reduction in crime attributed to the
DMIO program outweigh the costs? To answer this
question, we turned to the Institute’s benefit-cost
model.13 When there are fewer crimes, there are
fewer victims and taxpayers spend less on the
criminal justice system. We estimate the present
value of crime-related costs avoided over the
lifetime of a participant for both taxpayers and crime
12
Lovell (in press). Detailed cost data for the 100 DMIO
participants used in the recidivism analysis are unavailable.
13
S. Aos, R. Lieb, J. Mayfield, M. Miller, A. Pennucci. (2004).
Benefits and costs of prevention and early intervention
programs for youth. Olympia: Washington State Institute for
Public Policy.

Most violent felonies according to RCW 9.9A.030.
4

Exhibit 6

victims. To determine the economic “bottom line” of
the program, we subtract the cost of the DMIO
program from the present-value sum of its benefits
(including avoided costs).

Net Benefits of Research-Based Offender Programs

When research is based on a less-than-randomized
research design, we know the results have a larger
margin of error than a randomized design. Since
random assignment was not possible for this study,
we reduced the estimated effect on recidivism by 50
percent when calculating cost savings.14 That is,
since we cannot control for selection bias that may
result in an overestimation of the effectiveness of
the program, we apply a 50 percent discount factor
to the program effect when we perform our benefitcost analysis.

We estimate that the DMIO program costs about
$24,280 per participant over the first 2.5 years
post release and produces about $25,100 in
crime-reduction benefits (Exhibit 5). Of these total
benefits, $11,450 accrues to taxpayers in the form
of reduced criminal justice system expenditures;
another $13,650 accrues to society because there
are fewer crime victims. The result is an overall
return to society of $820 or $1.03 per dollar spent
on a DMIO participant.

$11,450

Costs (over 2.5 years)

$23,500

$24,280

Benefit/Cost Ratio

$1.03

$0.47

Net Benefits

$820

-$12,830

$13,738

Treatment-oriented intensive supervision

$11,563

Cognitive-behavioral therapy

$10,669

Drug treatment in community

$10,299

Correctional industries in prison

$10,054

Drug treatment in prison

$9,439

Adult basic education in prison

$7,836

Adult drug courts

$4,767

Employment and job training in the community

$4,359

Electronic monitoring in lieu of jail time

$870

DMIO program

$820

Sex offender treatment in prison with aftercare

$-3,258

Surveillance-oriented intensive supervision

$-3,747

The reductions in DMIO criminal recidivism found
during the first 1.5 years after prison release hold up
at the 2.5 year mark. Participation in DMIO is
associated with statistically significant decreases in
felony recidivism and in recidivism for combined
felony or misdemeanor offenses. The analysis was
unable to identify statistically significant effects on
violent felony recidivism. A benefit-cost analysis
indicates that the reduction in criminal recidivism
attributed to the DMIO program provides a net
economic benefit to crime victims and taxpayers,
although, net benefits are small compared with other
offender programs.

Taxpayers
Only

$25,100

Vocational education in prison

Conclusion

DMIO Program Benefits and Costs

Benefits (over life-time)

Net
Benefits

Adapted from Aos et al. (2006).

Exhibit 5

Taxpayers
and Victims

Adult Offender Program

The Institute has measured the cost effectiveness of
a number of other programs that target adult
offenders.15 The net benefits (total benefits minus
costs) attributed to each of these programs are
described in Exhibit 6. While the benefits of the DMIO
program exceed the costs, according to our analysis,
it does not rank highly in cost effectiveness compared
with other programs. These other programs,
however, may not necessarily be effective for
individuals eligible for the DMIO program.

The pre-existing differences between the DMIO and
comparison groups, such as recidivism risk and
gender, remain a potential weakness of this study.
Future analyses will include a larger sample of
DMIO participants and a comparison group that
more closely resembles the DMIO population. We
are also unable to account for a potentially important
difference between DMIO and comparison groups,
willingness to participate. Currently, we account for
this selection bias by discounting our findings by 50
percent. Future analyses, based on a substantially
larger sample of DMIO participants, will provide an
opportunity to address the issue of selection bias
more effectively.

14

The rationale for this discount is explained in Aos (2004).
S. Aos, M. Miller, & E. Drake. (2006). Evidence-based public
policy options to reduce future prison construction, criminal
justice costs, and crime rates. Olympia: Washington State
Institute for Public Policy.

15

5

Appendix: DMIO Program Costs

For further information, contact Jim Mayfield at
(360) 586-2783 or mayfield@wsipp.wa.gov

Cost estimates are based on a cohort of 116 DMIO
participants released from prison between July 1,
2002, and December 30, 2003. The actual billings
attributed to these clients were examined over a
two-year period and are reported in Exhibit A1.16
Administrative costs are based on average
enrollments of 165 during 2005, a period during
which program participation stabilized.
Exhibit A1

Total and Per-Capita Costs for
Two Years After Prison Release (2004 Dollars)
Per Capita Costs

Period

Total
Billings
(N=116)

Months
1-6(1)

$750,430

$6,580

$680

$7,260

Months
7-12

$419,250

$3,680

$680

$4,360

Months
13-18

$384,860

$3,380

$680

$4,060

Months
19-24

$345,090

$3,030

$680

$3,700

2-Year
Total

$1,899,630

$16,670

$2,720

$19,390

Months
25-30
(estimate)

--

$2,680

$680

$3,360

2.5-Year
Total
(estimate)

--

$19,350

$3,400

Billing

Admin(2)

Total

$22,740(3)

Notes: (1) includes transition costs for pre- and post-release planning,
engagement, and services; (2) based on stable program enrollments
of 165; (3) For the benefit-cost calculations in this report, 2.5 year
expenditures were estimated by extrapolating an additional six
months at per-person costs expected during the 25- to 30-month
period. Costs were then expressed in 2006 dollars ($24,280).

16

More details on these costs estimates are provided in Lovell
(in press).

Document No. 07-01-1902
Washington State
Institute for
Public Policy
The Washington State Legislature created the Washington State Institute for Public Policy in 1983. A Board of Directors—representing the legislature,
the governor, and public universities—governs the Institute and guides the development of all activities. The Institute’s mission is to carry out practical
research, at legislative direction, on issues of importance to Washington State.