Program evaluation, adult probation programs

              State of Arizona
Office
of the
Auditor General
PERFORMANCE AUDIT
Report to the Arizona Legislature
By Douglas R. Norton
Auditor General
ADULT PROBATION
PROGRAMS
March 1999
Report Number 99-4
2910 NORTH 44th STREET • SUITE 410 • PHOENIX, ARIZONA 85018 • (602) 553-0333 • FAX (602) 553-0051
DOUGLAS R. NORTON, CPA
AUDITOR GENERAL
DEBRA K. DAVENPORT, CPA
DEPUTY AUDITOR GENERAL
STATE OF ARIZONA
OFFICE OF THE
AUDITOR GENERAL
March 17, 1999
Members of the Arizona Legislature
The Honorable Jane Dee Hull, Governor
Mr. David K. Byers, Administrative Director
Administrative Office of the Courts
Supreme Court
Transmitted herewith is a report of the Auditor General, An Evaluation of the Adult
Probation Programs administered by the Supreme Court’s Administrative Office of the
Courts, Adult Services Division. This evaluation was conducted in response to a May
27, 1997, resolution of the Joint Legislative Audit Committee.
The report describes factors that contribute to adults successfully completing probation.
Of particular note was the fact that substance abusers who consistently attended or
successfully completed substance abuse treatment were much more likely to
successfully complete probation than was the general sample of probationers we
evaluated. About 85 percent of the individuals who completed drug treatment
succeeded in completing the terms of probation. And, 80 percent of those who
consistently attended alcohol abuse programs successfully completed probation. In
view of the known relationship between substance abuse and crime, plus Arizona’s
emphasis on drug treatment, the next step is to ensure that effective treatment is
available to all who need it, and to emphasize attendance in and completion of
substance abuse treatment programs. Since consistent employment and completion of
community service also predict successful probation outcomes, probation departments
should continue to emphasize these conditions of probation.
The evaluation identified differences in the ways counties manage probationers. Some
counties emphasize restitution and rehabilitation and therefore are more tolerant of
some forms of noncompliant behavior. Other counties appear more concerned with
community protection, and are quicker to revoke misbehaving probationers. Follow-up
evaluations could determine the relative costs and benefits of these approaches, but
improved recordkeeping is needed to facilitate further studies. Better standards and
guidelines for
March 17, 1999
Page -2
information collection, storage, and retrieval of both paper case files and electronic records,
and development of a statewide electronic database, is required to expedite future research
and evaluation. Such action is also important for decision-making, case management, and
protecting victims.
As outlined in its response, the Administrative Office of the Courts agrees with most of the
findings and recommendations presented in the evaluation. However, the Administrative
Office of the Courts does not agree to work with County Probation Departments to deter-mine
the long-term cost-effectiveness of tolerating different degrees of criminal and non-compliant
behaviors, while departments provide community-based rehabilitation. The
Courts indicate that implementing this recommendation would be too costly. The Courts
also disagree with the recommendation to develop specialized caseload assignments in
smaller counties for probationers with mental health problems. It was suggested that a dif-ferent
method would be used to address this concern.
My staff and I will be pleased to discuss or clarify items in the report.
This report will be released to the public on March 18, 1999.
Sincerely,
Douglas R. Norton
Auditor General
Enclosure
i
SUMMARY
The Office of the Auditor General has conducted an evaluation of the adult probation pro-grams
administered by the Supreme Court’s Administrative Office of the Courts, Adult
Services Division. This evaluation was conducted in response to a May 27, 1997, resolution
of the Joint Legislative Audit Committee. The evaluation focused on identifying factors that
contribute to adults successfully completing probation and examines the effect of substance
abuse treatment programs, differing probation management approaches counties have
adopted, and outcomes associated with gender and other demographic characteristics.
Under the Criminal Code, probation is a sentencing alternative for many crimes in Arizona.
Most typically, it involves a convicted felon being placed under supervision for three to
seven years for crimes ranging from theft, drug, or drunk driving charges to armed robbery
and sex offenses. Offenders who are sentenced to probation are required to abide by stan-dard
conditions such as remaining law abiding, submitting to searchers, and possessing no
weapons. Many sentences, in addition to probation, include some time in jail or prison.
Courts may also impose financial obligations on probationers, such as fees for probation
services and supervision, or restitution payments to a victim for economic loss, such as
medical expenses.
Probation is the most common way that offenders in Arizona’s criminal justice system serve
their sentences. At any given time in 1998, Arizona county probation departments were su-pervising
approximately 35,000 probationers. Given the recent dramatic rise in drug-related
arrests, many of these probationers have substance abuse problems. The current trend in the
Arizona criminal justice system is to emphasize substance abuse treatment rather than pun-ishment.
Since first- and second-time drug possession offenders can receive treatment dur-ing
probation, it is necessary to determine the degree to which drug treatments are success-ful.
Factors Associated with
Successful Completion
of Probation
(See pages 11 through 17)
This evaluation identified three factors that are associated with successfully completing
probation. These factors are probationers’ success in completing substance abuse treatment
programs, maintaining employment, and completing community service requirements.
Although the overall successful probation completion rate for this study was 63 percent,
probationers who consistently participated in drug abuse counseling had successful proba-tion
completion rates of 85 percent. Similarly, substance abusers who frequently attended
12-step drug or alcohol programs such as Alcoholics Anonymous had probation success
ii
rates of 80 percent. These probation completion rates are not only higher than those for pro-bationers
not successfully completing their treatment programs, but are also higher than
those for probationers never required to attend such programs.
Because completion of substance abuse treatment programs is associated with higher pro-bation
success rates, efforts to initially screen probationers for treatment needs and to em-phasize
that they attend substance abuse treatment programs should be stressed. Many
probationers were not identified as substance abusers and/or ordered to attended treatment
at the time they were placed on probation. Approximately 17 percent of the probationers not
initially ordered to attend treatment programs were later identified as having problems and
ordered to do so. In addition, approximately 30 percent of those ordered to attend treatment
never did. This group was more likely to have their probation revoked than those not or-dered
to treatment and those who completed treatment. Although in some cases this may
have been due to a lack of funding and availability of such programs, more treatment
should be available in the future through the Drug Treatment Education Fund established
in 1996 and from which monies became available in fiscal year 1998.
Probationers who were consistently employed during probation had a 90 percent probation
completion rate. Probationers with consistent employment also had significantly fewer sub-sequent
arrests and positive drug tests during probation than those who were only sporadi-cally
employed. In addition, probationers who completed their community service require-ments
had probation success rates of 85 percent. These probationers also had fewer positive
drug tests and tended to have fewer subsequent arrests than did those who did not com-plete
community service. Therefore, supervision should continue to emphasize consistent
employment and completion of community service
Counties’ Degree of Tolerance for Criminal
and Noncompliant Behaviors
Affects Probation Outcomes
(See pages 19 through 26)
Probation departments have a dual responsibility to provide probationers with opportuni-ties
for rehabilitation while at the same time protecting the community from crime. In Ari-zona
each county sets its own policies for balancing these two responsibilities and each var-ies
in the degree to which it tolerates criminal and noncompliant behavior in an effort to
provide community-based rehabilitation. The four counties in the study vary in the extent to
which they emphasize rehabilitation or community protection.
Variations in different levels of tolerance for criminal and noncompliant behaviors are re-flected
in various outcomes of the probation process. For example, Pima County revoked
more probationers to jail or prison than did Maricopa. In addition, Pima County revoked its
probationers more quickly than did Maricopa County, which tended to reinstate probation-ers
with additional conditions or more stringent levels of supervision. While arrest rates for
iii
probationers who were subsequently revoked are similar across the four counties, the extra
time Maricopa County probationers spend in the community before eventually having their
probation revoked does result in more crimes committed by these individuals.
In Maricopa individuals completing probation were more likely to pay full restitution to
crime victims, do their community service, and complete probation early. However, it is not
clear how the costs and benefits associated with probation departments’ varying levels of
tolerance for criminal and noncompliant behaviors compare to one another. A follow-up on
early probation completers versus probationers who serve their full terms, and a more in-depth
analysis of the behaviors of probationers who are eventually revoked, would help to
determine the costs and benefits associated with different levels of tolerance.
Probationer Success Rates
Vary by Gender and
Other Characteristics
(See pages 27 through 32)
Probation outcomes differ by gender. As a group, women are somewhat more likely to
complete their probation, and, even if they violate their conditions of probation, they are less
likely to have their probation revoked. However, while education, age, and income are all
positively correlated with probation success for men, these relationships are different for
female probationers.
Minorities are generally more likely to have their probation revoked and are less likely to
complete probation early. However, this is generally accounted for by the fact that African-
Americans whose probations were revoked had higher arrest rates and initially committed
more serious types of crimes than non-minorities who had their probations revoked.
In addition, appropriate treatment is important for the probation success of offenders with
mental health problems or mental disabilities. Probationers whose mental health problems
are identified at the time of sentencing and who receive appropriate treatment do just as
well in the probation system as the general population. However, probationers with mental
health problems that are not identified and addressed in their conditions of probation do not
fare well.
Finally, for the small group of probationers who abscond (flee and whose whereabouts are
unknown), only two characteristics differentiate them from the population as a whole. They
are more likely to fall into the 26-to-30-year-old age group and have an average lower edu-cation
level than other probationers.
iv
Recordkeeping Needs
to Improve
(See pages 33 through 36)
Accurate and complete documentation in probation files is important for decision-making,
case management, protecting victims, and research and evaluation. However, basic man-agement
information collected by county probation departments is difficult to retrieve and
use in a timely manner. Specifically, case files often were missing information, or informa-tion
was conflicting, disorganized, or illegible. In addition, electronic files are unreliable. The
Adult Services Division of the Administrative Office of the Courts should work with the
county probation departments to develop standards and guidelines for information collec-tion,
storage, and retrieval of both paper case files and electronic records, and to develop a
statewide, standardized electronic database.
v
Table of Contents
Page
Introduction and Background........................................................... 1
Finding I: Factors Associated
with Successful Completion
of Probation.................................................................................... 11
Substance Abuse
and Criminal Behavior .................................................................................................... 11
Completing Treatment Programs
Is Associated with Successful
Completion of Probation................................................................................................. 12
Consistent Employment and Community
Service Are Associated with Successful
Completion of Probation................................................................................................. 14
Repeated Arrests and Petitions
to Revoke Were Linked
to Probation Failure.......................................................................................................... 16
Greater Use of Treatment
Programs Could Further
Improve Probation Success............................................................................................. 16
Recommendations............................................................................................................ 17
Finding II: Counties’ Degree of Tolerance for
Criminal and Noncompliant Behaviors
Affects Probation Outcomes........................................................ 19
Counties Have Wide Latitude in
Setting Probation Policies................................................................................................ 19
Tolerance for Relapse to Criminal
and Noncompliant Behaviors
Differs Among Counties Studied.................................................................................. 20
Approaches Produce Varying
Probation Outcomes......................................................................................................... 22
vi
Table of Contents (cont’d)
Page
Finding II: (cont’d)
Approaches Produce Varying
Probation Outcomes......................................................................................................... 22
Additional Study
Necessary........................................................................................................................... 25
Recommendation.............................................................................................................. 26
Finding III: Probationer Success Rates
Vary by Gender and Other
Characteristics............................................................................... 27
Background....................................................................................................................... 27
Comparisons Between Men
and Women Show Substantial
Differences in Probation Outcomes.............................................................................. 28
Differences in Revocation
and Early Completion Found
for Different Ethnic Groups............................................................................................ 29
Without Appropriate Treatment,
Probationers with Mental Health
Problems or Mental Disabilities
Are Less Likely to Succeed............................................................................................. 31
Absconders Differ from
Other Probationers in
Education and Age........................................................................................................... 32
Conclusions....................................................................................................................... 32
Recommendation.............................................................................................................. 32
vii
Table of Contents (cont’d)
Page
Finding IV: Recordkeeping
Needs to Improve .......................................................................... 33
Good Recordkeeping Critical
to Probation Management .............................................................................................. 33
Many Recordkeeping
Problems Surfaced While
Auditors Assembled Data
for This Evaluation........................................................................................................... 34
Improvements in Case File
Accuracy and Electronic Formats
Could Improve Accountability...................................................................................... 35
Recommendations............................................................................................................ 36
Agency Response
Appendix A.......................................................................................... a-i
Appendix B.......................................................................................... b-i
References .......................................................................................... r-i
viii
Table of Contents (cont’d)
Page
Tables
Table 1 Adult Probation Outcomes Study
Statewide Adult Probation Programs
Schedule of Expenditures by Funding Source
Years Ended or Ending June 30, 1996, 1997, 1998, and 1999
(Unaudited).............................................................................................. 6
Table 2 Adult Probation Outcomes Study
County Adult Probation Departments
Populations Served in Standard, Interstate Compact,
and Intensive Probation Programs at February 28, 1998,
and Departments’ Staffing Levels at December 31, 1997
(Unaudited).............................................................................................. 7
Table 3 Adult Probation Outcomes Study
Percentage of Probationers Who Paid All Fines, Fees,
and Restitution and Performed All Community Service
As of September 1, 1998......................................................................... 23
Table 4 Adult Probation Outcomes Study
Revoked Probationers Comparison of Average Arrests
During Probation, and Average Number of Months Before
Revocation, by County as of September 1, 1998............................... 25
Table 5 Adult Probation Outcomes Study
Probation Outcomes for Men and Women
by General Education Level in Percentages
as of September 1, 1998.......................................................................... 29
Table 6 Adult Probation Outcomes Study
Summary of Sample Selection............................................................. a-ii
Table 7 Adult Probation Outcomes Study
Sample of Probationers by County
Percentage Assigned to Intensive Probation
Supervision and and Percentage Identified as Male
and Belonging to an Ethnic Minority................................................. a-iv
ix
Table of Contents (concl’d)
Page
Tables (concl’d)
Table 8 Adult Probation Outcomes Study
Sample of Probationers by County
Type of Crime Causing the Probation by Percentage..................... a-iv
Table 9 Adult Probation Outcomes Study
Sample of Probationers by County
Percentage of Sample with Prior Supervisions, Custodies,
Misdemeanors, and Felonies and Average Length of
Probation Sentence in Months............................................................. a-v
Figures
Figure A Adult Probation Outcomes Study
Percentage of Successful Probations
and Result of Final Drug Treatment
As of September 1, 1998......................................................................... 13
Figure B Adult Probation Outcomes Study
Consistency of Employment and
Percentage of Successful Probations
As of September 1, 1998......................................................................... 14
Figure C Adult Probation Outcomes Study
Percentage of Adult Probationers
Successfully Completing Probation
and Community Service Participation
As of September 1, 1998......................................................................... 15
Figure D Adult Probation Outcomes Study
Arrests During Probation by Type of Crime
Shown as a Percentage of All Arrests Recorded
As of September 1, 1998......................................................................... 24
Figure E Adult Probation Outcomes Study
Percentage of Probationers Who Successfully
Completed Their Probation by Gender and Age
As of September 1, 1998......................................................................... 30
x
(This Page Intentionally Left Blank)
1
INTRODUCTION AND BACKGROUND
The Office of the Auditor General has conducted an evaluation of the adult probation pro-grams
administered by the Supreme Court’s Administrative Office of the Courts, Adult
Services Division. This evaluation was conducted in response to a May 27, 1997, resolution
of the Joint Legislative Audit Committee. The evaluation focused on identifying factors that
contribute to probationers successfully completing their probation. It examines such issues
as the effect of substance abuse treatment programs, differing probation management ap-proaches
counties have adopted, and outcomes associated with gender and other demo-graphic
characteristics.
Drug and Alcohol Abuse Is Impacting
the Criminal Justice System
From 1980 to 1996, the number of people in the criminal justice system tripled, due primar-ily
to criminal activity linked to drug and alcohol abuse. Recent dramatic increases in ar-rests,
convictions, and incarcerations for drug-related crimes have been accompanied by
prison and jail overcrowding.
While Arizona's overall increase in the crime rate reflects the State’s population growth, the
number of alcohol- and drug-related offenses has increased more rapidly. Department of
Public Safety Uniform Crime Reports from 1990 to 1996 show an 80 percent increase in ar-rests
for illicit substance abuse, drug trafficking, and drug sales. During the same period,
however, the increase in the total number of all crimes was much more reflective of the
overall increase in population, and grew only slightly faster than the population.
The dramatic rise in drug-related crime imposes a heavy cost to taxpayers, businesses,
neighborhoods, and families. It is estimated that in Arizona, approximately 30 percent of
domestic violence incidents and nearly 80 percent of child abuse and neglect cases are at-tributed
to substance abuse. Law enforcement officials attribute as much as 80 percent of
theft and credit card fraud cases to substance abuse. While social costs are difficult to calcu-late,
the federal government spends billions of dollars on drug prevention treatment, drug
interdiction, and related child support and law enforcement. Law enforcement, prisons and
jails, courts, probation, and substance abuse treatment demand increasing public resources.
We estimate the combined taxpayer burden in Arizona to be approximately $105 million a
year: about $65 million in county and state dollars, and approximately $40 million in federal
monies.
The current trend in the Arizona Criminal Justice System is to emphasize substance abuse
treatment rather than punishment. Since first- and second-time drug possession offenders
can receive treatment during probation, it is necessary to determine the degree to which
2
drug treatments are successful. For example, to what degree can participation in substance
abuse treatment lead to more successful probation; does treatment substantially reduce sub-stance
abuse; and does successful substance abuse treatment lead to lower crime rates over
time? In addition, the recently created Drug Treatment and Education Fund will afford
more probationers the opportunity to receive substance abuse treatment.
Probation in the
Criminal Justice System
Under the Criminal Code, probation is a sentencing alternative for many crimes in Arizona.
Most typically, it involves a convicted felon being placed under supervision for three to
seven years for crimes ranging from theft, drug, or drunk driving charges to armed robbery
and sex offenses. Many sentences, in addition to probation, include some time in jail or
prison. Courts may also impose financial obligations on probationers, such as fees for pro-bation
services and supervision, or restitution payments to victims for economic loss, such
as medical expenses.
Offenders who are sentenced to probation are required to abide by standard conditions such
as remaining law abiding, submitting to searches, and possessing no weapons. Specialized
requirements may also be imposed, such as having no contact with a crime victim, or par-ticipating
in vocational or educational training or mandatory substance abuse treatment.
Failure to adhere to the conditions of probation may result in more stringent conditions be-ing
imposed, a lengthening of the probationary period, or in revocation of probation, where
the offender is sent to jail or prison to complete the sentence. Probation officers monitor of-fenders
to determine whether they are complying with the terms of their probation. They
can make recommendations ranging from discharging offenders from probation ahead of
schedule to revoking probation and sending the offender to prison or jail.
Probation is the most common way that offenders in Arizona’s criminal justice system serve
their sentences. Over the last several decades, more than 60 percent of all adults under
criminal justice supervision have served required sentences through probation in the com-munity.
At any given time in 1998, Arizona probation departments were actually supervis-ing
approximately 35,000 probationers. As of February 28, 1998, there were 33,513 active
probationers. The number increased to 35,694 by November 30, 1998.
Together with parole, probation constitutes the community corrections portion of the crimi-nal
justice system. Probation and parole are believed to provide better opportunities than
incarceration for rehabilitating offenders. Rehabilitation, in turn, is believed to lower the
chance of criminal recidivism (relapse into criminal behavior). Although there are efforts to
provide rehabilitation while inmates are incarcerated, former prison inmates have relatively
high recidivism rates.
3
Probation Must Address
Multiple Goals Affecting
Both Offenders and Victims
The probation system involves a number of different participants, such as offenders, vic-tims,
courts, county probation offices, and other law enforcement personnel, and involves
the difficult task of meeting multiple goals that range from protecting the victim to rehabili-tating
the offender. The six major goals of community corrections, as identified by the
American Probation and Parole Association, are listed below, together with a discussion of
the role probation departments and others play in meeting those goals.
n Enforcing sanctions ordered by the court—County probation departments are given
the responsibility to enforce all sanctions or conditions ordered by the courts. More of-fenders
with higher levels of risk are being placed on probation, often with more strin-gent
conditions, through programs such as the Community Punishment Program (de-scribed
further in the next section on pages 4 through 5), which range somewhere be-tween
incarceration and standard probation. Enforcement requires facilitation skills, on-going
monitoring, and timely response, both to signs of progress and incidents of non-compliance.
n Assisting offenders—While many offenders may genuinely have a motivation to
change, they are unlikely to do so unless they receive help in altering their behavior or in
learning new skills and forming different attitudes. To help offenders change from a
criminal lifestyle to a socially acceptable and productive one, their attitudes toward em-ployment,
peers, authority, and substance abuse must be assessed and addressed. Pro-bation
officers need to specify appropriate behaviors and to provide offenders with ac-cess
to programs that help them develop new skills, and learn appropriate coping skills.
n Protecting the community—Protecting the community has always been a primary ob-jective
of the criminal justice system. Probation officers must provide surveillance to
monitor an offender’s activities and social environment. Surveillance efforts include
home visits, contacts with employers, neighborhood contacts, and electronic monitoring.
n Assisting decision-makers—Probation departments have a responsibility to provide
decision-makers with complete and accurate information regarding offenders’ back-grounds
and their current risks and needs. Probation officers provide this information
by conducting presentence investigations and by preparing reports that are used during
alterations in the conditions of probation (such as revocation) as well as in determining if
an offender should be released from probation.1
1 Probation officers may also begin the revocation process, which ends probation by having the offender com-plete
his or her term in jail or prison.
4
n Supporting victims—Probation departments have a responsibility to support the rights
of crime victims and to provide benefits to all those involved. To meet this responsibility,
probation officers can assist victims through advising them of the offender’s custody
status, notifying them of arrangements for payment of restitution, informing them of su-pervision
conditions, and notifying them if there is potential danger.
n Coordinating services—Finally, probation departments have a responsibility to coor-dinate
services for probationers. Offenders under the supervision of probation depart-ments
require a variety of services throughout their rehabilitative process. While many
of these services are increasingly provided in-house, probation departments are still
largely dependent on other community service providers to meet these needs.
Probation Is Locally Administered
but Primarily State-Funded
With the adoption of A.R.S. §12-251 in 1927, the probation system in Arizona was estab-lished
as part of the superior court system in each county. Each county’s presiding judge is
responsible for the county probation department. Specifically, A.R.S. §12-251(A) designates
the presiding judge in each county to appoint a chief probation officer to manage the proba-tion
department.
Probation departments were initially funded exclusively by counties. The first significant
state funding occurred during 1985, and since that time the State’s monetary involvement
has continued to increase. For fiscal year 1999, over $76 million was scheduled for expendi-ture
on probation services, with the State contributing approximately $50 million and the
remainder provided by the counties and monies generated from fees and other sources. The
State provides financial support in the following programs:
n Standard Probation—Standard probation is a sentencing option whereby a convicted
offender is released into the community under the supervision of a probation officer in
lieu of incarceration. Offenders may be required to seek substance abuse treatment or
participate in other programs as directed by the court or the probation officer.
n Intensive Probation—This program for high-risk offenders, as measured by a risk as-sessment
used by probation departments, began in 1986. It is more highly structured
and requires stricter conditions than standard probation, such as more frequent report-ing
to probation officers, electronic monitoring, at-home surveillance, and abstaining
from drug or alcohol use.
n Community Punishment Program—The Legislature established this program in 1988
to provide enhanced services for high-risk offenders who might otherwise fail to com-plete
probation. Offenders who fail drug offender diversion programs, fail other re-
5
quirements for an existing probation term, or are newly sentenced to probation may be
enrolled in Community Punishment Programs. Each Arizona County operates Commu-nity
Punishment Programs differently, which may include the elements of electronic
monitoring, sex offender treatment, or substance abuse treatment.
n Drug Treatment Education Fund—Voters approved Proposition 200 in November
1996. This proposition, known as the Drug Medication and Control Act of 1996, created a
program that provides expanded drug treatment and education services to first- and
second-time drug possession offenders. Each of Arizona's counties receives some of the
Drug Treatment Education Fund monies for substance abuse treatment. Although the
Drug Treatment Education Fund is primarily designed as a treatment program for drug
offenders (not including alcohol), it can serve as a funding source for all substance abuse
programs. Excess Drug Treatment Education Fund monies may be spent on other of-fenders
with substance abuse problems after all offenders eligible under Proposition 200
are served. The increased resources afforded by this fund beginning in fiscal year 1998
allow the judicial system to better address the probationers’ substance abuse issues.
n Interstate Compact—In 1996, probation departments began supervising offenders who
were under supervision in other states but move to Arizona (this service was previously
provided by the Department of Corrections). While the funding for most counties’ Inter-state
Compact probationers is absorbed through the standard probation monies, four
counties (Maricopa, Mohave, Pima, and Yavapai) have enough probationers to require
at least one additional probation officer. As a result, the Legislature appropriated slightly
more than $1 million in fiscal year 1998 for these counties.
Table 1 (see page 6) provides an overview of statewide expenditures for adult probation
programs for fiscal years 1995 through 1998. As the table shows, most of the $45.9 million in
state aid in fiscal year 1998 went to standard and intensive probation.
6
Table 1
Adult Probation Outcomes Study
Statewide Adult Probation Programs
Schedule of Expenditures by Funding Source
Years Ended or Ending June 30, 1996, 1997, 1998, and 1999
(Unaudited)
Source/Program
1996
(Actual)
1997
(Actual)
1998
(Actual)
1999
(Estimated)
State monies:1
General Fund appropriations:
Adult Standard Probation $17,492,273 $18,468,955 $20,473,797 $22,163,585
Adult Intensive Probation 16,156,266 17,090,678 18,100,614 19,129,421
Community Punishment 3,440,737 3,447,883 4,115,147 4,182,932
Interstate Compact 776,433 1,051,563 1,120,011 1,313,900
Luxury privilege tax:
Drug Treatment and Education                  2,740   3,167,683 3,655,802
Total state expenditures   37,865,709 40,061,819 46,977,252 50,445,640
Other monies:
County general fund2 15,726,591 16,735,435 15,838,983 15,054,922
Probation service fees 4,562,026 4,971,753 6,897,990 10,913,031
Total expenditures $58,154,326 $61,769,007 $69,714,225 $76,413,593
1 Includes amounts disbursed directly to the counties, amounts paid by the State on behalf of the counties for
expenditures such as motorpool charges and vehicle purchases, and amounts paid by the State for admin-istrative
costs and centralized services such as research, automation, and certification.
2 For fiscal years 1996 and 1997, includes estimates for Gila, Graham, Greenlee, La Paz, Mohave, and Santa
Cruz Counties because they have combined juvenile and adult probation departments and do not sepa-rately
account for adult probation expenditures. For fiscal year 1998, this information was not accumulated
for the counties. Amount shown is an average of expenditures in fiscal years 1996 and 1997 and the amount
budgeted for fiscal year 1999.
Source: Auditor General staff analysis of the Administrative Office of the Courts’ financial reports, county
finance department reports, and county adult probation department reports.
7
As of February 1998, about 87 percent of offenders on probation in Arizona were in stan-dard
probation. Table 2 shows the distribution of probationers by county. Probationers in
the Community Punishment Program are included under standard probation. Probationers
receiving services through the Drug Treatment Education Fund are not accounted for sepa-rately.
Table 2 also shows staffing levels for each county’s probation department as of De-cember
1997. In all, the counties had 1,638.5 positions.
Table 2
Adult Probation Outcomes Study
County Adult Probation Departments
Populations Served in Standard, Interstate Compact,
and Intensive Probation Programs at February 28, 1998,
and Departments’ Staffing Levels at December 31, 1997
(Unaudited)
Probation Population
County Standard
Interstate
Compact Intensive Total
Total
Staff
Maricopa 18,762 762 1,518 21,042 898.0
Pima 3,481 238 608 4,327 284.8
Mohave 1,202 114 66 1,382 51.5
Yavapai 1,017 43 161 1,221 69.5
Pinal 894 31 83 1008 53.0
Yuma 722 32 214 968 68.5
Coconino 771 32 118 921 50.5
Gila 551 11 33 595 24.4
Navajo 555 16 44 615 26.5
Cochise 406 29 136 571 50.0
Santa Cruz 129 11 34 174 20.0
Apache 194 19 40 253 20.0
Graham 183 2 14 199 8.0
La Paz 116 10 5 131 6.3
Greenlee 95 2 9   106 7.4
Total 29,078 1,352 3,083 33,513 1,638.4
Source: Adult Services Division, February 1998 Monthly Statistical Report and 1998 fiscal year Midyear Closing
Report (as of December 31, 1997).
Audit Scope and
Methodology
This evaluation focused on identifying factors that contribute to probationers’ successful
completion of probation, with particular attention to substance abuse treatment, and to a
8
lesser extent, how counties vary in their management of probationers. The focus on sub-stance
abuse reflects the profound effect of drug-related crimes on the criminal justice sys-tem
in recent years.
The findings in this report are based on analysis of data generated from 845 probation files
in 4 of Arizona’s 15 counties: Cochise, Maricopa, Pima, and Yavapai. Approximately 80 per-cent
of all probationers in Arizona reside in one of these four counties.1 The 845 probationers
included in the analysis all began their probation terms in 1994. The study covers the period
from when each individual began his or her probation through the date the probation
ended.2 The 845 cases were drawn from the population of individuals who began probation
in 1994 in the four counties that were selected. The sample was drawn in such a way to
maximize the extent that statistically valid conclusions could be drawn. However, some
constraints in the data collection (see Appendix A, pages a-i through a-vii), limit the extent
to which conclusions regarding Cochise and Yavapai Counties can be made. As a result,
county comparisons in the report focus on Maricopa and Pima Counties.
The evaluation was specifically designed to allow an assessment of how substance abuse
treatment affected probationers. As a result, the sample included a group known to have a
substance abuse problem and to have received treatment. For Maricopa and Pima Counties,
this group included a large number of probationers who had Community Punishment Pro-gram
as a condition of their probation, because these two counties use the Community
Punishment Program to provide substance abuse treatment. However, offenders with
known substance abuse problems who were not in the Community Punishment Program
were also in the sample. The sample intentionally included substantial numbers of offenders
sentenced for driving under the influence, and who have a substance abuse problem.
The sampling design also called for a group to be included that had a known substance
abuse problem but who did not receive treatment. To serve as a comparison to the proba-tioners
who received substance abuse treatment as a condition of probation, this review in-cluded
known substance abusers who were not required to attend substance abuse treat-ment.
The sample also included probationers with no known substance abuse problem. See
Appendix A (see pages a-i through a-vii), for a more thorough description of the sampling
methods and the data collection procedures.
1 The methodology used in selecting the counties and files, and a discussion of data, are provided in Appen-dix
A (see pages a-i through a-v).
2 Some of the probationers were still on active probation at the time their files were reviewed. Their activities
through the time their files were reviewed were analyzed.
9
The following provides some key descriptions of the total sample:1
n 80 percent were male
n Almost half had not completed high school
n The average age was 31 years
n Eighty percent were not married when their probation began
n Most had committed a class 4 or 6 felony2
n More than three-fourths had no prior felonies
n Approximately 60 percent had committed at least one prior misdemeanor (the average
number of misdemeanors was 2.3)
n More than half had no prior supervision
n At the time of data collection the sample of 845 had the following probation dispositions:
Ü 49 percent had successfully completed their probation
Ü 9 percent were still on probation due to an initially longer-than-average term of pro-bation
Ü 5 percent were still on probation because their term had been lengthened or they had
an additional/concurrent probation term added
Ü 24 percent had been revoked to jail or prison
Ü 5 percent absconded
Ü 1 percent were decreased
Ü 3 percent were transferred with no known outcome, and
Ü 4 percent had dispositions that were unclear or unknown
1 Additional information on the sample by county is provided in Appendix A, Tables 7, 8, and 9 (see pages a-iv
and a-v).
2 Crimes are classified as felonies or misdemeanors. Felonies are more serious than misdemeanors. Felonies
are assigned a class of 1 to 6, with 1 the most serious. For example, homicides are felony class 1 offenses.
Robbery is a class 4 felony; sexual assault on a spouse is a class 6 felony.
10
Additional information was gathered through literature reviews; interviews with a variety
of individuals in the system, including a judge, chief probation officers, supervisors, and
treatment counselors; observations of substance abuse treatment services and 12-step pro-grams
for substance abuse; and review of county probation office documents regarding
processes and procedures.
Complete information on each of the 845 probationers was not available. As described in
Finding IV (see pages 33 through 36), some files contained conflicting or incomplete infor-mation.
In cases where there was incomplete or conflicting information, missing values are
assigned to that variable. As a result of this missing data, and the fact that not all of the 845
cases in the study had completed probation, the number of cases on which the analyses are
conducted varies, since a case with a missing value on any of the factors being analyzed is
excluded. A variety of methods, including linear and logistic regression, analysis of vari-ance,
and cross-tabulation were used to develop the findings in this report. Appendix B (see
pages b-i through b-iv) provides additional detail about the analyses that are reported in
Finding I (see pages 11 through 17), Finding II (see pages 19 through 26), and Finding III
(see pages 27 through 32).
This report presents findings in four areas regarding:
n The impact of substance abuse treatment and other factors on the successful completion
of probation;
n A comparison of probation completion rates and intermediate outcomes in the four
counties, as well as the counties’ treatment of probationers;
n Probation outcome variations among important subpopulations, such as women and
absconders (probationers who flee and whose whereabouts are unknown); and
n An assessment of probation recordkeeping.
The Auditor General and staff express appreciation to the Administrative Office of the
Courts, the Director and staff of the Adult Services Division, presiding judges, and the chief
probation officers and their staff for their cooperation and assistance throughout the audit.
11
FINDING I
FACTORS ASSOCIATED WITH
SUCCESSFUL COMPLETION OF PROBATION
The evaluation identified several factors that are associated with successful completion of
probation in Arizona. Probationers who successfully complete drug treatment programs or
actively participate in 12-step programs, such as Alcoholics Anonymous, are more likely to
successfully complete probation, less likely to be arrested while on probation, and less likely
to test positive for alcohol or drugs during probation. Additionally, probationers who re-main
consistently employed or who successfully complete required community service also
are more likely to complete probation than those who do not. Not surprisingly, the factors
most closely associated with failure to complete probation involved conditions that often
result in incarceration: being arrested during the probationary period or otherwise being the
subject of petitions to revoke probation. Arizona appears to have potential to make greater
use of treatment programs for substance abuse among those offenders who are placed on
probation.
Substance Abuse
and Criminal Behavior
The success of substance abuse programs is of particular interest to probation systems.
Many studies have found a relationship between drug and alcohol abuse and criminal ac-tivity.
For example, individuals who test positive for drugs at the time of arrest have longer
criminal records and have been imprisoned more often than those who do not. Forty-one
percent of first-time offenders have a history of regular drug use, but the proportion jumps
to 81 percent for those with five or more prior convictions. Studies have concluded that
treatment of substance abuse may reduce criminal activity and that the justice system is one
of the most important gateways to treatment delivery. With this in mind, in recent years the
national trend has been to emphasize treatment for offenders with drug problems as op-posed
to incarcerating them. Arizona voters approved the Drug Treatment Education Fund
in 1996, which created a program that provides expanded drug treatment and educational
services to first- and second-time drug possession offenders. It can also serve as a funding
source for all substance abuse programs, and excess monies may be spent on other offenders
with substance abuse problems. Increased resources afforded by the fund allow the Courts
to better address the substance abuse issues of probationers. The question is, how effective
are these programs at reducing substance abuse and crime, and do they contribute to suc-cessful
completion of probation?
12
Of the 660 probationers who had final probation outcomes in the current evaluation, 63 per-cent
completed probation successfully and 37 percent did not. If probationers successfully
completed probation, they were terminated on or before the predetermined date. Unsuc-cessful
probation involved revocations to jail or prison, or probationers absconding by
leaving the state, for example. The remaining 185 probationers in the sample were not in-cluded
in the analysis because they neither successfully nor unsuccessfully terminated pro-bation.
The majority of these individuals were still under supervision at the conclusion of
the evaluation.
Completing Treatment Programs
Is Associated with Successful
Completion of Probation
Probationers who consistently participated in drug or alcohol abuse counseling had signifi-cantly
higher probation success rates, lower subsequent arrest rates, and fewer positive drug
tests during probation than those who participated in but did not complete such programs.
Their success in completing probation was also better than that for probationers who did
not participate in such programs.
Completing drug treatment is associated with successful probation outcome—Individuals
who completed drug treatment or counseling programs were significantly more likely to
complete probation than were other probationers.1 As Figure A (see page 13) shows, 85 per-cent
of individuals who graduated from their most recent drug treatment or counseling
program also successfully completed probation. By contrast, only 22 percent of those who
did not graduate from treatment or a counseling session completed probation. Although
attending drug treatment/counseling was often a condition of probation, not completing it
did not necessarily lead to revocation.
As Figure A also shows, probationers who completed their treatment had a higher proba-tion
success rate than those who were not ordered to attend drug treatment or counseling.
Only 57 percent of those not required to attend drug treatment or counseling successfully
finished probation. The fact that those graduating had a significantly higher probability of
completing probation than those without a known drug problem further demonstrates the
impact of successful treatment.
Further, probationers graduating from treatment/counseling programs had significantly
fewer subsequent arrests than non-graduates did, and had significantly fewer positive drug
tests. Although treatment graduates had significantly fewer positive drug tests than non-graduates,
the graduates, on the average, still had some problems with drugs as indicated
1 Information was collected on up to four substance abuse treatment programs probationers attended while on
probation. The analyses are based on the outcome of the program they most recently attended.
13
by some probationers having positive drug tests after treatment. On the other hand, treat-ment
graduates had rearrest rates similar to those not sent to treatment. It may be that
treatment graduates were more motivated in general to complete the requirements of pro-bation.
Figure A
Adult Probation Outcomes Study
Percentage of Successful Probations
and Result of Final Drug Treatment
As of September 1, 1998
Source: Auditor General staff analysis of sample of probationers’ records who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Frequent attendance in 12-step programs predicts probation success—Similar results were
obtained for substance abusers who frequently attended 12-step drug or alcohol programs
such as Alcoholics Anonymous. These probationers were significantly more likely to com-plete
probation (80 percent), than were those who seldom attended (42 percent). And for
further comparison, only 57 percent of those not required to attend 12-step programs suc-cessfully
completed probation. In addition, substance abusers who frequently attended 12-
step programs also had significantly fewer arrests and positive drug tests than did those
who seldom attended programs. As was the case with drug treatment graduates, frequent
12-step attendees still experienced some substance abuse problems and some were arrested
57%
22%
85%
Treatment Not Treatment Not Treatment
Required Completed Completed
(n=283) (n=111) (n=204)
Percentage Completing
Probation
14
again, but abuse problems and recidivism rates were less than those observed for proba-tioners
who did not take full advantage of 12-step programs.
Consistent Employment and Community
Service Are Associated with Successful
Completion of Probation
While the issue of treatment programs relates to probationers with substance abuse prob-lems,
two other factors emerged as being strongly associated with positive probation out-comes
for all types of probationers. These two factors were consistent employment during
probation and completion of the community service requirement. In both cases, there was a
substantial difference in probation success rates for those who maintained consistent em-ployment
or who completed community service and those who did not.
Consistent employment is associated with successfully completing probation—Probation-ers
who were consistently employed throughout probation successfully completed proba-tion
at a 90 percent rate, as indicated in Figure B (see page 15). This success rate was some-what
better than that for probationers who were employed most of the time and signifi-cantly
better than that for probationers who were employed sporadically or not at all. Con-sistently
employed probationers also had significantly fewer arrests and positive drug tests
during probation than those who were employed only sporadically.
In addition, probationers who were employed on a full-time basis at the beginning or end of
probation completed probation at a significantly higher rate than individuals unemployed
or employed on a part-time basis at those times. Full-time employment at the beginning or
end of probation was also associated with fewer arrests and fewer positive drug tests.
Completion of community service is also associated with successful probation—Probation-ers
who completed their community service requirement completed probation more often
than did other probationers. As indicated in Figure C (see page 15), 84 percent of those
completing community service also completed probation successfully. Only 40 percent of
those who did not complete community service completed probation, and 61 percent of
probationers not required to perform community service completed probation. Again,
community service was a condition of probation, but not completing it did not always lead
to revocation. As was the case with probationers who frequently attended 12-step programs,
or completed drug programs, those who completed community service may have been
more committed to completing probation than were other probationers. These probationers
also had fewer positive drug tests and tended to have fewer subsequent arrests than did
those who did not complete community service.
15
4 1% 4 3%
7 8%
9 0%
Figure B
Adult Probation Outcomes Study
Consistency of Employment and
Percentage of Successful Probations
As of September 1, 1998
Figure C
Adult Probation Outcomes Study
Percentage of Adult Probationers
Successfully Completing Probation
and Community Service Participation
As of September 1, 1998
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Sporadically Employed Most Consistently
Unemployed Employed of the Time Employed
(n=59) (n=141) (n=125) (n=112)
Percentage Completing
Probation
Community Service Community Service Community Service
Not Required Not Completed Completed
(n=312) (n=179) (n=165)
61%
40%
84%
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Percentage Completing
Probation
16
Repeated Arrests and Petitions
to Revoke Were Linked
to Probation Failure
Higher numbers of arrests and petitions to revoke during probation were associated with
probation failure. As could be expected, new arrests lead to petitions to revoke, and peti-tions
to revoke lead to revocation to jail or prison. And, revocation to jail or prison is, by
definition, unsuccessful probation. By itself, an arrest or petition to revoke did not automati-cally
lead to revocation to jail or prison. Decisions on revocation are made on an individual
basis. Repeated arrests or petitions to revoke, however, were the factors most commonly
associated with probation being terminated unsuccessfully.
Those assigned to Intensive Probation Supervision were less likely to successfully complete
their probation than were other probationers. These individuals were considered high-risk
probationers, were supervised more closely, and were required by statute to be revoked
under certain conditions.
Greater Use of Treatment
Programs Could Further
Improve Probation Success
Conditions of probation, such as substance abuse treatment, afford probationers with op-portunities
for rehabilitation. Substance abusers who conscientiously took advantage of
those opportunities had higher probation success rates and lower recidivism rates than did
the probation population at large. However, almost one-third of the probationers ordered
by the courts to go to treatment did not attend. Fully utilizing these programs (as a condi-tion
of probation during sentencing), and strongly emphasizing completion through proba-tion
supervision may lead to more successful probations in the future.
Many ordered to substance abuse treatment did not attend—At the time this evaluation was
completed, a number of probationers with substance abuse problems did not receive drug
counseling or attend treatment programs. For example, 161 of 536 probationers ordered by
the courts to go to treatment did not attend treatment and were more likely than those who
did attend to have their probation revoked. Failure to attend may have been partially due to
lack of funding and availability of such programs. However, the Drug Treatment Education
Fund, established in 1996, provides funding and requires the courts to direct substance
abusers to treatment.
Additional screening could identify others who would benefit from treatment programs—
With the potential availability of substance abuse treatment increased, efforts to screen of-fenders
for abuse problems may be expanded. Auditor General staff analysis of the data
found that probation officers identified a number of substance abusers through the course of
probation and directed them to treatment programs. More specifically, 69 of the 309 proba-
17
tioners who were not initially ordered by the courts to attend treatment eventually were
directed to attend, and 33 of these probationers subsequently received treatment. More ef-fective
screening for abuse problems prior to sentencing may result in placing more abusers
in treatment and placing them there more quickly.
Recommendations
The Arizona Adult Probation System should:
1. Continue to screen offenders thoroughly for substance abuse issues prior to sentencing,
so that all probationers in need may be directed to substance abuse treatment or coun-seling
services.
2. Use monies from the Drug Treatment Education Fund to assure substance abuse treat-ment
options are available for all probationers who need them.
3. Supervise probationers with special attention paid to completing substance abuse treat-ments
and community service, and emphasize consistent employment as a condition of
probation, as these factors significantly predict successful probation outcomes.
18
(This Page Intentionally Left Blank)
19
FINDING II
COUNTIES’ DEGREE OF TOLERANCE
FOR CRIMINAL AND NONCOMPLIANT
BEHAVIORS AFFECTS PROBATION OUTCOMES
When probationers were arrested or otherwise violated the conditions of probation, the four
counties studied varied substantially in whether they allowed offenders to remain on pro-bation
or remanded them to jail or prison. These variations reflect differing levels of toler-ance
for relapse into criminal and/or substance-abusing or noncompliant behaviors. The
differences in their tolerance for relapse were reflected in various outcomes of the probation
process, such as how many probationers met their court-ordered financial responsibilities.
However, additional research would be needed to determine the costs and benefits of vari-ous
levels of tolerance.
Counties Have Wide Latitude
in Setting Probation Policies
Probation departments have a dual responsibility to provide probationers with opportuni-ties
for rehabilitation while at the same time protecting the community from crime. In Ari-zona,
each county sets its own policies as to how to balance these two responsibilities and to
what degree relapse into criminal or substance-abusing behavior or noncompliance is toler-ated
as part of the rehabilitation process. One of the key areas involved relates to deciding
when, and under what circumstances, to revoke probation. If a probationer commits an of-fense
or does not comply with the terms of probation, a determination must be made
whether to keep the offender on probation (perhaps with modifications to the conditions) or
to revoke the probation entirely. Because rehabilitation is generally more likely to occur in
community settings than in prison, keeping the offender on probation is often seen as an
attempt to provide additional opportunities for rehabilitation. However, many probationers
relapse into criminal or substance abuse behaviors. Probation departments can vary in the
tolerance they have for continuing rehabilitation in the community or sending the offender
to jail or prison, which may perhaps protect the community against the possibility that the
offender will commit further crimes during the probation.
Probation officers have a statutory responsibility to bring defaulting probationers into court
when they judge that the offender’s conduct justifies the sentence being revoked. Arizona
statute gives courts the discretion to revoke probation or modify conditions of probation.
Probation officers also have a responsibility to make recommendations to modify conditions
when the probationer’s behavior exceeds expectations.
20
The process of bringing offenders back to court for new sentencing involves Petitions to Re-voke
(PTR). PTRs, which are completed by probation officers, may be filed for probationers
who have committed a new crime or for technical reasons involving noncompliance with
probation conditions. When a PTR is filed, the court has the discretion of sentencing the of-fender
to jail or prison, ordering stricter monitoring, reinstating the offender with added
conditions, or dismissing the petition. PTRs can also be withdrawn by a probation officer.
However, the courts must revoke the probation and remand the individual to prison if the
probationer is on Intensive Probation Supervision at the time of the new arrest.
Tolerance for Relapse to Criminal
and Noncompliant Behaviors
Differs Among Counties Studied
The four counties in the study vary in the extent to which their policies, particularly toward
probation revocation, tolerate criminal relapse and noncompliant behaviors. The greatest
distinction between tolerance for relapse was that of Maricopa County, which places a great
emphasis on rehabilitation and tolerates more noncompliant and criminal behavior; and
Pima County, which allows probationers only limited leeway. Yavapai and Cochise are
somewhere between the two, with Yavapai closer to Maricopa, and Cochise closer to Pima.
Maricopa has higher tolerance for criminal and noncompliant behaviors—Maricopa
County’s focus is on rehabilitating of the probationer and promoting the use of intermediate
sanctions if probationers relapse into criminal, substance abuse, or noncomplaint behaviors.
The County’s written procedures instruct probation officers to use counseling and warn-ings,
or to modify supervision techniques if probationers are arrested for nonviolent crimes
or otherwise commit violations of their probation conditions. Similarly, the County Superior
Court has adopted a policy of intermediate sanctions, such as new conditions including
short jail time, or substance abuse treatment rather than incarceration. The County uses
PTRs as an intermediate sanction or warning, and also to refer probationers to additional
rehabilitation interventions.
The Pima County model is less tolerant of criminal and compliance relapses and offers
fewer options for intermediate sanctions. The County’s written procedures say that a gen-eral
unwillingness to abide by supervision requirements subjects a probationer to revoca-tion.
The probation department does not rely on PTRs as a sanction before revocation.
Differing tolerance levels are reflected in number and timing of revocations—Pima County,
revoked more probationers to jail or prison than did Maricopa County (38 vs. 29 percent).1
Differences in policies influenced not only the percentage of probationers revoked, but also
how soon the revocation occurred. Pima County revoked probations after an average of 14
1 No county differences in outcomes for probationers originally placed on Intensive Probation Supervision
were found.
21
months. Maricopa, by contrast, revoked them after an average of 24 months. No significant
differences in the reason for filing a PTR (new crime, technical violation, or both new crime
and technical) were found for PTRs that resulted in revocations.
Compared to Pima, Maricopa was more likely to reinstate probationers with additional
terms or change the level of supervision. The Maricopa County Criminal Justice System also
used concurrent probation more often than other counties. Concurrent probation means an
offender has committed a new crime and is serving two or more probation sentences si-multaneously.
These sentences overlap, meaning each day served on one term counts to-ward
completion of the others.1 Within the sample, 75 Maricopa probationers were serving
multiple probation terms simultaneously, with 17 of them serving three concurrent terms. In
the three other counties combined, 30 probationers were serving concurrent terms, and only
one of them was serving three concurrent terms.
An example of a revocation that occurred in each of the two counties can help illustrate the
differences in the counties’ approaches. Each example comes close to matching the county’s
statistical average for months on probation and number of PTRs issued before the probation
was revoked.
n Example 1: Maricopa County—Before being remanded to prison, this probationer
served 22 months in community supervision and had two previous PTRs. This offender
was convicted of drug possession and sentenced to three years’ probation beginning in
August 1994. In addition to conditions requiring payment of fees and fines, he was or-dered
to substance abuse treatment, prohibited from drinking alcohol, and ordered to
serve four months’ jail time. He seldom worked during probation, failed two drug
treatment programs, and tested positive for marijuana and cocaine. He was revoked in
November 1994 for technical violations and reinstated after serving more jail time. He
was revoked again for technical violations and noncompliance in August 1995 and was
subsequently reinstated. In June 1996, he was remanded to prison.
n Example 2: Pima County—Before being remanded to prison, this probationer served
13 months in community service and had no previous PTRs. He was sentenced to three
years’ probation for a drug offense in July 1994. He was ordered to serve two months in
jail, perform 24 hours of community service, and pay fines. He was revoked to prison in
August 1995 for a new arrest and technical violations.
Yavapai’s and Cochise’s approaches are between Maricopa’s and Pima’s—The two other
counties have approaches somewhere between Maricopa’s and Pima’s. Similar to Maricopa
County, Yavapai County appears to take a more traditional approach to relapse to criminal
1 If one term expires, however, offenders are not released until all the terms are served. Within the
constraints of the criminal code, judges are able to impose different conditions for each new conviction.
22
and noncompliant behaviors by probationers. Its probation officers consult with the appro-priate
judge to decide if a PTR will be filed. Cochise County’s approach is similar to Pima
County’s. Its probation officers make a decision about filing a PTR based upon the severity
and number of technical violations. In instances of a new felony arrest or an arrest for a
crime similar to their probation offense, Cochise County officers are directed to file a PTR on
the probationer.
Approaches Produce Varying
Probation Outcomes
The differing levels of tolerance for criminal and noncompliant behaviors produce different
outcomes in the probation process. Probation completers in Maricopa County were more
likely to pay full restitution, do their community service, and complete probation early.
However, revoked probationers in Maricopa County were also the most likely to have
committed additional crimes before revocation. The counties showed little difference, how-ever,
in the degree to which fines were paid.1
Maricopa County probationers pay full restitution and complete community service more
often—Maricopa County’s probation completers are more likely to pay full restitution than
are probation completers in Pima County. Probationers in Maricopa County paid full resti-tution
86 percent of the time. As seen in Table 3 (see page 23), Pima County probationers
paid full restitution 77 percent of the time. Since Maricopa County had a larger portion of
offenders who were employed when their probation began, it may be these probationers
had more opportunity to pay money owed.
Yavapai County probationers paid full restitution only 60 percent of the time. The signifi-cantly
smaller number of probationers in Cochise County paid full restitution only 33 per-cent
of the time.
Probationers in Maricopa County completed community service, which was a frequent
condition of probation, more often. Maricopa County probationers completed court-required
community service 68 percent of the time, compared to 58 percent for Pima and 53
percent for Cochise.
1 Only probation completers are included in the analyses specific to payment of restitution, completion of
community services, and payment of fines and fees.
23
Table 3
Adult Probation Outcomes Study
Percentage of Probationers Who Paid All Fines, Fees, and Restitution
and Performed All Community Service
As of September 1, 1998
County Fines
Probation
Fees
Victim
Restitution
Community
Service
Maricopa 67% 68% 86% 68%
Pima 60 61 77 58
Yavapai 75 69 60 56
Cochise 38 32 33 53
Percentage for sample
as a whole 64 64 74 64
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Early completions earned more frequently in Maricopa County—Maricopa County proba-tioners
are more likely to earn early completion than are probationers in the other three
counties. Fourteen percent or fewer of probationers in Pima and Cochise Counties earned
early closure. However, Maricopa County had twice as many probationers, or 28 percent,
earn early closure.
To complete probation early, probation conditions must be satisfied, including the paying of
all fines and restitution, and the completion of community service. Early completers not
only satisfy their responsibility to their community admirably and completely, but their
termination allows court and probation department resources to be used on remaining
caseloads.
The longer a probationer remains in the community, the greater the opportunity to re-offend—
Most probationers in the study were not arrested while serving their probation
sentence. However, some probationers, especially those whose probation was eventually
revoked, were arrested while on probation. Figure D (see page 24) shows the number and
types of crimes that resulted in probationer arrests. A very high percentage of the arrests
were for drug or alcohol use, or for crimes known to be related to drug and alcohol abuse,
such as shoplifting, theft, and family violence.
The arrest rates per month for these probationers were similar across all four counties.
meaning that over a 12-month period of time, the same number of crimes per probationer
would be committed regardless of the county. But, the extra time on probation afforded to
24
Property Crimes/
Disorderly
Conduct
13%
Assaults
12%
Sex Crimes
4%
Family Violence
16%
Weapons
5%
Drugs/DUI
29%
Theft/Shoplifting
21%
Maricopa and Yavapai probationers, as compared to their counterparts in Pima and Cochise
Counties, resulted in more total arrests. For example, for every five revoked probationers in
Maricopa County, an average of two more crimes are committed than occur for every five
revoked probationers in Pima County.
Table 4 (see page 25) indicates how the counties differ in the use of revocation and how that
relates to arrest rates. As indicated by the table, the average number of arrests while on pro-bation
for Maricopa and Yavapai County probationers who are eventually revoked is higher
than for Pima and Cochise probationers.
Figure D
Adult Probation Outcomes Study
Arrests During Probation by Type of Crime
Shown as a Percentage of All Arrests Recorded
As of September 1, 1998
Number of recorded arrests = 186 (excludes arrests for which charges were dropped).
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
25
Fees and fines payment similar across counties—The different approaches taken by proba-tion
counties appear to have no impact on whether court-imposed fees and fines are even-tually
paid. Maricopa, Pima, and Yavapai Counties had probationers paying all fees and
fines 60 to 75 percent of the time. The only notable difference among counties came from
Cochise County, where persons who completed probation paid 32 percent of their fees and
38 percent of their fines. However, Cochise County secures civil judgments for money that
probationers still owe at the time of termination. This procedure allows parties involved to
continue pursuing payments after probation.
Table 4
Adult Probation Outcomes Study
Revoked Probationers
Comparison of Average Arrests During Probation, and
Average Number of Months Before Revocation, by County
As of September 1, 1998
County
Average Arrests
Prior to Revocation
Average Months
Before Revocation
Maricopa .72 24
Pima .37 14
Yavapai .87 20
Cochise .20 13
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Additional Study
Necessary
It is not clear how the costs and benefits associated with probation departments’ different
tolerances of probationers’ criminal and noncompliant behaviors contrast with one another.
A follow-up on early completers and their criminal recidivism rates after probation, com-pared
to recidivism rates of probationers whose terms expire, would help answer these
questions. Since most recidivism occurs within a short period of release from supervision, a
long time span would not be necessary. A follow-up of one to two years should be sufficient
to answer the question. In addition, a comprehensive comparative analysis of behaviors,
while under supervision and after release, of probationers who are eventually revoked
would answer some of the questions regarding revocation practices and the costs to the
community. This question could also be analyzed using a design with random assignment
of high-risk probationers (those most likely to be revoked) to a group who receives interme-diate
sanctions and to another group who does not.
26
Recommendation
The Adult Services Division of the Administrative Office of the Courts should work with
County Probation Departments to determine the longer-term cost-effectiveness of different
degrees of tolerance of criminal and noncompliant behaviors, in order to provide commu-nity-
based rehabilitation.
27
FINDING III
PROBATIONER SUCCESS RATES
VARY BY GENDER AND
OTHER CHARACTERISTICS
Previous studies have found that probation outcomes vary with such characteristics as age,
gender, and education. Like these other studies, this evaluation found differences in success
rates, but it also found more complex relationships between the various characteristics ana-lyzed
than has been noted before. Outcomes vary between the sexes, between ethnic
groups, and between those receiving and not receiving needed mental treatment. Analysis
of probationers who abscond (flee the court’s jurisdiction and whose whereabouts are un-known),
shows that this group also has characteristics that differ from probationers who do
not abscond.
Background
Of particular interest among previous studies of probation outcomes is the Arizona Su-preme
Court’s 1995 study of adult probation outcomes. That study found the following:
n Female probationers were more likely to complete probation than males.
n The older the offender was at the time of sentencing the more likely he or she was to
complete probation.
n Of both male and female probationers, those with at least a high school education are the
most likely to complete their probation.
n Anglos were more likely to complete probation than were Hispanics or African-Ameri-cans.
The findings reported here generally replicate these findings, but they also show additional
complexities in the relationship of the characteristics analyzed.
28
Comparisons Between Men
and Women Show Substantial
Differences in Probation Outcomes
Probation outcomes differ by gender. As a group, women are somewhat more likely to
complete their probation, and, if they violate conditions of their probation, they are less
likely to have their probation revoked. However, while education, age, and income are all
positively correlated with probation success for men, these relationships are different for
female probationers.
Success rates and sanctions differ—For the sample group studied in this analysis, women
had a somewhat higher success rate in completing probation than men did. Sixty-five per-cent
of women successfully completed probation, compared with 60 percent for men. The
sanctions for violating terms of probation also differ. Women who do not fully comply with
terms of probation are more likely to have their probation term lengthened, while men who
do not comply are more likely to have their probation revoked and be sent to jail or prison,
or given some other more restrictive sentence. Women who did not fully comply with the
terms of their probation were three times as likely as men to have their time on probation
lengthened.
Relationship between probation completion and education, age, and income also varies by
gender—Three characteristics that are commonly associated with success rates in probation,
education, age, and income, were analyzed across gender lines, and the relationships were
not the same for women as for men.
n Education—Typically, higher education levels are generally related to probation com-pletion.
However, in the current evaluation the impact of education on successful com-pletion
of probation is much weaker for women than for men. Another way in which
education levels affect women differently is that more highly educated women are even
more likely to have their terms lengthened rather than revoked. These differences are
more fully illustrated in Table 5 (see page 29).
n Age—When men and women are considered together, the results show a positive rela-tionship
between age and successful completion of probation; that is, the older the per-son,
the greater the likelihood of successful completion. However, when the sexes are
analyzed separately, this relationship changes. For example, males 20 years of age and
younger are the least likely age group to successfully complete probation, but women of
this age group are more likely than all but the oldest group to be successful. Figure E (see
page 30), presents successful completion rates by age and gender.
n Income—Typically, probationers with higher incomes at the beginning of probation are
more likely to complete probation. However, when the sexes are analyzed separately,
the impact of income to probation completion disappears for women.
29
Table 5
Adult Probation Outcomes Study
Probation Outcomes for Men and Women
By General Education Level
In Percentages
As of September 1, 1998
Men Women
Less
than
High
School
High
School
Graduate
or GED
Some
College
Less
than
High
School
High
School
Graduate
or GED
Some
College
Unsuccessful
Probationer absconded 8% 3% 2% 1% 3% 2%
Probation revoked 36 34 25 30 21 17
No Final Outcome
Probation term
lengthened
4 5 6 12 9 20
Successful
Probation term expired 31 33 35 32 49 29
Early positive probation
termination 21 25 32 25 18 32
Number of cases 220 157 109 69 33 41
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Differences in Revocation
and Early Completion Found
for Different Ethnic Groups
While Anglos are more likely to complete probation than were minorities, the differences in
completion rates are not significant. However, when outcomes are examined more closely,
for example, completion versus early completion, some differences emerge. Hispanics and
African-Americans are more likely to have their probation revoked and less likely to have
an early termination of probation than Anglos. However, differences among ethnic groups
in the number of petitions filed to revoke probation, and severity of the crime that led to
probation, appear to account for these differences in outcomes.
30
Figure E
Adult Probation Outcomes Study
Percentage of Probationers Who Successfully
Completed Their Probation by Gender and Age
As of September 1, 1998
Number of cases = 649
Cases by gender and age group indicated in parentheses.
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Revocation rates were 27 percent for Anglos, 35 percent for African-Americans, and 32 per-cent
for Hispanics.1 Early termination rates were 29 percent for Anglos, 16 percent for Afri-can-
Americans, and 18 percent for Hispanics. For all three ethnic groups, the average num-ber
of months before probation was revoked was about 21.
African-Americans whose probation was revoked had a higher rate of arrests during their
probation and a higher rate of petitions to revoke filed than either Hispanics or Anglos.
They also were likely to be on probation for more serious crimes. The portion of offenders
1 The numbers of Native Americans and Asian-Americans in the sample were relatively small, so they were
excluded from this analysis.
0%
20%
40%
60%
80%
100%
Male Female
20 and 21 to 25 26 to 30 31 to 35 36 to 50 51 and older
younger
(68) (8) (98) (27) (99) (25) (110) (37) (125) (32) (17) (3)
Probationers’ Age
31
on probation for class 2 or class 3 felonies, and whose probation was subsequently revoked,
was 30 percent for African-Americans, 10 percent for Anglos, and 1 percent for Hispanics.
Without Appropriate Treatment, Probationers
with Mental Health Problems or Mental
Disabilities Are Less Likely to Succeed
Appropriate treatment is important for the probation success of offenders with mental
health problems or mental disabilities. Probationers whose mental disabilities or illnesses
are identified at the time of sentencing and who receive appropriate treatment do just as
well in the probation system as the general population. However, probationers with mental
problems that are not identified and addressed in their conditions of probation do not fare
well. Similarly, identification and subsequent assistance may help mentally disabled proba-tioners
succeed, but without early identification and specialized assistance, probationers
with mental problems are less likely to be successful.
The sample of probationers included 24 probationers who had been identified by our
evaluation as having a mental or emotional health problem or mental disability, but who
had no recognition of this condition or no prescribed treatment in their conditions of proba-tion.
These probationers had a successful completion rate of only 46 percent, compared to 62
percent for the general probation population. While the number of these probationers in the
sample is small they were significantly more likely to be revoked to prison or jail than those
with no such identified problems (44 percent vs. 27 percent) and much less likely to have
completed probation early (11 percent vs. 26 percent). By contrast, the sample included 66
probationers who were required to have mental health counseling as a condition of their
probation. This group had no significant differences in their probation outcomes from the
probation population as a whole.
These findings suggest that greater efforts at identification and treatment could yield better
chances of success for this segment of the probation population. For example, early identifi-cation
and recognition of mental health problems as a condition of probation may help to
improve these probationers’ chances of successful probation. Both the Maricopa and Pima
County Probation Departments assign probationers with mental health problems to proba-tion
officers with specialized training in supervising these offenders.1 It appears that these
assignments contribute to the probationers having positive outcomes. While based on small
numbers of cases, probationers in the other counties who were identified as having mental
health problems, but did not have the opportunity for specialized assignment, tended to
have very poor experiences and negative outcomes.
1 Such grouping also exists for sex offenders, and those convicted of domestic violence.
32
Absconders Differ from Other
Probationers in Education and Age
The approximately 6 percent of probationers who abscond are generally similar to the pro-bation
population as a whole except for two characteristics, education and age. In other
characteristics, such as employment or prior criminal history, they are generally similar. For
example, employment and marital status do not distinguish absconders from other proba-tioners.
Education levels lower for absconders—Absconders had an average lower education level,
completing an average of 10.06 years of schooling compared to 11.46 years for the general
probation population. Only about one-third of the absconders have a high school diploma
or GED, compared to about one-half of those whose probation was revoked and about 60
percent of those who complete probation successfully.
Absconders over-represented in one age group—Absconders are over-represented in the 26-
to 30-year-old age group. Forty percent of absconders fall into this age group, compared to
fewer than 20 percent of all probationers.
Absconders are not distinguished by employment or prior criminal history. Employment
status at the beginning of probation does not distinguish absconders from other probation-ers.
Absconders have risk levels and average numbers of prior supervisions very similar to
probationers who complete probation and different from those who are revoked. Abscond-ers
actually have fewer prior felonies than probationers who are revoked and fewer than
even the average probationer who completes his or her term. In addition, absconders have
fewer prior misdemeanors than either probation completers or probationers who are re-voked.
Conclusions
The various findings suggest that different groups bring different risks and strengths to
their probation and may require variations in supervision. Better understanding of these
subpopulations can help identify demographic characteristics that can be used as indicators
of probation outcomes and may help probation officers to make better decisions in their su-pervision
of probationers.
Recommendation
The Administrative Office of the Courts should work with the smaller counties to develop
specialized caseload assignments for probationers with mental health problems.
33
FINDING IV
RECORDKEEPING NEEDS TO IMPROVE
Basic management information collected by county probation departments is difficult to
retrieve and use in a timely manner. Poor information management can affect probation
officers’ ability to adequately supervise offenders and can inhibit effective program man-agement
and decision-making. The Administrative Office of the Courts should collaborate
with the 15 county probation offices to create minimum standards for paper and electronic
case files management.
Good Recordkeeping Critical
to Probation Management
Accurate and complete documentation in probation files is important for a number of rea-sons:
n Decision-making—Probation officers are responsible for providing information to
judges and administrators at a number of important times during the sentencing and su-pervision
of probationers. Besides being used during a probationer’s initial sentencing,
additional information about a probationer’s progress is used in making recommenda-tions
about revocation, adding or deleting conditions of probation, changing levels of su-pervision,
or determinations about whether probation has been successfully completed.
Recommendations based on insufficient or inaccurate information could be detrimental
to the defendant and/or the community.
n Case management—Poorly maintained records reduce the probation officers’ ability to
assess probationers’ performance either quickly or accurately. Officers need to under-stand
the risks and status of the people they supervise, particularly when assigned a new
and unfamiliar case. Probation officers need to have complete and accurate information if
they are to meet their responsibilities in protecting the community and assisting the of-fender
to change. Likewise, probation officers need to identify and document violations
quickly in order to intervene with timely sanctions. They also need to document proba-tioners’
progress to assist them in finding services and activities that can help them
change their behaviors.
n Protecting victims—Incomplete and inaccurate information could limit a probation offi-cer
from meeting his or her responsibilities regarding informing victims of restitution ar
34
rangements, the offenders’ supervisory conditions, and potential dangers from the pro-bationer.
n Research and evaluation—Inaccurate and incomplete information limits the Admin-istrative
Office of the Courts and the County Probation Departments’ abilities to meas-ure
short-term and intermediate goals and conduct cost-benefit analyses. Without good
data there is not an effective ongoing system of monitoring and evaluating probationers
and the system and no sufficient mechanism to communicate the impact it has on public
safety.
Many Recordkeeping
Problems Surfaced While
Auditors Assembled Data
for This Evaluation
In analyzing the sample of probationers selected for this evaluation, some situations were
encountered in which probation files contained conflicting information, lacked essential
details, were illegible, contained duplicate information, and were unobtainable. This lack of
information can make it difficult for probation officers to adequately supervise probationers
and places great limits on administrators’ abilities to assess and manage their departments
and programs. The types of problems encountered in the four-county review included the
following:
n Missing information—Basic information was often lacking. For example the offender’s
state identification number and standard criminal justice codes identifying their crimes
were often not available. A disturbing illustration of incomplete information is the in-complete
recording of new arrests.
n Conflicting information—Files also contained conflicting information about the proba-tioner’s
demographic information and probation outcome.
n Treatment details missing—While all four counties have probationers who were court-ordered
to obtain drug or alcohol abuse treatment, consistent details were missing about
treatment programs entered or completed.
n Incomplete information for transfers—Files for probationers who had transferred
from another county or another state were particularly difficult to garner information
from. In these cases, background information on the offender and personal details were
often missing.
35
n Files not found—Maricopa County had difficulty obtaining active probationers’ files
and locating others thought to be in storage. Other counties were also unable to find
some files selected for the study.
n Handwritten information—Handwritten conditions of probation were difficult to deci-pher
and confirmation of an offender’s conditions of probation had to be sought else-where
in the file record. Pima County’s paper files were the most problematic because
probation officers wrote their notes by hand. Officer notes on several files could not be
read at all, causing gaps in data collection.
n Disorganized files—Maricopa County paper files were the most disorganized of the
four counties. For example, they contained multiple copies of the same document and
inconsistent information was also found. Overall, Yavapai, and Cochise records were
better organized and consistently decipherable.
n Unreliable electronic files—Overall, as discussed in Appendix A, reliable electronic
files were not available from any of the four counties. For example, in Maricopa County,
electronic and paper files had inconsistent information for the beginning and ending
dates of an offender’s probation.
n County electronic files are not standardized—This evaluation was impeded by the
fact that counties did not collect, record, and report information in the same manner and
format. (See Appendix A, pages a-i through a-vii for additional discussion.)
Improvements in Case File
Accuracy and Electronic Formats
Could Improve Accountability
The Adult Services Division of the Administrative Office of the Court (AOC) should work
with the counties to create standards for paper and electronic filekeeping. At present, there
are no statewide standards for what should be included in paper or electronic files. Al-though
collecting data from paper files is time consuming and costly, developing standards
for paper files appears to be the place to start. If the State is to track the effectiveness of pro-bation,
working with paper files is currently necessary. Within the past five years at least
two other adult probation outcome studies in Arizona found that electronic files were es-sentially
useless.
Building on standards agreed upon by AOC and the county probation departments for pa-per
files, the AOC should then work with the counties to develop consistent and accurate
electronic case files. Improved electronic filekeeping is necessary for courts to assess and
direct probation activities. It would also allow more timely assessment of probation pro-grams.
An adequate data system would allow work that currently consumes four months
because paper files must be reviewed to be completed in a matter of days.
36
Recommendations
In cooperation with the county probation departments, the Adult Services Division of the
Administrative Office of the Courts should:
1. Set and enforce minimum standards of information maintained in paper files for active
and closed adult probation files.
2. Set and enforce minimum standards of electronic information maintained for active and
closed adult probation files.
3. Explore the feasibility of creating a standardized electronic database for recording and
maintaining probation case files.
Agency Response
(This Page Intentionally Left Blank)
March 5, 1999
Mr. Douglas R. Norton, Auditor General
Office of the Auditor General
2910 North 44th Street, Suite 410
Phoenix, Arizona 85018
Dear Mr. Norton:
Thank you for the opportunity to comment on the performance audit report of the Administrative
Office of the Courts’ Adult Services Division. Before addressing each of the recommendations, I
would like to make several comments.
We are very pleased that the evaluation by your staff focused on identifying factors contributing to
an adult’s successful completion of probation and the positive results from an offender’s
participation in substance abuse treatment. Your findings validate what those of us who manage
community corrections programs have believed for some time; probationers who maintain
employment and complete community work service requirements have a greater likelihood of
successfully completing probation. Your evaluation further validates an outcome study conducted
by the Administrative Office of the Courts in 1995, as well as fifteen years of national research
which concluded that substance abusing probationers who consistently participate in
treatment/counseling programs are much more successful on probation. It is now clear that in
Arizona, not unlike the rest of the country, probation supervision coupled with treatment increases
positive behavioral change of the offender and decreases revictimization of the community.
The citizens of Arizona also recognized the benefit of, and need for, resources for substance abusing
offenders by establishing the Drug Treatment and Education Fund (DTEF), which became
operational during fiscal year 1998. During it’s first year of operation, this fund provided substance
abuse treatment services to 2,622 probationers statewide. It is anticipated that the number of
probationers receiving treatment/counseling services funded by the DTEF will rise to approximately
3,496 during fiscal year 1999.
While the availability of DTEF monies will continue to have a considerable impact on the
availability of treatment resources for drug abusing probationers, the evaluation by your staff did not
note that community resources are limited. Your report notes that Department of Public Safety
Uniform Crime Reports show an 80% increase in the number of drug-related arrests from 1990 to
1996, and that Arizona law enforcement officials attribute a significant portion of crime to substance
abuse. However, there is no acknowledgment of the lack of agencies/programs statewide to
accommodate
2
the number of substance abusing individuals in Arizona, especially in the rural areas. The audit
findings were shared with each of the probation departments included in the evaluation and all agree
that probationers are competing with other citizens for treatment placements and may have to wait
weeks or months before space is available in a residential or group setting. While the state can
provide treatment funds, the positive impact of those dollars is diminished by the lack of program
capacity. It is also important to note that pursuant to statute, the DTEF can not be utilized for the
treatment of alcohol abuse, which also plagues Arizona.
The report points to the significance of employment and community work service in the successful
completion of probation. However, it does not recognize the need for expanded job readiness and
development programs, as well as an increase in not for profit agencies to provide probationers
opportunities to complete community service requirements.
The audit suggests the need for standardized electronic record keeping. For several years the
Supreme Court has been requesting funding from the Legislature to develop and deploy an adult
probation automated case management system and funding has not been provided. Absent funding
and a statewide system, standardized electronic files are next to impossible.
Finally, the analysis excluded 185 probationers in the sample who “neither successfully nor
unsuccessfully terminated probation”as the majority of them were still under supervision at the
conclusion of the evaluation. These probationers are, and should be considered, successful, as they
are in the community and complying with court orders.
RESPONSE TO RECOMMENDATIONS
FINDING I
The Arizona Adult Probation System should:
1. Continue to screen offenders thoroughly for substance abuse issues prior to sentencing,
so that all probationers in need may be directed to substance abuse treatment or
counseling services.
The finding of the Auditor General is agreed to and has been part of the probation process.
Risk and needs assessment tools have been used by the probation departments for over 10 years to
assist with identifying and prioritizing the resources needed by probationers. Maricopa and Pima
County operate an Assessment Center, designed to evaluate probationer need and identify
appropriate resources during the presentence process. These counties are also in the process of
examining and developing additional tools to facilitate probationer assessment. Finally, the Drug
Treatment and Education Fund program requirements mandate the use of two statewide screening
and assessment tools (Arizona Substance Use Survey and Offender Substance Abuse Profile) to
determine probationer need and appropriate treatment placement. These assessments were
completed on all
3
2,622 probationers receiving substance abuse treatment services through the DTEF in fiscal year
1998.
2. Use monies from the Drug Treatment and Education Fund to assure substance abuse
treatment options are available for all probationers who need them.
The finding of the Auditor General is agreed to and has occurred.
As previously noted, DTEF monies were utilized to provide substance abuse treatment to 2,622
probationers in need in fiscal year 1998. Of these probationers, approximately 10% participated in
substance abuse education, less than 1% in day treatment, 74% in outpatient counseling, 9% in
intensive outpatient counseling, 3% in short term residential treatment, and 4% in a long term
residential program.
3. Supervise probationers with special attention paid to completing substance abuse
treatments and community service, and emphasize consistent employment as a
condition of probation, as these factors significantly predict successful probation
outcomes.
The finding of the Auditor General is agreed to and will be implemented.
The Adult Services Division will request probation officers place greater emphasis on probationer
compliance with substance abuse treatment and community service requirements. Consistent
employment will also be stressed. However, the need for increased statewide treatment capacity,
not-for-profit agencies to provide opportunities for community service, and job readiness and
development services must be acknowledged by the Auditor General when setting probation
outcome expectations.
Finding II
The Adult Services Division of the Administrative Office of the Courts should work with
County Probation Departments to determine the longer-term cost-effectiveness of different
degrees of tolerance of criminal and noncompliant behaviors, in order to provide community-based
rehabilitation.
The finding of the Auditor General is acknowledged.
The audit primarily attributes the varying degrees of tolerance for criminal and noncompliant
behavior to the local probation departments and fails to acknowledge the significant roles of the
other entities involved. Local law enforcement, county attorney charging and plea bargaining
practices, and judicial discretion all substantially contribute to the varying degrees of “tolerance”
cited in the audit report. The issue spans the entire criminal justice system and can not be examined
solely from the perspective of the local probation department.
4
To conduct the type of evaluation suggested would be a tremendous undertaking requiring vast
resources to collect data across a multitude of criminal justice agencies. The Adult Services Division
does not have the personnel or funding available for such a project and the cost implications were
not included in the biennial budget request to the Legislature. Additionally, the basis for the funding
is fundamentally incorrect.
On a technical note, the revocation examples from Maricopa and Pima County presented on page
21 are markedly different (the Maricopa County case is limited to technical violations of probation
while the Pima County case involves technical violations plus a new arrest) and can not be
compared. Furthermore, no information has been provided on the probationer’s criminal history or
the nature of the new arrest.
It should also be mentioned that Cochise County is not the only jurisdiction which seeks civil
judgments for unpaid court assessments upon completion of probation, as indicated in the audit
report. In fact, Maricopa County was the first to initiate this practice and civil judgments are also
sought in Pima County.
Finding III
The Administrative Office of the Courts should work with the smaller counties to develop
specialized caseload assignments for probationers with mental health problems.
The finding of the Auditor General is not agreed to and a different method will be utilized to address
the concern.
Appendix A of the audit report states that “there was not an expectation to generalize statewide from
the four counties” yet this recommendation is based upon only 24 mental health cases and is
generalizing statewide. We do not believe that each local department can justify and support
specialized caseloads for mentally ill probationers. What is needed, however, is a continuum of care
for these probationers which is not being provided by all Regional Behavioral Health Authority’s
(RHBA) statewide. The Adult Services Division will commit to working with the local probation
departments to develop relationships with their local RHBA to facilitate the delivery of services to
mentally ill probationers. The Division will also work with the probation personnel and the RHBAs
to establish a training curriculum for probation officers to assist them in better managing the special
needs of these probationers.
Finding IV
In cooperation with the county probation departments, the Adult Services Division of the
Administrative Office of the Courts should:
1. Set and enforce minimum standards of information maintained in paper files for active
and closed adult probation files.
5
The finding of the Auditor General is agreed to and a different method of dealing with the finding
will be implemented.
It is agreed that establishing some minimum standards for what should be retained in a case file is
a desirable goal. However, Arizona’s probation system is decentralized and it would be
inappropriate for the state to mandate local file set up. The more appropriate recommendation would
be for the Legislature to provide funding for the development of an adult probation automation
system.
2. Set and enforce minimum standards of electronic information maintained for active
and closed adult probation files.
The finding of the Auditor General is acknowledged.
This recommendation should be directed at the Legislature. Minimum standards for electronic
record-keeping is a worthwhile goal and we stand ready to implement them if funding is provided.
3. Explore the feasibility of creating a standardized electronic database for recording and
maintaining probation case files.
The finding of the Auditor General is not agreed to and can not be implemented without further
resources.
The feasibility of a statewide adult probation database has already been addressed and the
Administrative Office of the Courts is actively involved with representatives from Maricopa and
Pima County’s in the development of an automated case tracking and management system. A
prototype of this statewide automation system is scheduled to be piloted in one field office in
Maricopa County this year, expanding to one field office in Pima County and one rural county during
the summer of 2000. Full implementation will, however, require funding from the state, the county
or both.
Again, we are very pleased that the evaluation by your staff validates our belief and national research
that probation supervision coupled with treatment results in positive behavioral change of the
offender and decreases revictimization of the community. I hope your findings will result in
legislative support for additional funding to increase treatment capacity and for the development of
an adult probation automation system.
I thank your staff for the professionalism demonstrated throughout the course of this audit and assure
you that those recommendations we agreed to will be implemented.
Sincerely,
David K. Byers, Director
Administrative Office of the Courts
DKB:des
cc: Barbara Broderick, Director, Adult Services Division
Mike DiMarco, Budget Director, Administrative Services Division
Appendix A
(This Page Intentionally Left Blank)
a-i
APPENDIX A
SAMPLING AND DATA COLLECTION
The findings in this report are based on analysis of data generated from 845 probation files
in four counties. The counties represented are Maricopa, Pima, Yavapai, and Cochise. Data
is for probationers who began probation in 1994.
Data was generated from analysis of 845 files. Electronic data was available from some
counties in Arizona. However, these data were incomplete, unreliable, and not always com-parable
to other counties; i.e., all counties do not collect the same information.
Probationers who began probation in 1994 were selected because most individuals placed
on probation in 1994 would have finished or be finishing their probation in 1998 (since pro-bation
terms are typically 3 years) and would have an outcome. At the same time, files for
individuals who began probation in 1994 and who had already terminated should not have
been destroyed. In addition, the time period is recent enough to have relevance for current
supervisory practices. However, approximately 21 percent of the cases did not have a final
probation outcome. This includes probationers who were still active on probation because
they had long terms, including lifetime terms, and probationers whose terms were ex-tended.
This group also includes probationers who died, or were transferred.
Maricopa and Pima Counties were selected because Pima and Maricopa County probation
departments supervise the majority of probationers in the State. Seventy-five percent of all
probationers in Arizona reside in these counties, with 62.3 percent in Maricopa and 12.7 per-cent
in Pima. Yavapai and Cochise counties were selected for comparison and illustration
and because they provided probationers with some options for substance abuse treatment, a
key element of the study. Convenience also entered into selecting Yavapai and Cochise
Counties. These counties were selected over other options because they provide a cross-section
of rural counties and allowed Auditor General staff to conduct research within the
time and resource constraints. Probationers from these four counties represent approxi-mately
80 percent of all probationers in Arizona’s 15 counties.
Sample Selection
The original sample size was set for Maricopa and Pima Counties at the 95 percent level of
confidence and 5 percent sample precision based on the outcome variable (successful com-pletion)
occurring at 50 percent. Yavapai and Cochise Counties, chosen for comparison and
a-ii
illustration, were not set at the same sample precision as Maricopa and Pima. Table 6 sum-marizes
the sampling process and confidence levels and sample precision.
Table 6
Adult Probation Outcomes Study
Summary of Sample Selection
County Strata
Number in
Strata
Number in
Sample
Number in
Data File
Confidence
Level
Sample
Precision
Maricopa Community Punishment
Program 155 155 145
Driving Under Influence 747 75 66
Drug offense 1,004 75 66
Drug possession 1,551 75 60
No substance abuse 427 75 62
Non-drug, crime, sub-stance
abuse history 3,872 150 128
Total Maricopa 7,756 605 527 95% 5%
Pima 1 Community Punishment
Program 80 50 24
Drug and alcohol abuser Unknown 50 46
Drug offender Unknown 50 33
Non-drug offense Unknown 50 29
Non-drug offense,
likely abuser   50 44
Total Pima 1,600 2 250 176 95% 7%
Cochise No strata
Total Cochise 155 75 74 95% 10%
Yavapai No strata
Total Yavapai 472 75 69 95% 10%
Total 9,983 3 1,005 846
1 Actual number of strata in Pima County is unknown.
2 Estimated number of individuals who began probation in Pima County in 1994.
3 Estimated.
Source: Auditor General staff analysis of data provided by Cochise, Maricopa, Pima, and Yavapai Counties.
a-iii
Maricopa accounts for more than 60 percent of all probationers statewide. Therefore, the
sample precision and confidence level set were based on the materiality of Maricopa County
to probationers statewide. It was also determined that comparison among the strata would
be made so strata needed to be large enough to do that. Six strata in Maricopa County were
defined: 1) attended Community Punishment Program; 2) convicted of DUI felony; 3) drug
offense other than possession; 4) drug possession; 5) non-drug crime and no substance
abuse history; and 6) a non-drug crime, but a documented substance abuse history. Strata 2
through 6 exclude all probationers originally sentenced to the Community Punishment Pro-gram.
A sample was also drawn from Pima County. However, Pima County did not provide a
reliable sampling frame and the sample was drawn from the presentence report file that
included all probationers sentenced to any form of punishment in Pima County. The origi-nal
file contained over 3,700 cases but approximately 1,000 of these were required to serve a
sentence in the Department of Corrections (DOC). In addition, the remaining 2,700 included
offenders who may have been sentenced to DOC and should not be included in the sam-pling
frame. However, a confidence interval of 95 percent with sampling precision of 5 per-cent
was planned. Since the level of detail provided on the offenders was not the same as
that provided by Maricopa County, it was not possible to draw the same strata. However,
since there was an inadequate sampling frame, strata were constructed from the available
information allowing for stratified sampling in order to minimize variance. It was expected
that over sampling 50 in each stratum would yield a minimum of 30 per strata for statistical
runs. Due to differences in the type of data provided by Pima and Maricopa County the
Pima strata are similar to, but not the same, as the Maricopa County strata. Five strata in
Pima County were defined: 1) Community Punishment Program; 2) non-drug offense but
no known abuse; 3) original charge is drug related; 4) known drug and alcohol abuser; and
5) non-drug offense but likely abuser. Due to budget and calendar constraints, the stratified
250 sample seemed a reasonable number. However, the resulting sample was actually lower
than 250 owing to coding errors from the poor sampling frame. Although a 95 percent con-fidence
is retained with the resulting 176 cases from Pima County, the sampling precision is
only 7 percent in contrast to the 5 originally planned.
Yavapai and Cochise Counties were not set at the same sample precision as Maricopa and
Pima for two main reasons. First, the small number of probationers in the two counties did
not warrant the same evaluation effort as Maricopa or Pima. Second, since there was not an
expectation to generalize statewide from the four counties, there was no need to have a
sample precision equal to Maricopa and Pima. A confidence interval of 95 percent with 10
percent sample precision for Yavapai and Cochise seemed both feasible and appropriate.
Tables 7, 8, and 9 provide additional background information on individuals in the sample.
a-iv
Table 7
Adult Probation Outcomes Study
Sample1 of Probationers by County
Percentage Assigned to Intensive Probation Supervision and
Percentage Identified as Male and Belonging to an Ethnic Minority
County
Intensive
Probation
Supervision Male Minority
Cochise (73) 30% 95% 52%
Maricopa (527) 11 77 33
Pima (176) 16 82 60
Yavapai (69) 20 87 16
Percentage of total (845) 14 80 39
1 Number of cases in the County’s sample is indicated in parentheses.
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Table 8
Adult Probation Outcomes Study
Sample1 of Probationers by County
Type of Crime Causing the Probation by Percentage
Type of Crime
Felonies
County
Class
2
Class
3
Class
4
Class
5
Class
6
Class Unknown
or
Undesignated Misdemeanor
Cochise (73) 7% 3% 19% 7% 21% 11% 32%
Maricopa (527) 4 9 28 8 32 0 19
Pima (176) 8 7 17 6 8 51 3
Yavapai (69) 0 3 8 2 25 49 13
1 Number of cases in the County’s sample is indicated in parentheses.
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
a-v
Table 9
Adult Probation Outcomes Study
Sample1 of Probationers by County
Percentage of Sample with Prior Supervisions, Custodies,
Misdemeanors, and Felonies
and Average Length of Probation Sentence in Months
Prior Actions
County Supervisions Custodies Misdemeanors Felonies
Average
Sentence
Length
Cochise (73) 34% 26% 47% 15% 39
Maricopa (527) 44 42 67 29 40
Pima (176) 42 30 57 11 36
Yavapai (69) 35 39 64 19 32
Average for the
total 42 38 63 23 38
1 Number of cases in the County’s sample is indicated in parentheses.
Source: Auditor General staff analysis of sample of records of probationers who began probation in 1994 in
Cochise, Maricopa, Pima, and Yavapai Counties.
Critical Data Elements
Collected During the File Reviews
Data was gathered in seven basic areas: 1) demographics, 2) criminal history, 3) conditions
of probation, 4) criminal behavior while on probation, 5) substance abuse treatment while
on probation, 6) life events while on probation, and 7) probation outcomes including pay-ments
of fees, fines and restitution, completion of community service, and probation status.
A sample of files from Pima and Maricopa Counties was reviewed prior to developing the
data collection strategy.
1. Demographics—The following data was sought in each probation file: date of birth, eth-nicity,
marital status, level of education, income, and employment. Ethnicity was used as
both a categorical variable and as a recoded binary variable of minority and non-minority.
Education was also recoded into number of years of education completed. In
addition, marital status was used as a categorical variable and was also recoded into a
binary variable of married and not married. State identification numbers and social secu-rity
numbers were also recorded. More than one social security number was recorded
where necessary.
a-vi
2. Criminal history—data was collected to include total number of prior felony convic-tions,
prior misdemeanor convictions, prior supervisions including parole and proba-tion,
prior incarcerations1, and prior juvenile offenses. The severity of the crime sen-tenced
for was also coded for felony or misdemeanor and class of offense from 1 through
6. Need and risk scores as measured by need and risk assessments used by the counties
were also recorded. The need and risk scores summarize demographics and criminal
history. The total possible need and risk scores were also recorded to account for varia-tions
across counties. Concurrent probation with the same county, other counties, cities,
other states, and federal was also recorded.
3. Original terms of probation—and additional and revised conditions of probation were
coded. Any conditions of probation on the basic conditions of probation forms used by
courts, and which could vary, were recorded. In addition, comment fields allowed for
special conditions not appearing on the standard forms to be recorded. Total amounts of
fees, fines, restitution, and community service ordered were recorded. Most conditions
were ordered as binary variables. Conditions that were added or subtracted after the
original terms were also coded.
4. Data was collected on criminal justice system interaction while on probation—Dates of
arrests were recorded. Up to six arrests could be recorded. Initially it was expected that
NCIC codes would be available for new arrests, but this data was generally not available
and is not useable in the data set. In addition, while conviction data were also expected
to be gathered, these data were not available. The number of and length of incarcerations
while on probation was recorded. Dates of petitions to revoke probation were recorded
along with the reason for the petition and the outcome of the petition.
5. Data on life events—such as marriage, divorce, loss through death of a loved one, and
having children was also recorded. Stability in employment was also estimated using an
ordinal level scale.
6. Attendance and completion of substance abuse treatment program(s) was recorded—
Substance abuse treatment information included beginning date, ending date, type of pro-gram
(i.e., residential or outpatient), and outcome from the program (failure, gradua-tion).
Up to four treatment programs while on probation were recorded. In addition, the
results of drug testing were recorded. The total number of tests given and the total num-ber
positive were recorded along with the substance(s) found on the first and last posi-tive
test.
7. Data on probation outcomes were recorded—The reasons for the probation termination
were collected. Outcome was recoded into an ordinal level variable (see discussion in
Appendix B, pages b-i through b-iv) and further recoded into a binary variable. In addi-
1 Consistent with the files, any incarceration is counted. Using this method, one night in jail is equivalent to 10
years in prison.
a-vii
tion to their probation outcome, outcomes specific to completing community service and
paying fees, fines, and restitution were recorded. It was necessary to construct ordinal
levels of measurement for these outcomes since files did not have complete information
in these areas. Dates specific to benchmarks were recorded including date probation be-gan,
date probation was due to end, date probation actually ended, and new date pro-bation
was due to end if an extension was made. These dates, along with date of birth,
dates of arrests, petitions to revoke, and substance abuse treatment dates made it possi-ble
to compute a variety of time intervals. Comment fields also allowed for the recording
and coding of information not routinely collected.
a-viii
(This Page Intentionally Left Blank)
Appendix B
(This Page Intentionally Left Blank)
b-i
APPENDIX B
Methods
A variety of analytic methods have been used in the report. Discussion of the specific meth-ods
is not included in the body of the report. A brief summary of the statistical techniques
employed for each finding is provided below.
Results are referred to as significant if the relationship is found to be statistically significant.
The significance level is the probability that a statistical result as extreme as the one ob-served
would occur if there were no differences between or among groups. A probability of
.05 or less is considered statistically significant throughout this report.
Finding I
Analyses reported in Finding I (see pages 11 through 17) were conducted on data weighted
for population by county. The 845 cases were weighted so that each of the four counties was
represented in the sample in direct proportion to the number of cases they represent in the
probation population for these four counties. Weighting allows for generalization of the
findings across the four counties included in the sample.
The results reported in Finding I are based primarily on a series of one-way analysis of vari-ance.
The independent variables used for the analyses of variances were identified through
a statistical modeling of the data. The first step in the modeling was a factor analysis that
resulted in four main factors identified. Variables with large weights in the factor analysis
were selected for further analyses and variables with lower weights were eliminated thus
limiting both the numbers of variables analyzed and reducing multi-colinearity. The inde-pendent
and dependent variables do not meet stringent data requirements of linear regres-sion.
Logistic regression is used for analysis of dichotomous dependent variables; however,
it also requires stringent data requirements for independent variables. These variables were
then used to develop both linear and logistic regression models. The methods were both
used for exploratory purposes and to identify variables that were predictive using both
methods.
The results of the two regression models were used to identify the factors that appeared to
have the most impact on probation outcomes and were subsequently included in the one-way
analyses of variance.
n Statistical Model—Both linear and logistic regression models were used to estimate
how well probation outcomes could be predicted.
b-ii
The final standardized model has an R squared of .54. The standardized coefficients for the
variables that enter the model and the R squared change for each are:
Variable Beta R Squared Changed
Total petitions to revoke
filed
-.559 .312
Completed community
service
.313 .129
Total treatments com-pleted
while on proba-tion
.253 .069
IPS as a condition of pro-bation
-.163 .023
Ethnicity .135 .018
Since the reliability of the model is limited by the data, a logistic regression model was also
utilized. Logistic regression is useful for situations in which one wants to predict the pres-ence
or absence of a characteristic or outcome based on values of a set of predictor variables.
It is similar to linear regression but is suited to models where the dependent variable is di-chotomous.
In the Logistic Model the dichotomous outcome variable of 0 = unsuccessful and 1 = suc-cessful
was used as the dependent measure. This variable excludes all probationers who are
still active and those with unknown final outcomes.
In the logistic model, without any predictor variables it is possible to predict with 61.45 per-cent
accuracy the dichotomous probation outcome.1 At best, the accuracy is increased to
64.08 percent if knowledge about ethnic group and the probationer’s risk score is known.
However, the predictive accuracy increases to 87.67 percent using only one condition of
probation variable and 4 activity variables and no background variables. This model pro-vides
an improvement in accuracy of 26.22 percentage points. The independent variables in
the logistic models are 1) whether the probationer is assigned to IPS as a condition of proba-tion,
2) total number of petitions to revoke filed, 3) monthly arrests, and 4) number of treat-ments
completed while on probation and consistency of employment.
Using SPSS we estimated a stepwise linear model. Variables enter the linear regression if
they are statistically significant at the .05 level, and are deleted in a step if their significance
falls below the .01 level. The linear regression model is used to i