Page 1
THE AMERICAN JOURNAL OF DRUG AND ALCOHOL ABUSE
Vol. 29, No. 3, pp. 647–670, 2003
Factors Associated with Probation Officers’ Use
of Criminal Justice Coercion to Mandate
Alcohol Treatment
Douglas L. Polcin1,* and Thomas K. Greenfield2
1Haight Ashbury Free Clinics, Inc., San Francisco, California, USA2Alcohol Research Group, Berkeley, California, USA
ABSTRACT
Alcohol problems are widespread among individuals in county criminal
justice probation systems. However, it is unclear why only a small
fraction of these problem drinkers receive treatment. In this study, self-
administered questionnaires were mailed to 145 probation officers in
nine California counties to identify factors that predicted probation
officers’ use of coercion to mandate alcohol treatment. The questionnaire
measured characteristics of probation officers, characteristics of their
caseloads, and perceptions about their departments. Principle compo-
nents analysis combined some of the items into six factor-based scales.
*Correspondence: Douglas L. Polcin, Haight Ashbury Free Clinics, Inc., 612 Clayton
St., San Francisco, CA 94117, USA; E-mail: [email protected] .
647
DOI: 10.1081/ADA-120023463 0095-2990 (Print); 1097-9891 (Online)
Copyright D 2003 by Marcel Dekker, Inc. www.dekker.com
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
B980
Page 2
Multiple regression analyses identified two significant predictors of use
of coercion into treatment: a belief that treatment was effective and a
belief that one’s peers in the department were using coercion frequently.
Implications for increasing treatment entry of probationers with drinking
problems include educating probation officers about the effectiveness of
substance abuse treatment in general and about coerced treatment in
particular. Probation departments are encouraged to develop manage-
ment styles that facilitate shared normative beliefs about assessing and
managing alcohol problems among probationers.
Key Words: Coercion; Criminal justice; Alcohol; Probation; Mandated
treatment.
INTRODUCTION
Numerous studies indicate that alcohol problems are common among
individuals in the criminal justice system (1–12). For example, Greenfield
(2,3) investigated 413 California prison inmates using the Diagnostic
Interview Schedule (13) and found that 55% had a lifetime prevalence of
alcohol abuse or dependence. Similarly, proportions for lifetime alcohol
abuse or dependence among inmates in Michigan prisons were reported to be
47% (8). Kermani and Castaneda (5) reviewed the literature on psychoactive
substance use in forensic psychiatry and cited evidence from the National
Institute of Justice (7) that incarcerated inmates had a history of drinking
alcohol on a daily basis three times as frequently as the general population.
In a study specifically focused on adult probationers, Mumola and
Bonczar (6) surveyed a national representative sampled of 2000 individuals
and found that more than 20% were on probation for driving under the
influence (DUI), 25% of the non-DUI probationers had been drinking at the
time of their arrest, and 41% had received treatment for alcohol abuse. In
another study focused on probationers, Blevins, Morton, and McCabe (1)
used the Michigan Alcoholism Screening Test (MAST) and found that
60% of a sample of 133 adult probationers had, at a minimum, problem
drinking tendencies.
The relationship between alcohol problems and crime is also evident
from general population surveys. For example, in a study that controlled
for demographic characteristics, Greenfield and Weisner (4) surveyed
county and national samples and found that lifetime drinking problems
strongly predicted current criminal behavior, arrests, and convictions. The
National Institute on Alcohol Abuse and Alcoholism summarized the role
of alcohol in specific crimes and reported that the misuse of alcohol was
involved in a large proportion of illegal activities (9). Studies suggest that
648 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 3
alcohol plays a role in 50% of homicides, 52% of rapes and other sexual
assaults, 48% of robberies, 62% of assaults, and 39% of motor vehicle
deaths (9).
Despite the role of alcohol in illegal behavior and criminal justice
incarceration, few alcohol abusers in the criminal justice system receive
treatment. In the general U.S. population, only 1 in 10 individuals with
alcohol problems receives any form of treatment (9). Treatment entry rates
among individuals in criminal justice are also low. Tam, Schmidt, and
Weisner (11) studied a sample of one county’s criminal justice system and
found that only 12% of the problem drinkers reported any contact with an
alcohol treatment program over the past year.
Detecting those individuals most in need of services may be prob-
lematic as well. Polcin and Weisner (14) found that clients entering alcohol
treatment programs who reported receiving coercion from the criminal
justice system tended to have lower levels of alcohol problems than clients
who received coercion from family members, friends, work supervisors, or
health care professionals. This raises the question about whether clients in
the criminal justice system with the most serious alcohol problems are
being detected and referred. It could be that the social and legal context of
drinking (e.g., drinking while driving) rather than the severity of alcohol
problems is the critical variable in determining who is coerced into treat-
ment from the criminal justice system.
At the same time that studies are showing underutilization of treatment
among probationers, research has documented substantial reductions in cri-
minal justice costs when individuals use alcohol and drug abuse treatment
(e.g., Refs. (15,16)). In a comprehensive meta-analysis of the DUI lit-
erature, Wells-Parker and colleagues (17) documented an 8–9% reduction
rate in DUI recidivism as a result of mandated referral to treatment.
Comparisons of voluntary vs. coerced outcomes are more common among
drug than alcohol treatment studies (18). However, research that has exam-
ined outcomes for heterogonous groups of drug and alcohol clients suggests
that coerced treatment is about as effective as voluntary treatment (16,19).
Thus it is important to identify the mechanisms by which treatment uti-
lization might be increased among the probation population, such as the
role of coercion in facilitating treatment entry.
In the current literature on criminal justice coercion to alcohol treat-
ment, most studies have examined the role of alcohol in illegal activities,
the prevalence of alcohol problems among the criminal justice population,
or the effectiveness of various treatment efforts. Few studies have examined
why individuals with alcohol problems in the criminal justice system re-
ceive little or no treatment. In particular, the existing literature has not
addressed how characteristics of criminal justice personnel or the institutions
Coercion to Mandate Alcohol Treatment 649
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 4
in which they work affect referral practices or the use of coercion to
facilitate treatment utilization.
The primary purpose of this study was to examine probation officers’ use
of coercion when referring individuals to alcohol treatment and to assess how
individual probation officer characteristics, caseload characteristics (i.e.,
prevalence of alcohol problems and alcohol-related arrests), and perception
about organizational factors in the probation departments predict the use of
coercion. Our study was conceptually rooted in previous alcohol treatment
entry studies conducted at the Alcohol Research Group (e.g., Refs. (14,20)).
Coercion to treatment for drug problems was not assessed and is an area in
need of further study. Coercion was defined as using threats of legal
consequences if individuals refused to comply with a referral to treatment.
METHODS
Procedures
Sixteen chief probation officers in California were contacted and in-
vited to take part in the study. The selection of departments that we
contacted was based on geographic proximity to the San Francisco Bay
Area and the number of potential subjects in them. Nine departments in
Northern California agreed to participate and the officers in these de-
partments were mailed a 13-page self-administered questionnaire, along
with a self-addressed stamped envelope to return it. The survey took ap-
proximately 20–30 minutes to complete. A follow-up reminder letter was
mailed in approximately 2 weeks that encouraged officers to complete the
questionnaire if they had not yet returned it. Two subsequent remailings of
the questionnaire to nonrespondents were conducted.
As an incentive to participate, probation officers were paid $20 for
completing the survey. Most departments required them to complete the
survey on their own time, outside of work. One department prohibited ac-
ceptance of any incentive. In that department, probation officers who com-
pleted the survey indicated a charity, to which the $20 incentive was donated.
Sample
Self-administered surveys were mailed to 226 probation officers in 9
Northern California county probation departments. The counties varied in
size and included a mix of urban (n = 2) and suburban or rural (n = 7) areas.
To be included in the study, officers needed to be adult probation officers
and carry an active caseload of probationers whom they monitored. After
follow-up efforts (see Procedures) 145 surveys were returned for a response
650 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 5
rate of 64%. Fifty-three percent of the sample were women and the racial
distribution consisted of 54% white, 21% Hispanic, 13% African-American,
and 11% other. The mean age was 40, and 43% of the respondents in-
dicated they were married.
Self-Administered Survey
The questionnaire consisted of 62 items, which were divided into four
sections (see Appendix A). Most of the items were rated on 5-point Likert
scale, although a number were categorical. Many items were taken or adapted
from previous surveys used at the Alcohol Research Group (e.g., Ref. (20)).
The first section of the survey addressed caseload characteristics, the
practice of coercion to mandate treatment (which specified that the
probation officer used the threat of legal consequences if the probationer
did not comply with the referral), and referral to treatment (which omitted
the threat of legal sanctions for noncompliance). Examples of coercion
items included asking respondents how often during the last 12-month they
had used coercion 1) for probationers who were arrested for offenses where
alcohol played a role and 2) when the probationer had not been arrested for
an offense where alcohol played a role, but whom the probation officer
believed had a drinking problem. These same questions were also presented
using the term ‘‘referral’’ in place of ‘‘coercion’’ (for verbatim wording,
see Appendix).
We also presented respondents four brief vignettes in which the
probation officers were asked how often they would mandate alcohol
treatment for the probationer described in the vignette. Vignettes varied
along three dimensions: 1) whether the offender was arrested for an offense
in which alcohol played a role, 2) the offender’s level of motivation for
treatment, and 3) the severity of the alcohol problem. We also presented
probationers the same vignettes and asked them how often they would
‘‘refer’’ the probationer for treatment. (These items omitted the term
coercion [i.e., the threat of legal consequences]).
The second section of the survey addressed probation officer per-
ceptions about a variety of characteristics of their department and the larger
criminal justice system. For example, subjects were asked about the
adequacy of treatment resources, whether they felt supported by judges and
state laws, their views about the disposition of probationers who were
noncompliant with coerced treatment, their views about whether their peers
used coercion, and whether their peers supported using coercion.
The third section addressed probation officers’ beliefs about proba-
tioner resistance to addressing alcohol problems, the effectiveness of
treatment, the effectiveness of coerced treatment, and their own sense of
self-efficacy in dealing with alcohol problems among probationers.
Coercion to Mandate Alcohol Treatment 651
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 6
The final section assessed demographic characteristics of probation
officers, their own drinking during the past year, their experience with
recovery services (e.g., attended Alcoholics Anonymous, received treatment
for an alcohol problem, worked in a substance abuse treatment program),
whether they and whether anyone in their current family or family of origin
had experienced a drinking problem, and the degree of effect that drinking
problems in the family had on them. Principal components analyses were
used to identify factor-scales such as coercion, used as the dependent
variable in the regression analyses (see Results).
Data Analysis
Principle components analysis (PCA) was used to combine items into
common constructs being measured (Table 1). Bivariate analysis included
t-tests to assess the relationship between coercion and categorical variables
and Pearson correlations to assess the relationship between coercion and
continuous variables. Two multiple regression models taking coercion as the
dependent variable were estimated to investigate the predictive utility of a
number of variables, others controlled. The first multiple regression assessed
probation officer and caseload characteristics and the second assessed
variables related to probation officer perceptions about their departments.
Bivariate analysis (t-tests and Pearson correlations) did not result in signi-
ficant relationships between coercion and demographic characteristics; there-
fore, demographics were not entered into the regression models (see results
in the following section). Two separate regressions were conducted because
we were interested in a variety of probation officer and departmental factors
among a limited sample size.
Table 1. Principle components analysis combining items into scales.
Scale name Items Cronbach’s alpha
Coercion Part 1, 7 items
(6, 8, 10, 12a, 12b, 12c, 12d)
0.77
Referral Part 1, 7 items
(5, 7, 9, 11a, 11b, 11c, 11d)
0.64
Family drinking problems Part 4, 3 items (13–15) 0.59
Recovery experience
(AA or treatment)
Part 4, 2 items (8, 9) 0.59
Belief about treatment
effectiveness
Part 3, 7 items (10–16) 0.73
Self-efficacy Part 3, 4 items (1–3, 5) 0.55
652 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 7
RESULTS
Caseload Characteristics
Probationer caseload characteristic varied considerably. Responses to
questions on probation officer perceptions of caseload size, percent of
caseload having a drinking problem, percent arrested for offenses where
alcohol played a role, and percent arrested for driving under the influence
all had broad distributions. Because of these ranges, medians are reported:
caseload size, 74 probationers; percent of caseload having a drinking
problem, 27%; percent of caseload arrested for an offense in which alcohol
played a role, 19%; and percent arrested for driving under the influence, 5%.
Measures
Table 1 shows characteristics of the factor-based scales derived from the
principle components analysis. The table indicates the scale name (selected
based on conceptual content) and gives the number of items loading on the
scale, which questions loaded on each scale (Appendix A provides specific
questions), and Cronbach’s alpha coefficients for each scale. The six scales
had adequate internal reliability for research purposes, especially three of the
scales each with seven items—coercion with an alpha of 0.77, referral with
an alpha of 0.64, and beliefs about the effectiveness of treatment at 0.73.
Other scales showed only marginal reliability because, in part, of their
smaller number of items (e.g., self-efficacy with four items and an alpha of
0.55). In no case did removal of an item increase internal reliability.
In general, the specific items that loaded on the coercion scale tended
to support the probation officer’s use of the construct ‘‘coercion to or
mandating treatment.’’ The level of support for coercion, though fairly
high, varied based on the specifics of the situation that probation officers
were presented with in the questions. For example, when simply asked how
often they used coercion for individuals arrested for an offense where
alcohol played a role, 51% indicated they used coercion to mandate
treatment all or most of the time. However, when additional information
was added in vignettes presented to them, the result differed. An example is
an item asking the respondent how often they would mandate treatment for
a person who was arrested for an offense where alcohol played a role, there
appeared to be a serious drinking problem, and the person was highly
resistant to treatment, 89% indicated they would coerce such a person to
treatment all or most of the time.
Probation officers also often made referrals to treatment without the
threat of legal coercion. For example, when asked how often they referred
individuals to treatment who were arrested for an offense in which alcohol
Coercion to Mandate Alcohol Treatment 653
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 8
played a role, 93% indicated they made a referral to treatment all or most
of the time, contrasting with 51% using coercion to mandate treatment on
the equivalent item. In response to the vignette in which the probationer is
described as requesting alcohol treatment, when not arrested for an alcohol
offense and not appearing to have an alcohol problem, 79% of the re-
spondents indicated they would refer such a person to treatment all or most
of the time, whereas only 25% indicated they would use coercion all or
most of the time. Table 2 summarizes the comparison of probation officers
preferences regarding ‘‘referral’’ vs. ‘‘coercion’’ to treatment on the four
vignettes presented to probation officers. It can be seen that in every
vignette the likelihood of referring probationers to treatment was higher
than using coercion to mandate treatment.
Items loading on the family drinking problems scale asked whether
anyone in the respondents current or family of origin had a drinking problem,
how many such family members were (measured as a count), and the impact
that these relative’s drinking had on the probation officer’s life (measured on
a 4-point scale from none to great). Seventy-nine percent of the respondents
indicated that they had at least one problem drinker in their family and 48%
indicated that the drinking had a moderate or great impact on their life. This
scale had a negative correlation with the probation officer’s frequency of
drinking over the past year (r = –0.21, P < 0.05), a positive correlation with a
scale measuring the respondent’s experience with recovery services (as a
client, staff, or member of Alcoholics Anonymous) (r = 0.28, P < 0.01), and a
positive correlation with the belief that treatment is effective (r = 0.22,
P < 0.05). Although a variety of explanations are possible, these findings no
doubt reflect the influence of family of origin on probation officers’ personal
histories with alcohol use, dependence, and treatment. Those in recovery
from alcohol problems may come from more alcohol-involved families or at
least be more aware of family members’ alcohol problems, and maybe more
likely to score higher on this measure than others. Their own efforts to
establish and sustain recovery would result in more experience with recovery
services as well as less frequent drinking over the past year.
A factor-based scale measuring beliefs about treatment effectiveness
comprised seven items. Taken together, these items revealed a moderate
belief that treatment is effective. For example, 71% indicated that treatment
for alcohol problems was effective ‘‘some of the time,’’ and 61% believed
that coerced treatment was effective ‘‘some of the time.’’ Similarly, 68%
indicated that coerced treatment was effective in reducing criminal behavior
‘‘some of the time.’’ Despite mixed views about the effectiveness of
treatment, a majority of respondents indicated that treatment was worth
trying. For example, 70% indicated that treatment was valuable even for
highly resistant clients, and 63% indicated that coerced treatment helped
facilitate probationers’ entry into alcohol treatment ‘‘all’’ or ‘‘most’’ of the
654 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 9
time. Sixty-two percent indicated that coercion improved client attendance
‘‘all’’ or ‘‘most’’ of the time and 92% indicated that that coercion helped
avoid treatment dropout at least ‘‘some of the time.’’
A final scale loaded four items into a measure of self-efficacy. In
general, respondents felt confident in their abilities to deal with alcohol
problems. For example, 65% agreed with a statement that ‘‘probation
officers can help probationers reduce alcohol consumption.’’ Seventy-nine
percent disagreed with the statement, ‘‘I feel I do not have much to offer
when working with problem drinkers.’’
Several variables measuring probation officer views and personal char-
acteristics used single items rather than scales for measurement. Examples
include items measuring perceptions of client resistance and frequency of
alcohol consumption during the past year. On the measure of client
resistance, 63% of the respondents disagreed with a statement that dis-
cussing alcohol problems with probationers increases resistance. On the
Table 2. Comparison of probation officers’ referral vs. coercion of probationers to
treatment on four case vignettes that differed by motivation, level of alcohol problem,
and whether alcohol played a role in the offense (means and standard deviations based
on a 5-point likert scale).
Characteristics of the
vignette and items used Referral to treatment Coercion to treatment
Highly motivated
not perceived a problem
drinker alcohol did not play
a role in arrest (Part 1,
items 11a and 12a)
1.72 (1.15) 3.72 (1.54)
Resistant to treatment not
perceived a problem
drinker serious offense,
alcohol played a role
(Part 1, items 11b and 12b)
1.94 (1.15) 2.58 (1.32)
Highly resistant to treatment
perceived to have a moderate
drinking problem alcohol
did not play a role in arrest
(Part 1, items 11c and 12c)
2.53 (1.06) 2.89 (1.23)
Highly resistant to treatment
perceived to have serious
drinking problem alcohol
played a role in arrest
(Part 1, items 11d and 12d)
1.39 (0.82) 1.61 (1.01)
Note: Lower scores reflect more referral and coercion.
Coercion to Mandate Alcohol Treatment 655
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 10
frequency of alcohol consumption during the past year, nearly half (49%)
indicating they drank 2–3 times per month or less.
Perceptions About Probation Departments and
Criminal Justice Systems
The variables in this section assessed a range of issues relevant to
respondents’ views about their department and the larger criminal justice
system. Unlike some of the measures described previously, which combined
items into scales, variables here were assessed using single items. Included
were issues such as assessment procedures, adequacy of treatment resources,
the extent to which their peers parole officers use and support coercion, and the
extent to which they feel supported by judges’ rulings, state laws, and sanctions
for probationer noncompliance when they use coercion to mandate treatment.
Eighty-three percent of the respondents indicated that their departments
did not use structured questionnaires to assess substance abuse problems.
However, 71% of the respondents felt that outside programs that assess
alcohol problems among their probationers were accurate at least most of
the time.
There were mixed views about the adequacy of treatment resources and
the spectrum of service levels available. Fifty-six percent disagreed or were
undecided in response to an item stating the department had adequate al-
cohol treatment resources available to probationers. Similarly, 54% dis-
agreed or were undecided about whether the resources available represented
an adequate mix of inpatient, residential, and outpatient services. Probation
officers indicated that the most common sources for coerced referrals were
to outpatient programs (43% of coerced referrals) and specialty driving
under the influence programs (24% of coerced referrals).
Probation officers tended to feel supported by others in the criminal
justice system for using coercion (e.g., expressing the view that their peers
supported the use of coercion to get probationers into treatment [71%
agreement]). They also felt that their peers used coercion a moderate
amount to coerce probationers into treatment, with 88% indicating their
peers used coercion to mandate treatment at least some of the time. Sixty-
four percent felt that judges’ rulings and 59% felt that state laws either
supported or strongly supported their ability to coerce probationers into
treatment. When probationers failed to comply with mandated treatment,
61% felt the sanctions were about right.
Bivariate Analysis
For dichotomous variables, we used t-tests to assess whether there were
differences in use of coercion to mandate treatment. No significant differences
656 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 11
were found by gender, race (dichotomized as white vs. other), marital status
(married vs. other), or department use of formal screening instruments (yes/
no). Pearson correlation did not find a relationship between coercion and age.
To assess whether there were differences in how these variables were
associated with ‘‘referral’’ to treatment as opposed to the use of coercion, we
ran the same analyses using the ‘‘referral score’’ as the dependent variable
measure. No significant differences were found. To assess whether the nine
different departments differed in their use of coercion, an analysis of variance
F-test was conducted assessing differences in the coercion score. No
differences in use of coercion were found among the nine departments.
The relationship between coercion and other continuous variables were
assessed using Pearson correlations. Not surprisingly, there was a sig-
nificant relationship between coercion scores and referral scores (r = 0.34,
P < 0.001). In terms of individual probation officer and caseload
characteristics, there was a significant correlation between the beliefs about
treatment effectiveness score and coercion (r = 0.37, P < 0.001). In terms of
perceptions about the probation department and larger criminal justice
system, there was a significant correlation between use of coercion and the
belief that ones peers used coercion (r = 0.37, P < 0.001).
Multivariate Analysis
To assess whether the above correlations could predict coercion, con-
trolling for the influence of other variables, two multiple regression
analyses were conducted. The first assessed whether probation officer and
caseload characteristics could predict the use of coercion. Table 3 indicates
the variables entered into model and the results.
The overall model entering all seven variables resulted in an F = 2.44,
P < 0.05, and adjusted R-square = 0.083. Examination of the influence of
individual variables shows that the strongest predictor was the belief about
treatment effectiveness variable (t = 3.74, P < 0.001), R-square = 0.116. Al-
though beliefs about treatment effectiveness accounted for nearly 12% of
the variance of coercion, caseload and other probation officer characteristics
were not significant.
The second multiple regression analysis assessed variables related to
probation officer perceptions about their departments and the larger
criminal justice system. The overall model with seven variables entered
resulted in an F = 2.34, P < 0.05, and R-square = 0.077. The belief that
one’s peers used coercion was the strongest predictor, accounting for 9% of
the variance in the model (Table 4).
To assess whether different factors might predict referral to treatment
as opposed to coercion, we ran the same regressions using the referral to
treatment score as the dependent variable (not shown). The overall model in
Coercion to Mandate Alcohol Treatment 657
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 12
the first regression, which assessed caseload and individual probation
officer characteristics, resulted in a very modest trend: F = 1.80, P = 0.096,
adjusted R-square = 0.05. Beliefs about the effectiveness of treatment was
again the only significant predictor, accounting for 6% or the variance
(t = 2.64, P < 0.01). The second regression, assessing probation officer
views about their department and the larger criminal justice system, was
Table 3. Multiple regression analysis of probation officer and caseload
characteristics predicting use of coercion to treatment.
Variable
Parameter
estimate
Standard
error t-value P R-square
% Problem drinkers –0.007 0.009 –0.77 0.442
% Arrested for
alcohol offenses
0.006 0.008 0.80 0.423
Beliefs about resistance 0.042 0.175 0.24 0.811
PO drinking past 12 months 0.012 0.083 0.14 0.886
Family drinking problems 0.025 0.131 0.19 0.857
Experience with
recovery services
0.042 0.136 0.31 0.758
Beliefs about
treatment effectiveness
0.375 0.100 3.74 0.0003 0.116
Note: The model with all seven variables entered resulted in F = 2.44, (P < 0.05),
adjusted R-square = 0.083.
Table 4. Multiple regression analysis of probation officer perceptions of probation
departments and criminal justice systems predicting use of coercion to treatment.
Variable
Parameter
estimate
Standard
error t-value P R-square
Department uses
assessment instruments
0.196 0.407 0.48 0.631
Peers support coercion 0.056 0.158 0.35 0.724
Peers use coercion 0.629 0.191 3.30 0.001 0.089
Judges effect 0.007 0.184 0.04 0.969
State law effects 0.139 0.207 0.67 0.504
% Noncompliant 0.002 0.004 0.60 0.549
Severity of sanctions 0.130 0.269 0.48 0.630
Note: The model with all seven variables entered resulted in F = 2.34, (P < 0.05),
adjusted R-square = 0.077.
658 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 13
not significant with all variable entered and did not have any individual
variables that predicted referral.
DISCUSSION
The results of the study suggest that probation officers in general
support the use of coercion to mandate treatment for alcohol problems.
However, they seem to weigh a number of factors when judging whether to
make a coerced referral. These include: 1) whether the offender was arrested
for an offense where alcohol played a role, 2) the offender’s level of
motivation for treatment, and 3) the severity of the alcohol problem.
The results suggest that probation officers frequently refer probationers
to treatment without coercion. Although coercion and referral scores
were highly correlated, there were important differences. In some vignettes
where probation officers were unlikely to coerce probationers to treatment,
they were quite likely to make a referral. Making referrals to treatment that
are not relevant to the probationer’s legal status suggests that probation
officers often take a human services view of their work, trying connect
probationers with services that they need irrespective of the relevance to
probation status.
Probation officers were also moderate in their beliefs that probationers
could be helped in treatment. Very likely they have some subjective sense
of when and for whom treatment is effective. Although our study did not
access the conditions under which probation officers believed that treatment
could be effective, this may be an interesting area for further study.
Findings suggest that probation officers generally feel pleased with the
response that their departments and the larger criminal justice system has in
regards to their use of coercion to mandate treatment. More often than not,
they report feeling support for their use of coercion from judges, peers, and
state laws. They were also generally pleased with the quality of assessments
made by programs in their communities. However, there appeared to be
some concern about the availability of treatment resources and the level of
services offered.
The majority of potential predictors of coercion were markedly un-
related to it. Predictors of referral were even weaker. This serves to
highlight the two predictors that were significant: beliefs that treatment is
effective and perception of peers use of coercion. The former is clearly not
surprising; probation officers who believe treatment is effective are more
likely to mandate it, especially if they believe mandated treatment is
effective, which is one of the items that loaded on the beliefs about treat-
ment effectiveness score. The second predictor is more complicated. It is a
Coercion to Mandate Alcohol Treatment 659
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 14
bit like the adolescent finding that teenagers who believe their peers are
using drugs are more likely to use drugs themselves, irrespective of the
accuracy of their belief. Here, probation officers who believe their peers are
using coercion are more likely to use it themselves.
Because probation officers were asked about ‘‘their peers in the de-
partment’’ and the nine departments did not differ in their use of coercion,
it appears that the perception about use of coercion is key, whether or not
the perception is accurate, rather than perceptions of actual differences in
use of coercion. If departments actually differed in their use of coercion it
would make sense that respondents from higher use of coercion
departments would report more frequent practice of coercion among their
peers. This was not the case in our study.
However, one limitation relevant to this finding is that respondents may
not have actually thought about the overall department when asked about
‘‘peers in the department.’’ Instead, they may have thought about those
probation officers whom they worked closest with on a daily basis and
disregarded the overall department. In this scenario, there may be subgroups
within departments that actually differ in their use of coercion. Respondents
may have been reporting these real subgroup differences rather than un-
substantiated normative beliefs. Disentangling these differences more def-
initively would require additional research.
Other limitations of the study include the limited geographical area and
the lack of assurance that our sample was representative of probation
officers in California. However, as we noted in the results, use of coercion
and referral to treatment did not differ among demographic subgroups or
different departments, some of which were urban (n = 2), whereas others
were suburban or rural (n = 7). It should be remembered that our study
focused on coercion to treatment for alcohol problems and coercion to
treatment for drug problems could yield different results.
There are two implications from our findings for increasing the pro-
vision of services to the large number of probationers who have alcohol
problems. First, facilitating the belief that treatment is effective may help to
increase coercion to treatment. Several strategies are suggested.
First, managers of probation departments should expose probation of-
ficers to the extensive literature documenting the effectiveness of alcohol
treatment in general and coerced treatment in particular (e.g., Refs. (15–
17,21–26)). The National Institute on Alcohol Abuse and Alcoholism
publishes treatment reviews regularly in provider friendly publications such
as ‘‘Alcohol Alert.’’ Publications within the probation field should make
dissemination of alcohol treatment outcome findings a top priority.
Managers could also facilitate presentation of successful cases in pro-
bation officer meetings to expose officers to real successes within the
660 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 15
department. An excellent suggestion would be for managers to support
long-term follow-up of coerced referrals to document successes and make
modifications to improve outcome.
Managers could also adopt strategies to facilitate normative beliefs
that probation officers’ peers in the department frequently practice
coercion to treatment. Many modern theories of management in general
(e.g., Ref. (27)) and management of human service organizations in
particular (28) suggest that developing shared views of the department
among all personnel and shared responsibility for operations is critical.
The theory is that the nature of current organizational environments
necessitates that workers be empowered to help develop policies based on
a shared understanding of their organization. In this management style,
leaders of probation departments would involve probation officers in
developing normative beliefs and practices in regards to the use of
coercion. To the extent that these norms and practices supported coercion,
mandated referrals would increase.
There are a number of suggestions for further research. Although few
of the individual probation officer characteristics, caseload characteristics,
or views about probation departments and criminal justice systems were
associated with the practice of coercion or referral to treatment, these
factors may be associated with other factors relevant to coerced treatment.
For example, it would be interesting to assess whether the variables used
here were associated with probationers compliance with coerced referrals or
their success in treatment.
Because family drinking problems were so prominent among re-
spondents and apparently had a substantial impact on the lives of probation
officers, this variable deserves closer scrutiny. Family drinking problems
correlated with less frequency of drinking over the past year, a belief the
treatment was effective, and more experience with recovery services. It is
likely that many of these individuals are in recovery themselves. It would
be interesting to assess whether their work with problem drinkers differs
from other probation officers in respects other than use of coercion.
Finally, a variety of human services, medical, and legal professionals
have contact with individuals who have alcohol problems (14,20). Mea-
surement of their referral practices and use of coercion deserves more
attention in the literature. The scales that were developed to measure
coercion and referral in this study resulted in strong alpha coefficients.
They could be used to measure the disposition of problem drinking cases
from the perspective of other criminal justice personnel, such as judges
or parole officers. They might also be modified to measure the responses
to alcohol problems from social workers, health care professionals, or
work supervisors.
Coercion to Mandate Alcohol Treatment 661
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 16
APPENDIX A
662 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 17
Coercion to Mandate Alcohol Treatment 663
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 18
664 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 19
Coercion to Mandate Alcohol Treatment 665
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 20
666 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 21
Coercion to Mandate Alcohol Treatment 667
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 22
ACKNOWLEDGMENT
This work is supported by NIAAA R03 AA12692-02.
REFERENCES
1. Blevins LD, Morton JB, McCabe KA. Using the Michigan alcoholism
screening test to identify problem drinkers under federal supervision.
Fed Probat 1996; 60(2):38–44.
2. Greenfield TK. A study of major mental disorders in California prisons:
survey methodology and preliminary estimates. In: Paper Presented at
the Annual Meeting of the American Academy of Psychiatry and Law,
San Francisco, CA, October, 1988.
3. Greenfield TK. Voices from California state prisons: utilization of and
amenability to treatment. In: Paper Presented at the Annual Meeting of
the Society for the Study of Social Problems, Washington, D.C, August,
1990.
4. Greenfield TK, Weisner C. Drinking problems and self reported
criminal behavior, arrests, and convictions: 1990 US alcohol and 1989
county surveys. Addiction 1995; 90:361–373.
5. Kermani EJ, Castaneda R. Psychoactive use in forensic psychiatry. Am
J Drug Alcohol Abuse 1996; 22(1):1–27.
6. Mumola CJ, Bonczar TP. Substance Abuse and Treatment of Adults on
Probation. Special Report. Bureau of Justice Statistics: Washington,
D.C., 1998.
668 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 23
7. National Institute of Justice. Research in Action: Drug Use Forecasting.
United States Department of Justice: Washington, D.C., 1988.
8. Neighbors HJ, Williams DH, Gunnings TS, Lipscomb WD, Broman C,
Lepkowski J. The Prevalence of Mental Disorders in Michigan Prisons
(Final Report to the Michigan Department of Corrections). Institute for
Survey Research: Ann Arbor, Michigan: July, 1987.
9. National Institute of Alcohol Abuse and Alcoholism. Improving the
Delivery of Alcohol Treatment and Prevention Services: Executive
Summary. NIH Publication No. 97-4224, Bethesda, MD, 1997.
10. Schmidt L, Weisner C. Developments in alcoholism treatment. In:
Galanter M, ed. Recent Developments in Alcoholism (Vol. 11). New
York: Plenum, 1993.
11. Tam TW, Schmidt L, Weisner C. Patterns in the institutional encounters
of problem drinkers in a community human services network. Addiction
1996; 91(5):657–669.
12. Weisner C. Coercion in alcohol treatment. In: Institute of Medicine,
Broadening the Base of Treatment for Alcohol Problems. Washington,
D.C.: National Academy Press, 1990.
13. Robbins L, Helzer JE, Orvaschel H, Anthony J, Blazer D, Burnam MA,
Burke JD. The diagnostic interview schedule. In: Eaton WW, Kessler L,
eds. Epidemiology Field Methods in Psychiatry: The NIMH Epidemi-
ological Catchment Area Program. New York: Academic Press, 1985,
143–170.
14. Polcin DL, Weisner C. Factors associated with coercion in entering
treatment for alcohol problems. J Drug Alcohol Depend 1999; 54:63–
68.
15. Gerstein DR, Johnson RA, Harwood HJ, Fountain D, Suter N,
Malloy K. Evaluating Recovery Services: The California Drug and
Alcohol Treatment Assessment (Contract’ No. 92-001100). Califor-
nia Department of Alcohol and Drug Programs: Sacramento, CA,
1994.
16. Center for Substance Abuse Treatment. National Treatment Improve-
ment Evaluation Study (NTIES). US Department of Health and
Human Services: Washington, D.C., 1996.
17. Wells-Parker E, Bangert-Drowns R, McMillen R, Williams M. Final
results from a meta-analysis of remedial interventions with drink/drive
offenders. Addiction 1995; 9(7):907–926.
18. Polcin DL. Criminal justice coercion in the treatment of alcohol
problems: an examination of two client subgroups. J Psychoact Drugs
1999; 31(2):137–143.
19. American Society of Addictive Medicine (ASAM). Alcohol, illicit
drugs both factor in criminal activity. ASAM News 1999; 14:2.
Coercion to Mandate Alcohol Treatment 669
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �
Page 24
20. Weisner C, Schmidt LA. The community epidemiology laboratory:
studying alcohol problems in community and agency based populations.
Addiction 1995; 90:329–341.
21. Doyle H, Tobin J, Delaney W. A study of the legal involvement of new
patient attenders at a Dublin alcohol treatment unit. Ir J Psychol Med
1992; 9(2):93–95.
22. Institute of Medicine. Broadening the Base of Treatment for Alcohol
Problems. National Academy Press: Washington, D.C., 1990.
23. Institute of Medicine. Managing Managed Care: Quality Improvements
in Behavioral Health. National Academy Press: Washington, D.C.,
1997.
24. Polcin DL. The etiology and diagnosis of alcohol dependence: differ-
ences in the professional literature. Psychotherapy 1997; 34(3):297–
306.
25. Polcin DL. Professional therapy versus specialized programs for alco-
hol and drug abuse treatment. J Addict Offender Couns 2000; 21(1):2–
11.
26. Project MATCH Research Group. Matching alcoholism treatments to
client heterogeneity: project MATCH posttreatment outcomes. J Stud
Alcohol 1997; 58:7–29.
27. Bradford DL, Cohen AR. Managing for Excellence. Wiley and Sons:
New York, 1984.
28. Polcin DL. Administrative planning in community mental health.
Community Ment Health J 1990; 26(2):181–192.
670 Polcin and Greenfield
� ��� ����� � ��� �� � ��� �� �� ��� �� ��� �� � ��� ���� �� �� ��� � �� � ����� !�"�
# �$ ����� �� �� � �