U.S. D
epartment of Justice
Office of Justice Program
s
National Institute ofJustice
Michael J. R
ussell,Acting D
irector D
ecember 1993
Assessin
g th
e Imp
act o
f Dad
e Cou
ntv
's
The extraordinary
growth in the drug-
related criminal caseload during the 1980's
and the perceived impact of illicit drugs on
public safety in Dade C
ounty prompted
Florida's Eleventh Judicial C
ircuit to im
plement a court-based
drug abuse treat-m
ent approach. The innovation w
as guided
P'y the notion that an effective and flexible
, ,im
grarn of court-superviseddrug treat-m
ent could reduce demand for illicit drugs
and hence involvement in crim
e and reinvolvem
ent in the court system by
substance abusers.
by John S. Goldkarnp and D
oris Weiland
What has com
e to be known as the "M
iami
Drug C
ourt model" has tw
o principal com-
ponents-a nontraditional role for officials
in the courtroom and a specially adapted
program of "outpatient" drug abuse treat-
ment.' O
ther diversion approachesrefer drug defendants to treatm
ent programs, but
the courtroom-based
team approach-and
particularly the centraljudicial role---dis-tinguishes D
ade County's initiative.
Research questions
The em
pirical assessmentof the D
rug C
ourt initiative had three basic purposes:
To exam
ine the program's im
pact in the E
leventh Judicial Circuit.
To serve as a factual basis for inform
ing the C
ircuit Court and participating agen-
cies on ways to im
prove or reshape, if nec-essary, the program
in its next phases.
Felo
ny D
rug C
ou
rt
issuesand Findings
Discussed in this E
valuahahonBulk-
tin. Florida's Eleventh Judicial
Circuit in 1989
adopted a court-based approach to treatm
ent for felony drug abuse in D
ade County
(Miam
i).This research study as-
sessed the program.
Key issues: T
he treatment approach,
helping defendants function more
normally in society,can conflict w
ith the crim
inaljustice approach,for re-ducing crim
e and improving
public safety. M
easuring a program's suc-
cess is thus a policy issue for public officials to decide. E
stablishingclear
expectations and criteria for program
outcomes should be done before a
program is im
plemented and should
be modified, if necessary, on the
basis of program experience.
Majorpndings. T
he researchers focused on defendants over an 18-m
onthperiod and com
pared them
to similar defendants not in the pro-
gram. T
hey found the Drug C
ourt defendants had:
+ Fewer cases dropped.
+ Low
er incarceration rates.
+ Less frequent rearrests.
+ Longer tim
es to rearrest.
+ Higher failure-to-appearrates,
caused mainly by the m
ore frequent appearancesrequired of D
rug C
ourt defendants.
Strengthsof the Dade C
ounty Drug
Court system
included strong support
To share w
ith the comm
unity of A
merican courts lessons other court
systems could draw
from study of the
Miam
i Model.
for the Drug C
ourt from all participants
in the criminaljustice system
, an active judicial role, specially designed treat-m
ent programs, and a flexible approach
to program participants' problem
behavior.
Key challenges for a drug court
program identified by the researchers
confront the Dade C
ounty program
and have implicationsfor other
jurisdictions:
+ Need for fast, accurate inform
ation about defendants.
+ A clearly defined target population
for the program, w
hich should avoid net w
idening and help set the basis for screening criteria.
+ Need for different treatm
ent plans for different levels and types of drug abuse.
A B
rief Xistory of
the Miam
iDrug C
oufi
Since 1989,w
hen the Drug C
ourt first visits or as
mm
y days
r opened, d
efe
nh
ts have been referred to achieve 7 consecutivenegaril
primarily to the D
AW
(diversion and tests. In phase
the numkr of req
us~
u
rreament progrm
), an outpatient pro-visits w
as genemlly reduced to e
e
or
gram w
ith centers in four locations in even tw
o per week, w
ith a mine test at
Dade C
ounty. There w
as also an option each visit. D
uring phaw 111, a~
end
ance
for defendants who lived in other juris-
requ
i~m
entsm
ight continueto be The dictions to participate in @
am
en
tpra-sam
e or, given a client's progess and gram
s outside Dade C
ounty aslong as
work scheduleor school obligations,
regular reports were m
ade to the court. relaxed som
ewhat.
The D
rug Court w
as initially designed l'hree consecutive un:
l failures to accept defendm
ts charged with
to keep roquirsdclinic ilyt~
uu~unenls
at third-degree felony drug possession
any time w
ould result in the client's
offenses and no prior convictions. placem
ent in "p
hw
V"-infom
d suspension. A
client rewin
g after
The D
AT
P drug abuse rream
ent pro-such an absence
would be reinsm
ted in gram
required panicipation
whatever
he or had been in. If
drug-invo'ved felony defendant' "'"8 a client failed to appear f
ar
consecu-w
hich the defendant would proceed
cive days, DA
TP
was required in
from detoxification (phase I), to coun-
compliance
Slateregulations to seling (phase 111, to educationdl/voca-
close C
lienlswere
assessment and E
Gning (phase
comm
only after such an en-
'''), and then ''
graduation (phase lV)'
tended absence, hilt they were required
Phase I was intended to require a m
ini-la start over in phase 1.
mum
of 12consecutive days of clinic
-T
o accomplish these research aim
s, data collection for the assessm
ent focused on these areas of inquiry:
The im
pact of Drug C
ourt on criminal
processing, including its identification and enrollm
ent of defendants who other-
wise w
ould have been adjudicated in the norm
al fashion.
Com
parison of Drug C
ourt defendant case outcom
es with those of defendants
who faced charges of sim
ilar severity both before the D
rug Court initiative (during the
summ
er of 1987)and at the time of the
study (August and Septem
ber 1990).
Performance of D
rug Court defendants,
including treatment program
outcomes.
Public safety implications of the D
rug C
ourt program.
Research design. U
se of an experimental
design to study the impact of the D
rug C
ourt was precluded for practical reasons.
qu
id
ve urine
.,
The court had already been in operation al-
most 2 years, and random
allocation of de-fendants to treatm
ent and experimental
groups would have too greatly disrupted
the ongoing program. Instead, researchers
designed a next best approach that focused on (nonequivalent)com
parison groups of relevant felony defendants to help gauge the effect of the program
. These included
contemporaneous and historically anteced-
ent samples of noneligible felony drug
cases and nondrug cases.l
The initial and principal sam
ple was a co-
hort of defendants admitted to the D
rug C
ourt program in A
ugust and September
1990.This group is identified as S
ample I
(n = 326) in exhibit I. Selection of the
sample period w
as guided by two con-
cerns: a) to ensure that the study would
fairly examine the program
at a stage som
etime after its im
plementation
"in-fancy"; and b) to perm
it use of a sufficient
observation or followup period (1 8
months) for study of defendant perfor-
mance from
the point of admission to
the program.
Defining and m
easuring "success":a policy concern
The E
leventh Circuit's D
rug Court is a hy-
brid combining elem
ents of both crirnina! justice and drug treatm
ent approaches to address an im
portant portion of the drug-involved population am
ong criminal
offenders (defendants in this case). Key
elements include the special role for judge
and criminal courtroom
personnel, the fun-dam
ental treatment orientation, and the
diversion-like framew
ork.
This attem
pt to integrate disparate ele-m
ents has meant joining tw
o perspectives accustom
ed to different methods and
sometim
es competing aim
s regarding drug involvem
ent and its reduction. The result-
ing uneasy marriage of crim
inaljustice and drug treatm
ent goals embodied in the D
ade C
ounty initiative complicated design of an
empirical assessm
ent.
Adapting the courtroom
setting to assist the aim
s of treatment is not necessarily
compatible w
ith the usually more form
at and adversarial aim
s and procedures of crim
inaljustice. From the view
point of drug abuse treatm
ent, the drug court seeks to reduce drug abuse so that defendants can function norm
ally in society. From the
criminal court perspective, the program
tries to reduce the im
pact of the drug caseload on case processing resources di-verting the flow
of cases,reducing drug crim
e among participants, and thus im
-proving public safety.
In contrast, a treatment perspective w
ould probably not view
a "three striker" ap-proach to program
compliance as realistic.
Indeed, treatment staff w
ould understand that, to the extent that serious drug abusers are encouraged to enter the program
, the road to progress ir likely to be very diffi-cult, w
ith initial failures routinely to be ex-pected. T
hir difference in perspectives translates into differences in expectations fi about the perfom
lance of drug court
PExhibit 1. D
efendant-Based S
ampling S
trategy for Evaluation
of Dade C
ounty Felony "Drug C
ourt"
Potential P
opulation
kdawea
wefarw
ants
Relevan
Felony 2&
(6
Sum
mary of S
amples
I Drug C
ourt In DA
TP
(100%, n=326)
II AssignedJN
ot-In DA
TP
(100% n=89)
Ill Drug C
aseINot A
ssigned (10%
, n=199) IV
Non-D
rugC
aseINot E
ligible (5%
, n=185) V
1987 F
3 & F
2 Drug
(n=302) V
I 1987 F
3 & F
2 Non-D
rug (n=536)
Dlug
Defen.-
-
As
we
d'
In Drug
Court
5% S
ample
Pre-D
rug Court
Sam
ple 1987
F3 &
F2
F3 &
F2
Cases
(n=838) Including F
ollow-up
0th
~Cases Mot
No
te 1: Th
~scategory lncludes flve defendants
ad
m~
ttedafter the
from A
ug.-Sepf. 1990
sample p
er~
odand not Included
~nthe counts below
No
te 2: One defendantw
as admttted after the sam
ple pe
r~o
d
Aug.-Sept.
1990 N
ote 3: N
o treatment f~
lescould be found for f~
veof these
defendants reduclngthe flnal sam
ple to 326 cases Note
that the sample Includes both defendants
w~
thftltngs enterlng D
ru C
ourt and those ad
m~
ttedto treatm
ent ~n
August and le
pte
mb
er 1990
100% S
ample
No
te 4: Includessix defendants
not shown as assigned
but later determ
inedto have been targeted.
defendants and, as a result, into potentially different w
ays to measure outcom
es.
Measurem
ent of a drug program's out-
comes is, therefore, a problem
because there are a num
ber of ways to m
easure "success," all of w
hich could be valid depending on the perspective adopted. In fact, this issue-the
definition and meas-
urement of success-represents
a major
policy task faced by officials designing and operating drug court program
s. For purposes of this study, program
outcomes
were defined as "favorable" or "unfavor-
able" after discussion and debate by m
embers of the judicial w
orking group guiding the research process. T
his ap-proach w
as critical to the assessment so
that the research could avoid making
policy assumptions that m
ay not have been intended by site officials.
The D
ade County D
rug Court diversion
and treatment program
(DA
V) w
as planned to require defendant progress through three phases to an eventual favor-able outcom
e, graduation in about 1 year. Program
outcomes w
ere cataloged by re-view
ing both the treatment agency files
and the criminaljustice data m
aintained by the court system
.Wsing these sources,the
specific program outcom
es recorded for the Sam
ple I group at the end of the 18-m
onth observation period (which began
at the point of entry into treatment) in-
cluded the following types:
Unfavorable D
ropped out T
erminated
Cases still active
With no alias capiases (bench w
anants) W
ith alias capiases
Transferred
Other jurisdiction
Other local agency
Other D
ied
Charges dropped W
ithin 35 days
Graduation im
plied N
olle prossed (ceased prosecution) N
olle prossed, tracking
Exhibit 2
. Program
Outcom
es for Drug C
ourt Defendants
Adm
itted to
Treatment, A
ugust-Septem
ber 1990
Percen
t of D
rug
Co
urt
Defen
dan
ts
Unfavorable
Favorable
Transferredlother
Charges
Cases S
tillActive
Dropped
35 Days
Pro
gram
Ou
tcom
es (n
=326)
Sealed Sealed, tracking Probation only
As a first step in organizing these program
outcom
es, exhibit 2 shows 34 percent of
defendant outcomes as clearly "favorable,"
23 percent as clearly "unfavorable," and 43 percent as falling into the other categories w
hose classification was not self-evident.
To illustrate the role of policy assum
ptions in the m
easurement of success,this rough
grouping of favorable and unfavorable out-com
es could be further collapsed into more
narrowly defined categories (referred
to in this report as version 2) by applying the follow
ing assumptions:
The sm
all number of defendants w
ho w
ere transferred to other jurisdictions re-m
ained the responsibility of the Drug
Court. H
owever, one could argue that they
should also be excluded from evaluation of
treatment program
outcomes because they
became the responsibility of other agencies
or jurisdictions and, therefore, were not ap-
propriate tests of the impact of the D
rug C
ourt in Dade C
ounty.
Defendants w
ho had active or open cases at the end of 18 m
onths either should be counted as provisionally having re-corded favorable outcom
es (as long as they did not record alias capiases),or be
counted provisionally as having unfavor-able outcom
es, if they absconded from
the program and did not retum
to active participation.
Defendants w
ho dropped out because their charges w
ere dropped within 35 days
should be excluded from the analysis of
outcomes because they did not participate
in the program for a m
eaningful period of tim
e (i.e., they were "false starts").
Even this classification of program
out-com
es, however, could be further refined
by adopting yet another assumption from
the drug treatm
ent perspective:
Because som
e minim
um period of par-
ticipation should be required before it is reasonable to evaluate the im
pact of the program
on defendant behavior, all per-sons dropping out w
ithin the first 3 weeks
of admission (notjust those w
ith charges dropped)should be excluded from
out-com
e measures. T
his is tantamount to ar-
guing that it is inappropriate to evaluate the im
pact of an antibiotic if the patient does not take the m
edication for a sufficient period as
prescribed.
Exhibit 3 excludes these "false start" cat-
egories to contrast the outcomes of only
the "relevant" defendant categories: of these, 40 percent had unfavorable out-
C( com
es, and 60 percent had favorableout-
comes. O
ther working definitions of
fi'favo
rable"
and "unfavorable" outcomes
Impact on crim
inal processing. Monthly
many m
ore of their cases were dropped
-could be em
ployed; the important point is
admissions to the D
AT
P were equivalent
or dismissed (including cases m
arked that such decisions m
ust be determined as
to about 7 percent of third-and second-"no action").
a policy matter by relevant court and
degree felony filings during the months
Low
er incarceration rates. During the
agency officials. ~
tud
ied.~
study period, far few
er Drug C
ourt defend-C
ase outcomesand duration. A
s ex-ants than other felony drug and nondrug
Selected findings
pected, "diversion" types of outcomes
defendants were sentenced to incarceration
Length of participationlretention
in treatm
ent. The m
edian length of time
spent by Drug C
ourt defendants in the D
AT
P program, m
easured from the date of
the intake interview to the last day in treat-
ment, w
as 331 days-almost
11 months-
excluding defendants whose charges w
ere dropped. E
xhibit 4 displays the median
time in the program
for Drug C
ourt defen-
(diverted, nolle prossed, case sealed)were
much m
ore frequently recorded for Drug
Court defendants. A
lso as expected,Drug
Court cases took longer to com
plete than those of other felony defendant groups. N
early one-third of Drug C
ourt cases were
still open (unadjudicated) by the end of the 18-m
onth observation period. Tw
o phe-nom
ena largely explain this finding:
for terms of m
ore than a year.
Few
er rearrests. Drug C
ourt defendants generated som
ewhat low
er rates of re-offending, as indicated by rearrests, than non-drug felony defendants in 1990. T
hey accounted for notably low
er rates of re-offending than other felony 2 and 3 drug defendants w
hose cases were not handled
by the Drug C
ourt. (See exhibit 5.)
dants for each of three categories of ver-D
efendants who w
ere permitted to stay
in treatment m
uch longer than origin-A
t the same tim
e, when com
pared tosion 2 program
outcomes: unfavorable,
ally planned. felony drug defendants in 1987,2
years be-favorable, transferred/dropped/other.
fore the Drug C
ourt was im
plemented, the
As now
would be expected by definition,
Defendantsw
ho absconded from the
1990D
rug Court defendants show
ed much
length of program participation (retention
program, leaving their cases indefinitely in
lower rates of rearrest,even after research-
"active" status.in treatm
ent) and program outcom
es ers controlled for possible differences in
closely corresponded. Defendants w
ith un-It w
as difficult to determine w
hether the sam
ple composition.
favorable outcomes had m
edian program
(?stays (225 days) less than tw
o-thirds the length of defendants w
ith favorable out-com
es (364 days). Defendants w
ith the other outcom
es, by definition, showed the
shortest median program
participation, about 19 days.
Drug C
ourt's longer completion tim
es con-tributed to greater use of court resources than norm
al processing would.
How
ever, greater proportions of other felony defendant groups in 1987 and 1990 apparently m
oved more quickly out of the
criminaljustice system
, in part because
Exhibit 3. P
rogram O
utcomes (V
ersion 2) for Drug C
ourt Defendants
Adm
itted to Treatment, A
ugust-September
1990: R
elevant Defendants O
nly
Percen
t of "R
elevant"
Defen
dan
ts
m'.
Unfavorable
Favorable
Pro
gram
Outcom
es (n
z245)
Longer tim
es to rearrest. When D
rug C
ourt defendants were rearrested, the
lengths of time to their first rearrests aver-
aged from tw
o to three times longer than
those of comparison groups (exhibit 6).If
this is generalized across the more than
3,000 Drug C
ourt defendants admitted
since the program began, this finding has
important im
plications for the criminal
caseload of the circuit court as a whole:
Drug C
ourt defendants not only appear to re-offend less often, but those w
ho did re-offend did so only after considerable tim
e had elapsed.
High failure-to-appearrate. A
ll of the D
rug Court defendants are required peri-
odically to appear in court throughout their treatm
ent, and more than half of them
re-corded failures to appear (F
I'A's), com
-pared w
ith from 2 to 11 percent of other
felony defendants. The high rate clearly
results from the requirem
ent for so m
any more court appearances than
associated with norm
al processing of crim
inal charges.
This phenom
enon is similar to that experi-
enced by many program
s granting provi-sional liberty to defendants and suggests that auuroaches should be devised to
monitor appearances m
ore closely to of key them
es that may be of interest not
prevent FT
A's.
only to the jurisdiction itself, as Dade
County plans further efforts to address the
. .
Themes em
erging from the
challenge posed by its drug-involved
Drug C
ourt assessment
caseload,but also to other jurisdictions undertaking or considering sim
ilar drug T
he empirical assessm
ent of Dade
court initiatives. C
ounty's Drug C
ourt revealed a number
Exhibit 4. Length of P
articipation of Drug C
ourt Defendants
Adm
itted to Treatem
ent, August-S
eptember
1990, by Version 2 P
rogram O
utcomes
Med
ian
Days In
Treatm
ent
"
All E
ntering U
nfavorable F
avorable T
ransferrediDroppedi
Defendants
Other
Program
Outcom
es (n
=304)
Strong system
support. A key to the
functioning of the Drug C
ourt is the strong e
joint support shown for the program
by the judiciary, the prosecutor, and the defender. D
rug Court depends on this support to
transact its business in a "team" fashion.
The prosecutor's and defender's roles
are unorthodox and team oriented; both
appear more supportive than adversarial
in encouraging a defendant's pursuit of treatm
ent.
Active judicial role. T
eamw
ork notwith-
"
standing, the leadership role of an actively involved
judge who is fam
iliar with drug-
influenced behaviors is an essential ele-m
ent in the court's capacity to function as w
ell as it does.
The judge can be encouraging and support-
ive in the many brief hearings he orders on
admission of a defendant or to review
a defendant's progress. H
e also is called upon to im
pose sanctions, however, w
hen the defendant show
s poor performance or
has to be returned to the Drug C
ourt on an alias capias.
Exhibit 5. R
earrests During 18-M
onthO
bservation Period: 1990 D
rug Court D
efendantsC
ompared to 1987 and 1990
Felony 2 & 3 D
rug Defendants
Percent of
40 D
efend
ants 30 "
N
o rearrest O
ne T
wo
Three or m
ore
Rearrests
During l&
Mo
nth
Observation P
eriod
Op
ec
ially
designed treatment resources.
One of the critical elem
ents of the Drug
Court in D
ade County w
as development of
a custom-designed substance abuse treat-
ment program
for the programm
atic needs of the D
rug Court specifically. T
he ap-proach focused notably on com
munity-
based outpatient treatment, w
hile making
provision for residential placements for a
very limited num
ber of individuals.
There w
as not in Dade C
ounty (and often m
ay not be in otherjurisdictions) a preex-isting treatm
ent program. Instead, the treat-
ment program
was tailorm
ade to address the target population identified by court of-ficials. Just as the crim
inal court adapted to the treatm
ent goals of the Drug C
ourt pro-gram
, the treatment program
had to modify
practices to respond to the procedures of the D
rug Court, particularly in the areas of
program eligibility and term
ination criteria.
Tolerance for addicts' behaviors. Plan-
ning for the Drug C
ourt sought to recog-nize realistically the sorts of behavior ikely to be associated w
ith drug-involved q
nd
ivid
ua
ls. Within clearly defined public
safety boundaries (defendants would be
transferred out of the program if they w
ere arrested for new
offenses more serious
than those specified by the eligibility crite-ria), the D
rug Court has im
plemented a
flexible or partly tolerant approach to prob-lem
behaviors within treatm
ent. This ap-
proach contrasts clearly with approaches
that specify punishments for program
mis-
steps (such as the days-in-jail ordered for positive drug test results proposed in the D
istrict of Colum
bia's new program
).
Needs for fast, accurate inform
ation. T
he drug court concept and the linking of drug treatm
ent and criminaljustice goals
relies heavily on the need for up-to-date, accurate, and im
mediately accessible data
about defendants, their treatment progress,
and their criminaljustice related problem
s and developm
ents.
In Dade C
ounty, this capacity at first de-veloped at a slow
er rate than the program's
ability to handle cases; it clearly represents m
e of the m
ajor operational challenges ofP the M
iami m
odel in trying to bridge the in-form
ation gap between drug treatm
ent pro-gram
s and the criminal court.
Exhibit 6. M
edian Time to First R
earrest During 18-M
onth Observation
Period: 1990 D
rug Court D
efendants Com
pared to 1987 Felony C
omparison S
amples
1990:Drug C
ourt
1990 Other D
rug
1990 Other D
rug F3
1990: Other D
rug F2
1990:N
on-Drug
1987:lgB7.F3 D
rugD
rug k 1987 F2 D
rug r':%.k
I
--1987 N
on-Drug
k I
0
30
60
90
120 150
180 210
240 270
300
Median D
ays to First Rearrest
Defining (and redefining) the target
population. A m
ajor policy step in imple-
menting the D
rug Court program
was
defining the initial target population. Care-
ful targeting can ensure that the treatment
resources will be deployed efficiently to
process a sufficiently challenging group of defendants w
ith no adverse impact on
public safety. By setting sights too low
(for exam
ple, to deal with very m
inor offend-ers), program
resources could easily have been overw
helmed by a large volum
e of cases, thus preventing benefit from
accru-ing to efforts to address the crim
inal caseload processing or problem
s associ-ated w
ith jail capacity. On the other hand,
assessment findings suggest that the crite-
ria for eligibility might be broadened to
include other types of drug-involved felony-level defendants w
ho may not be
charged with drug offenses.
Targeting to avoid net w
idening. Som
e D
rug Court defendants self-reported that
they engaged in no or very minor levels of
drug abuse, while som
e others tested nega-tively for drugs upon entering the treat-m
ent program. Setting aside questions
about the reliability of such data, the possi-bility that som
e in the treatment program
did not appear to have "serious" drug abuse problem
s raises important questions
about targeting and screening procedures.
Similarly, the finding that D
rug Court de-
fendants had their criminal charges
dropped or dismissed m
uch less frequently than other types of felony defendants raises the possibility that som
e would not have
ventured very far into criminal processing
had they been processed in other criminal
courts or during an earlier period.
Although the assessm
ent found no evi-dence that the M
iami D
rug Court notice-
ably "widened the net," particularly given
its selective felony-level focus, the possi-bility of net w
idening as an inadvertent side effect should be kept in m
ind by the D
ade County program
itself and by other jurisdictions considering sim
ilar efforts.
By setting sights too low
, the Drug C
ourt m
ay "sweep" into its "net" persons w
ho ordinarily w
ould not require many or any
of its scarce resources. By targeting cat-
egories not usually fully processed by the crim
inal courts, such a program m
ight un-w
ittingly add to the court workload and the
jail population as well as intervene w
hen intervention is not necessary.
Screening for eligible candidates and "hitting" the target population. G
iven a suitable target population policy, a separate elem
ent critical to effective implem
enta-tion of a D
rug Court is establishm
ent of a rigorous screening m
echanism that identi-
fies persons eligible for the program at the
earliest stages of processing. Mechanism
s that "m
iss" large portions of the target population or that carelessly include indi- viduals not m
eeting the eligibility criteria can adversely affect the D
rug Court's abil-
ity to meet its objectives.
Defining "success" in outcom
es as a m
atter of policy. Defining "success"-
what w
ill constitute favorable and unfavor- able outcom
es-is an im
portant policy m
atter to be resolved by debate and con- sensus am
ong key officials. This policy
debate is best carried out in advance of im
plementation and evaluation. Such a
policy should clearly detail the behaviors of participants that are acceptable, that are tolerated but sanctioned in som
e specified fashion, or that som
ehow cross the
boundary into unacceptable. program-
terminating actions. T
he implications of
enforcement of such a policy approach
would m
ost helpfully be analyzed in ad- vance of im
plementation, and m
odifica- tions m
ay be necessary periodically and be m
ade on the basis of program experience.
Strengthening reliability of information
relating to defendant drug abuse. A key
to effective early classification and effi- cient subsequent treatm
ent of drug- involved felony defendants m
ay be closer coordination and com
puter information ex-
change between Pretrial Services (or other
early processing agency) at the postarrest interview
stage and treatment intake staff.
A com
bination of carefully structured self-report questions about drug use at the Pretrial Services and treatm
ent intake stages and selective initial drug testing, for exam
ple, may contribute to im
proved targeting and program
ming of D
rug Court
candidates.
Developm
ent of defendant classifica- tions for risk and treatm
ent planning. C
lassification of defendants at the earliest stages based on estim
ated drug involvem
ent and risk to public safety can be developed to assist in the targeting of approp2ate candidates for D
rug Court and
in planning for treatment and supervision
in the comm
unity during Drug C
ourt involvem
ent.
Need for different treatm
ent programs.
In differentiating entering defendants according to estim
ated drug-involvement
and public safety risk, an improved initial
stage classification approach can help tar- get D
rug Court defendants efficiently to
treatment regim
ens of possibly different substance and length, w
hile still ensuring equitable treatm
ent of defendants overall.
Such a classification could maxim
ize effi- cient use of resources by assigning low
er-risk and less drug-involved defend- ants to som
ewhat shorter program
s of treatm
ent to be complem
ented by other nonincarcerative options w
hile channeling m
edium-risk and m
ore seriously drug- involved defendants into longer and m
ore intensive program
s. For equity, an aim
would be to provide equivalent diversion-
ary programs so that defendants are treated
similarly overall, even given their
different content.
The role of drug testing. T
he use of this technology should be carefully reexam
ined as a m
atter of policy: either it should be de- ployed m
ore effectively and selectively, lim
ited to initial tests, used more system
- atically w
ith self-reported drug use infor- m
ation, or even eliminated, if necessary, to
save costs. Inconsistent use of drug testing contributes little to the inform
ation require- m
ents of the Drug C
ourt program.
The role of acupuncture. A
cupuncture is em
ployed in the Dade C
ounty Drug
Court's treatm
ent program on a voluntary
basis as an adjunct to treatment for defend-
ants attending the outpatient treatment
program. A
s such, acupuncture has not been view
ed by the program as a specific
treatment m
odality. Instead it is employed
as a resource for stabilizing defendants, particularly during the early phases of treatm
ent, and for increasing amenability
for treatment.
The failure-to-appear problem
. A clear
implication of the court-based, judge-
supervised model of D
rug Court is that the
much m
ore frequent scheduling of defend- ants before the judge ultim
ately translates into m
any more failed appearances (alias
capiases issued) when the experiences of
Drug C
ourt defendants are compared to
8
those of "normal" defendants. (T
his may
be true even if the ratio of absences-to- scheduled- hearings m
ay not be worse
among D
rug Court defendants.) T
hus, strategies to address this D
rug Court
"side-effect" without overw
helming court
or jail resources should be made early in
the planning stages of Drug C
ourt efforts.
The resource im
plications of the Drug
Court program
. Court system
s have a practical interest in learning about the "cost-effectiveness" of the D
rug Court ap-
proach. Because this assessm
ent was not
designed as a cost-effectiveness study, clear conclusions about the resource im
pli- cations of this approach are not offered. N
evertheless, cost considerations are criti- cal to an overall appraisal of the D
rug C
ourt's promise. Such an analysis is com
- plicated, and its outcom
es would depend
heavily on the assumptions m
ade about costs and savings in a variety of areas. Principal focuses for such an analysis w
ould need to consider the costs associated w
ith, for example, a) operating one or
more courtroom
s dedicated strictly or largely to D
rug Court transactions, b) the
costs of treatment, and c) the costs of
missed appearances and program
mis-
starts, as well as possible savings in d) case
processing, e) confinement, and f) reduced
or slowed rates of reoffending.
The need for routine experim
ental evaluation. T
his assessment has revealed
but not resolved a number of them
es and issues relating to the use of the D
ade C
ounty Drug C
ourt. As other jurisdictions
proceed with their plans to im
plement
Drug C
ourts or continue with efforts al-
ready underway, serious consideration
should be given to simultaneous im
ple- m
entation of more rigorous, experim
ental evaluations. Fuller evaluation can point to the strengths and w
eakness of the Miam
i D
rug Court m
odel, and the advantages and disadvantages of the variety of initiatives now
underway in other court system
s. Fail- ure to incorporate plans for rigorous evalu- ation at the in the initial stages of planning and im
plementation preordains the use of
less than optimal evaluation m
ethods at a later stage w
hen questions of program im
- pact m
ay become critical.
Character of defendant
Op
rog
res
s through D
ade's D
rug Court
For practical reasons, the assessm
ent adopted a fram
ework that sought to record
defendant program, case and public safety
outcomes, as of an arbitrary point in tim
e 18 m
onths after defendants were adm
itted to the treatm
ent program. Som
e officials have argued that this could result in a very "flat" or "one-dim
ensional" accounting of the perform
ance of defendants in the pro-gram
. They have pointed out, in fact, that
the Drug C
ourt's overall approach was
based in part on the operating assumption
that the behavior of drug-involved indi-viduals w
ould be, almost by definition, er-
ratic and generally irresponsible-at least
in the earliest stages of treatment. T
hus, these officials reasoned, a sim
ple, quantita-tive m
easure of program outcom
es would
fail to convey the "ups and downs," "zig-
zags," and other kinds of "real-life" be-havior actually involved in treatm
ent program
progress.
In addition, great concern was expressed
by Dade officials that som
e defendants w
ho had great initial difficulty in the pro-gram
might be view
ed as "failures" under this approach, w
hen, had the observation period extended farther, ultim
ate success w
ould have instead been recorded as the final result. A
lthough it is perhaps com-
mon for officials responsible for program
s undergoing assessm
ent to feel that quanti-tative approaches som
ehow m
iss capturing the character of the program
experience, their point seem
s well supported in the
Dade C
ounty study by selected case his-tory illustrations. Several draw
n from the
full research report are presented here.
Case 1. R
. was in her m
id-30's when ad-
mitted to the treatm
ent program. She w
as arrested in Septem
ber 1990on cocaine
possession charges and assigned imm
edi-ately to D
rug Court. She had a substantial
history of criminaljustice involvem
ent, w
ith 13prior arrests (but only 1
within the
past 3 years) and 9 convictions, 5 for felony property offenses. She had no prior
-arrests for drug offenses. She w
as single,a high school graduate, living alone and
Exh
~b
~t
7. Felony 2 and 3 Drug D
efendants Entering C
ircuit Court in
A
ugust-Septem
ber 1990, A
ssigned to Drug C
ourt, by A
ctual Adm
ission to D
rug Court
Never adm
itted A
dmitted to D
rug Court
-16%
A
ug.-S
ept. 1990
69% (&
Adm
itted after 8
Aug.-S
ept. 1990
15%
7 -
Target Po
pu
lation
of Felony Drug D
efend
ants
Iden
tifiedfor D
rug Court
(n=305)
working full-tim
e. In her intake interview,
R. said she had been using drugs since age
17. She admitted current use of heroin,
marijuana, alcohol, and cocaine. A
dmitted
to treatment during the study period on
September 17, 1990,she had poor attend-
ance initially and consistently tested posi-tive for drugs, show
ing little motivation
for treatment.
In early Novem
ber, after a 2-week ab-
sence, she returned to treatment citing the
demands of her w
ork as the reason for m
issing appointments. She w
as not then seen again until the end of D
ecember.
From this point on, she show
ed slight im-
provement. A
lthough her attendance con-tinued to be poor, her drug tests, w
hen she did com
e, were usually negative. In Febru-
ary 1991,her attendance improved, but in
April she once again stopped attending
treatment.
In May, the defendantreturned once m
ore to D
AT
P, although the length of her ab-sence is not specified. H
er attendance im-
proved somew
hat, and her urine tests were
generally clean during the next months. In
mid-July, after 10 m
onths of participation, she w
as finally transferred to phase I11 aftercare. A
t the end of the 18-month
ob-servation period, the defendant's case w
as still open and she was still active
in treatment.
Interestingly, her records further showed
that as late as September 1992,or nearly 2
years after her initial admission, she did in
fact complete treatm
ent successfully with
the result that her criminal charges w
ere nolle prossed.
Case 2. C
., a man about 20, w
as arrested in early July on cocaine possession charges. H
e entered DA
TP on Septem
ber 26, 1990, after his case w
as transferred to Drug
Court. H
e had two prior arrests, both for
misdem
eanors. Although he w
as charged w
ith possession of cocaine, at his intake in-terview
he admitted only to using m
ari-juanahashish. H
e reported being a drug user since age 18,but this w
as his first tim
e in treatment. A
lthough he initially ap-peared m
otivated for treatment, on N
o-vem
ber 19, 1990,he stopped attending.
On A
pril 22, 1991, C. w
as once again re-ferred to D
AT
P following another arrest
for possession. After a m
onth, he was
again responding poorly and testing posi-tively for drugs. O
ne month later he w
as again reported to have stopped com
ing to treatm
ent. In January of 1992,the defend-ant w
as once more readm
itted by the Drug
Court judge. A
lthough he was still active
in treatment at the close of the 18-m
onth observation period, he dropped out again shortly after. R
ecords show his pattern of
behavior continued. He w
as readmitted in
late September of 1992 and again discon-
tinued treatment just over a m
onth later.
Case 3. Y
., 42 and an imm
igrant, was m
ar- ried but living apart from
her husband at the tim
e of her arrest in a sting for cocaine purchase and possession. She w
as college educated but unem
ployed and was earning
a living as a freelance translator and teacher. A
dmitted to D
AT
P on September
6, 1990, she admitted at her intake inter-
view infrequent cocaine use (less than once
per week), as w
ell as alcohol use. She re-
ported also that she had been using alcohol since 1967 and cocaine since 1983. A
t ad- m
ission she tested positive for both co- caine and am
phetamines.
Y. w
as reported motivated and cooperative
throughout treatment. She transferred to
phase I1 on October 2, 1990, and continued
good progress, attending treatment and
having negative drug tests until her transfer to phase I11 on D
ecember 3, 1990, w
hen she recorded a positive drug test. A
cupunc- ture and individual counseling helped her through this period. H
er attendance and at- titude continued good, and the "binge" did very little to slow
her completion of the
program. She w
as recomm
ended for graduation on A
ugust 28, 1991, slightly less than a year after adm
ission, and her case w
as later nolle prossed. File notes state that in addition to helping her w
ith her drug problem
, counselors tried to help her address problem
s related to employ-
ment and her m
arital situation.
Case 4. C
., 38 at the time of her adm
ission, entered the D
rug Court program
after ar- rest for cocaine possession on A
ugust 24, 1990, despite a long crim
inal justice his- tory under a num
ber of aliases. This w
as reportedly her first tim
e in drug treatment.
She was adm
itted on September 14, failed
to attend the program, and adm
itted again on O
ctober 9, 1990. On D
ecember 12 she
was reported to have discontinued treat-
ment. O
n March 15, 1991, the defendant
was again ordered readm
itted by the court. O
n April 26, 1991, she w
as reported to have failed to return. N
o further notes were
found after that date.
Case 5. S., 24, w
as single and had a high school education. H
e should have been a success story. H
e was arrested in D
ecem-
ber 1989 on charges of cocaine possession, 6 m
onths before Drug C
ourt was estab-
lished. His case w
as assigned to Drug
Court m
ore than a year and a half later, on A
ugust 28, 1990. At his intake interview
, he denied any drug use and case notes indi- cated that counselors w
ere inclined to be- lieve him
, based on his consistently negative drug test results, his cooperative m
anner, and his physical appearance. On
September 14, 1990, he w
as transferred to phase I1 and on O
ctober 22, 1990, after clean urine tests and good progress in treat- m
ent, to phase 111. During the program
, he obtained full-tim
e employm
ent and made
plans to further his education. Treatm
ent records show
he continued to do well, his
attitude was good, and he w
as drug free. H
e was w
orking long hours and was re-
quired to attend only weekly.
In March 199 1, he w
as placed in phase V
for "tracking" due to unexcused nonatten- dance, but he returned several days later and explained that his absence had been due to a fam
ily emergency. In July he w
as briefly jailed after being involved in a fight at a flea m
arket. On A
ugust 29, 1991, he w
as to have been recomm
ended for gradu- ation. S
. failed to appear for his scheduled court date and an alias capias w
as issued. H
e also failed to keep a clinic appointment.
On A
ugust 30, his father informed the
counselor that his son had been robbed and killed.
Reaching the
target population
Exhibit 7 portrays all felony defendants
charged during August and Septem
ber 1990 w
ho were assigned to D
rug Court.
About one in three (3 1 percent) of defen-
dants identified as meeting the chargelpri-
ors criteria and assigned to be processed in D
rug Court appeared not to be adm
itted to treatm
ent by the program im
mediately, for
any number of reasons. A
lthough this pro- portion suggests that D
rug Court w
as pro- cessing fully tw
o-thirds of the identified population of eligible defendants as they entered court processing, it raises questions about w
hy some eligiblelassigned defend-
ants were "m
issed" or did not participate in the voluntary diversion and treatm
ent
program once identified. Several phenom
- ena m
ay explain the "miss" rate.
Some m
ay merely have decided to de-
cline, preferring to take their chances with
traditional criminal processing.
Others m
ay have chosen to post bond im
mediately after arrest, thus elim
inating their opportunity for com
ing in contact w
ith Drug C
ourt.
Pretrial services staff might on occasion
have missed som
e defendants. A few
may
have simply w
alked away w
ithout an in- take interview
.
Drug C
ourt officials pointed out that initially, a sm
all number of defendants
agreed to report to the Model C
ities Clinic
for intake procedures, but never made
their appointments after they w
ere given pretrial release.
A sizable m
ajority of those eligible ap- peared to have been "enrolled," how
ever, and careful em
pirical examination casts
doubt on the initial finding of a 3 1 -percent m
iss rate. As m
any as 40
of the 83 defend- ants in this group (Sam
ple 11, as shown in
exhibit 1) may have entered treatm
ent through D
rug Court at som
e time during
the 18-month observation period, just not
in the August-Septem
ber sample period.
These findings suggest that, in fact, the
Drug C
ourt may have had a fairly effective
reach. The researchers estim
ate that the m
iss rate ultimately m
ay have been as sm
all as 17 percent. This finding of de-
layed enrollment com
plements the earlier
finding that about one-third of admis-
sions to Drug C
ourt were of defendants
whose charges had been filed during an
earlier period.
Conclusion
This research has focused on the innova-
tive efforts of one jurisdiction, the Elev-
enth Judicial Circuit in D
ade County, as it
shifted from the then prevailing paradigm
guiding crim
inal courts in the response to the drug-related caseload tow
ard a court-based approach to the treatm
ent of felony drug defendants. T
hroughout this research, and particularly as this report w
as reaching com
pletion, word of interest in
and efforts to develop Miam
i-type drug courts in m
any other criminal court
ystems in the U
nited States grew increas-
ingly frequent. Anecdotal reports of initia-
tives in other sites pointed to the possibility that a variety of interesting and potentially effective variations on the M
iami m
odel m
ay be underway in locations across
the nation.
As other jurisdictions proceed w
ith plans to im
plement drug court program
s. serious consideration should be given to conduct- ing sim
ultaneous, rigorous evaluations. Such evaluations can point to strengths and w
eaknesses of the Miam
i Drug C
ourt m
odel as well as those of other initiatives
around the Nation. Failure to incorporate
plans for experimental evaluations in the
initial planning and implem
entation stages preordains the use of less than opti- m
al evaluation methods at a later stage
when questions of program
impact m
ay becom
e critical.
Notes
I. The Drug C
ourt's treatment em
phasis is rim
arily on outpatient modalities. H
ow-
er, in 199 l, the Drug C
ourt arranged through the Florida system
for prioritized access to m
ore than 200 residential place- m
ents for selected defendants with particu-
larly difficult drug abuse problems. A
s of spring 1993, an average of about 40 such placem
ents were in use at a given m
oment.
2. The objective of this m
ultisample, com
- parative approach w
as to be able to view
processing of Drug C
ourt defendants in the context of felony defendants overall. O
ne lim
itation of this approach-shared by an
experimental approach as w
ell-is that
prior or subsequent cohorts could have recorded different outcom
es than those described in this report. N
evertheless, the rationale for this approach assum
es that de- fendants entering during the study period w
ere fairly typical.
3. When inform
ation about a defendant's status w
as uncertain or conflicting, priority w
as given to criminal justice inform
ation sources.
.Although the D
rug Court initially only
rgeted third-degree felony drug posses- sion cases w
ith no prior convictions, by 1990, persons w
ith initial charges
involving selected second-degree drug felonies (purchase of drugs) w
ere consid- ered for the program
as well as som
e de- fendants w
ith prior convictions. One w
ay of estim
ating the impact of the D
AT
P is, therefore, to determ
ine the proportion of relevant cases that w
ould have been eli- gible for D
AT
P and the proportion actually entering the program
. Given that m
onthly adm
issions include some cases filed during
previous months, the researchers deter-
mined the com
parative figures.
References
Am
erican Bar A
ssociation, 1993. The State cfC
riminu1 Justicp. W
ashington, D.C
.: A
BA
, Section on Crim
inal Justice.
Belenko, S., 1990. "T
he Impact of D
rug O
ffenders on the Crim
inal Justice System"
in Dnrgs, C
rime and the C
r-inzitzal Justic.e System
, ed. R. W
eisheit. Cincinnati and
Highland H
eights: Academ
y of Crim
inal Justice Sciences and A
nderson Publishing C
o., 27-78.
Gerstein, D
.R., and H
.J. Harw
ood, eds., 1990. Treating D
rug Problem
s. Vol. 1 :A
Study ofthe E
volution, Efi~
~ctivenes.~
, and
Financing c$
Puhlic and P
ritaate Drug
Treatment Systenzs. W
ashington, D.C
. : N
ational Academ
y Press.
Gerstein, D
.R., and H
.J. Harw
ood, eds., 1992. Treating D
rug Problem
s. Vol. 2:
Com
missioned P
apers on Ni.storicu1, Insti-
tutional, and Econom
ic Conte.xtLs c!f'l)rug
Treatment. W
ashington, D.C
.: National
Academ
y Press.
Goerdt, J.A
., C. L
omvadias. G
. Gallas, and
B. M
ahoney, 1989. E,x-amlining C
ourt De-
lay: The Pace c$
Litigation in 26 Urburl
Trial Courts, 1987. W
illiamsburg, V
ir- ginia: B
ureau of Justice Assistance and
National C
enter for State Courts, U
.S. D
epartment of Justice.
Goerdt, J.A
., and J.A. M
artin, Fall 1989. "T
he Impact of D
rug Cases on C
ase Pro- cessing in U
rban Trial C
ourts." State Caur-I
Journal, 4-1 2.
Goldkam
p, J.S., 1987. "Prediction in C
riminal Justice Policy D
evelopment" in
ClassiJi'cation and P
redic.tion in CI-im
inal
.lustice Dc~cisions, eds. D
.M. G
ottfredson and M
. Tonry. C
rime and Justice 9. C
hi- cago: U
niversity of Chicago Press.
Goldkam
p, J.S., M.R
. Gottfredson, and D
. W
eiland, 1990. "Pretrial Drug T
esting and D
efendant Risk." Journal of C
riminal Law
and C
riminology 8
8, 3:585-652.
Goldkam
p, J.S., M.R
. Gottfredson,
P.R. Jones, and D
. Weiiand, 1993.
Pel-.sonu1 Liberty and C
omm
unity Safety: P
~-c.triulRelease in the Crim
ina! Courts.
Forthcoming.
, 1990. Vol. 2, A
ssessing the In?- p
utt c!f'D
rug-R~
latedCriminal C
ases on P
ublic Safety: Drug-R
eluted Recidivisnz.
Philadelphia: Tem
ple University. D
epart- m
ent of Crim
inal Justice.
, 1990. Vol. 3,A
ssessing the Im-
pact ($Drug-R
elated Crim
inal Cases on
the Judic.iul P~
OI.V
SS
, C
row'djng and P
ublic Sufetv: Sum
mary and Inzplicutions. Phila-
delphia: Tem
ple University, D
epartment of
Crim
inal Justice.
Goldkam
p, J.S., P.R. Jones, D
. Weiland,
and S.H
. Smith, 1992. lssues irt M
anaging the D
rug-1nt~olt.ed Qfgndeirr- in the Com
- m
unity: Findings F
rom u N
ational Surl-ey i$P
robationlParoIe and D
t-u,q Ahuse
Treatn~
entAgencies. Philadelphia: Crim
e and Justice R
esearch Institute.
Goldkam
p, J.S., and D. W
eiland, 1991. A
ssessittg the Impact of D
ude County's
Felony D
rug Court. Philadelphia: C
rime
and Justice Research Institute.
, 1993. The Impact of C
ourt-R
elated Practices on the C
orrectional P
opulation in Dude C
ounty. Phil adel phia: C
rime and Justice R
esearch Institute.
Johnson, B.D
., P. Goodm
an, E. Prebie, J. Schm
eidler, D. L
ipton, B. Spunt, and T
. M
iller, 1985. Taking Care of B
usiness: The E
conomics of C
rime by N
et-oin A
husecs. Lexington, M
assachusetts: L
exington Books.
National A
ssociation of Pretrial Services A
gencies, 1978. Pe~
orm
an
ces Standards ancl G
oals jbr Pretrial R
elease and D
iversion: Diversion 5.
-
President's Com
mission on L
aw E
nforce-m
ent and Adm
inistration of Justice, 1967. 1
~IcU
mtpis p
rofe
m
c stice at T
emple U
ni The C
hallengeof Crim
e in a Free Society.
knflaae,paia M~w
eilandis W
ashington, D.C
.: U.S. G
overnment Print-
aq
so
cia
teat the crim
ea ing O
ffice. Justice Institute, Philadelphia
Smith, B
.E., R
.C. D
avis, and S.R.
Goretsky, 1991.Strategiesfor C
ourts To
Cope W
ith the Caseload P
ressures of Drug
Cases: Trial R
eport. Washington, D
.C.:
Am
erican Bar A
ssociation. --
Findings and conclusions of the research re-ported here are those of the researchers and do not necessarily reflect the official position or policies of the U
.S. Departm
ent of Justice.
TheN
ationalInstitute of Justice isa com
po-nentof theO
fice ofJustice Program
s,which
also includes the Bureau of Justice A
ssist-ance, B
ureau of Justice Statistics, Ofice of
Juvenile Justice and Delinquency P
reven-tion, and the O
ffice for V
ictims of C
rime.
This E
valuation Bulletin is based on the
complete R
esearch Report, A
ssessing the N
CJ 145302
lmpact of D
ude County's F
elony Drug
Court,N
CJ 144524.A
detailed description of the M
iami D
rug Court's operations
can be found in the NIJ Program
Focus, M
iami's "D
rug Court':' A
Different
Approach, N
CJ 142412. B
oth of these docum
ents can be obtained through the N
ational Crim
inal Justice Reference
Service, 800-85 1-3420.
U.S. D
epartment of Justice
Office of Justice Program
s
National Institute qf Justice
BU
LK
RA
TE
PO
STA
GE
& FE
ES PA
ID
DO
JNIJ
Permit N
o. G-9 I
Washington. D
.C. 20531
Official B
usiness Penalty for Private U
se $300