Mitchell O., Wilson D.B., Eggers A. et al.
Journal of Criminal Justice: 2012, p. 60–71.
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr Mitchell at firstname.lastname@example.org. You could also try this alternative source.
Synthesising the results of 154 studies, the most thorough and extensive investigation of the crime-reduction credentials of drug courts finds the evidence bulky but lacking quality, yet sufficient to support courts for adult illegal drug users if not (or not yet) teenagers or drink-drivers.
Summary Drug courts specialise in closely supervising (through regular urine tests and court appearances) and ordering the treatment of suitable drug-related offenders to improve compliance with treatment as an alternative to prosecution or imprisonment. In what is intended to be a non-adversarial environment, judges impose sanctions or offer praise or more tangible rewards and adjust treatment depending on progress. The incentive for suspects or offenders is that charges or penalties will be reduced or dismissed upon successful completion of the court's orders, which typically takes at least a year, a point reached by about half the caseload.
In just two decades, drug courts have grown from a single court to an international movement with over 2400 operative in the USA and others in the UK, Canada, New Zealand, Australia, South Africa, Bermuda, and Jamaica. Most US courts (which account for most of the evidence of their impacts) restrict their intake to non-violent, substance dependent 'clients' and many exclude drug dealing offenders or those with extensive and serious criminal histories or serious mental health problems. As a result, most eligible offenders have been charged with drug or property crime and have relatively few prior convictions for serious offences.
The featured review and meta-analysis A study which uses recognised procedures to combine quantitative results from several studies of the same or similar interventions to arrive at composite outcome scores. Usually undertaken to allow the intervention's effectiveness to be assessed with greater confidence than on the basis of the studies taken individually. aimed to pool the results of many more drug court evaluations than previous reviews, while distinguishing between drug courts targeting adult illicit substance users, those for juvenile offenders, and those for adults convicted of driving while under the influence of drink. It also aimed to discover what makes drug courts more or less effective, whether their impacts last, and whether their effectiveness is supported by the most rigorous trials.
To this end the analysts sought studies which compared the subsequent criminality of offenders adjudicated by drug courts with those adjudicated in usual court systems. Pooling of results from these studies was done in a way which did not assume there was one 'true' degree to which drug courts affected recidivism which varied only by chance, but that features of the study, court, offenders or other factors might create real differences in impact. Where available, results in terms of actual drug use were also analysed. In all 154 such studies were obtained, all but eight of US courts.
The analysts found this body of work generally methodologically weak. Just eight of the 154 studies had randomly allocated offenders to drug courts versus alternative judicial procedures, the securest way to eliminate bias due to differences in offenders who do or do not choose and/or get referred to and accepted by drug courts. Most studies obtained their comparison groups in ways (such as offenders tried before the courts opened or those turned down by the courts) which do not convincingly eliminate risk of bias due to the selection of offenders for the courts or other factors.
Across all 92 studies of courts for adult drug users, those processed through drug courts were moderately but significantly less likely to later reoffend – and in particular to commit drug crimes – than offenders subject to usual judicial procedures. Assessed by just four studies, drug use itself was also reduced, but not significantly. The figures meant that if 50% of offenders reoffended after normal procedures, 38% would have done so had they been through a drug court. This effect was not significantly attenuated among non-US studies or over follow-up periods beyond one year and up to three years, and nor was it significantly less when the follow-up period extended partly or wholly beyond the offender's sentence.
However, the most methodologically rigorous studies registered the lowest impacts. Across the three to have randomly allocated offenders to drug courts versus alternative procedures, recidivism over the next year was lower among drug court offenders, but the finding was not statistically significant. This was perhaps because one of the studies atypically randomised offenders to a comparison condition which entailed more intensive drug testing than the drug court. In this study too, while ineffective at one year, the drug court was associated with reduced offending at the three-year follow-up. Findings in the remaining two studies, or including the exception's three-year follow-up, were statistically significant, and recidivism reductions were about as great as across all studies.
The next most sound studies typically attempted instead to match drug court and comparison offenders on key variables, or to adjust the findings for their relative risks of offending. Across these 20 studies, recidivism was modestly and significantly lower among drug court offenders, but such research designs have limited power to iron out important potential differences between offenders who are or are not referred to (or choose to be tried by) drug courts.
Enough suitable studies were available to assess which features of adult drug courts (or studies of them) were associated with the greatest success. In respect of both general and drug offending, courts in which over half the offenders successfully completed their orders had the best records. Of the 18 remaining tests (nine each for reoffending in general and drug offending), two features were associated with lower rates of drug crime: being tried by a court which sees offenders at least twice a month, or by one which could hold out the prospect that success would expunge the original offence. The same was not true to of recidivism in general; instead, the minority of courts which accepted violent offenders had less impact than the remainder.
Across the 34 relevant studies, only general offending was significantly (but modestly) reduced by juvenile drug courts, figures which implied that if 50% of offenders reoffended after normal procedures, 42% would have done so had they been through a drug court. But across the most rigorous studies, findings were less impressive (equivalent to a reduction from 50% to about 44%) and not statistically significant. There was essentially no impact on drug offending in particular, and pooled results from the three studies to assess drug use recorded modest reductions which did not reach statistical significance.
Across the 28 relevant studies, offenders supervised by drink-driving courts were moderately but significantly less likely to later reoffend – and in particular to commit drug/alcohol crimes – than offenders processed through usual judicial procedures. Drug use itself – assessed by just two studies – was reduced but not significantly. The figures meant that if 50% of offenders reoffended after normal procedures, 38% would have done so had they been through a drink-driving court. Again the strongest effects were seen in the weaker studies. Across the four randomised trials, effects were small and non-significant, heavily influenced by one study which found the courts tended The finding was not statistically significant. to be counter-productive. [Editor's note: Due to legal requirements, this study in California was forced to compare drug court extensions to sentencing and supervision against what already was an intensive treatment programme and regular court hearings. The drug court also had little leverage over offenders due to an inability to waive charges on successful completion and to prison overcrowding, which meant jail terms were rarely implemented.]
Drug using offenders adjudicated by drug courts are less likely to reoffend than those normally adjudicated, but how much varies with the type of drug court. In adult drug courts, the average effect is equivalent to a reduction in general and also specifically drug recidivism from 50% to 37–38%, reductions which persist for at least three years after the start of the sentences. Thus, the accumulated evidence supports the continued funding, development, and operation of adult drug courts. Drink-driving drug courts have comparable impacts on recidivism but findings from the most rigorous, randomised trials are ambiguous, and more such trials are needed to definitively demonstrate effectiveness.
In contrast to the moderate effects of drug courts for adult illicit drug users and drink-drivers, juvenile drug courts have a minor impact on recidivism. Possible reasons are that, unlike many adult courts, these courts see relatively high risk offenders, and that their requirements (eg, in terms of drug testing, court hearings, programme duration) are less demanding.
Even if overall effective, the effects of different courts vary greatly. Exploring the possible reasons uncovered some evidence supporting the importance of 'leverage' and intensity, in that drug courts which dismissed charges for successful participants and courts with more frequent hearings had larger reductions in recidivism, but these findings were statistically significant only in respect of drug-related recidivism.
Courts which dealt only with non-violent offenders were more effective in reducing crime than those which extended to violent offenders. At first this seems to contradict analyses which have found that violent offenders do as well in drug courts as non-violent offenders. But these analyses reported on offenders, the featured analysis on courts. It remains possible that courts which admit violent offenders are for some reason less effective than other courts, and at the same time, that within each court violent offenders do as well as non-violent.
An important issue (which goes beyond the data collected for the featured analysis) is whether in other ways drug courts unduly restrict eligibility, for example, by excluding people charged with supplying drugs. Many supply to support their own drug use and may benefit from drug court treatment. Likewise, many courts also exclude offenders with extensive criminal histories and serious mental health issues. Though more of a risk to public safety, offering effective drug treatment to such offenders promises to reduce reoffending. Outside a drug court context, some evidence (of which the Breaking the Cycle evaluation is the most prominent) suggests that expanding the drug court model to broader populations of offenders can be effective.
commentary This analysis shows there is no simple answer to the question, 'Do drug courts reduce crime?' The answer is – 'It depends'. What it depends on is partly what the court is compared against. Weaker studies which cannot exclude the possibility that more promising offenders find their way to drug courts also find the largest effects. Presumably crucial variables – like how committed the offenders is to succeed, their social and family support, or professional assessments of how well suited they are to a drug court regimen – are rarely available to researchers so cannot be adjusted for. Effects remain even in randomised trials which should eliminate this source of bias, but perhaps partly because these are so rare, pooled results from these trials are not statistically convincing.
Results also vary with another feature of the comparator – the degree to which it is like the drug court it is being compared against. This probably works both ways: the gap between the two may be narrowed because the drug court is unable to fully implement a drug court model, or because the comparator already incorporates features of this model. Both influences may account for the lack of impact of Scottish drug courts in the main UK study to report on recidivism.
The two features found by the featured analysis to be associated with increased drug court effectiveness – seeing offenders at least twice a month and holding out the prospect that success would expunge the original offence – were among the effective ingredients also identified in a major study funded by the US Department of Justice of 23 drug courts. However, in the featured analysis these were among three significant findings to emerge from
Excluding graduation rate and whether the study was formally published.
of different features of drug courts. Had the analysis adjusted for the
Especially if multiple outcomes tend not to covary, the more are measured, the more likely it is that some will reach the threshold for a statistically significant difference purely due to chance variations in the samples rather than any real impact of the interventions being tested. For example, by convention, if a difference would happen only 1 in 20 times by chance, it is considered a non-chance occurrence possibly due to the intervention. But if, say, 20 independent outcomes are measured, more often than not one would cross this threshold purely by chance. To cater for this, it is recommended1 that researchers consider raising the threshold (in the example, according to some adjustment methods to as high as 1 in 400) before each of the outcomes is considered to have reflected a statistically significant difference.
1 International Conference on Harmonisation Of Technical Requirements for Registration of Pharmaceuticals for Human Use. "ICH harmonised tripartite guideline statistical principles for clinical trials." Statistics in Medicine: 1999, 18, p. 1905–1942. that some of these tests might have been found statistically significant purely by chance, it could be that none would have crossed this more stringent threshold. Though across a caseload they may on average reduce recidivism, the benefits of frequent hearings have been found to be limited to high risk In the linked example, those characterised by antisocial personality disorder or with a history of treatment for drug abuse problems. offenders.
The featured analysis is the newest, most recent and perhaps too the most probing of recent reviews of drug courts. Its findings largely accord with other recent reviews which find the bulk of the evidence on the side of drug courts, but also that this evidential weight is not matched by methodological quality in sufficient amounts to be confident that drug courts causes the overall recidivism reductions, as opposed to other factors the studies were unable to eliminate.
A review of interventions for drug using offenders produced for the Scottish Government accepted that research on Scottish drug courts had shown these cost more than normal procedures (which themselves embodied some drug court features including drug testing and treatment) yet did not further reduce crime. Like the featured analysis, the review warned that the most rigorous international trials which had randomly allocated offenders to drug courts versus other judicial options found only weak crime reduction impacts which fell short of statistical significance.
Reservations in the Scottish review cited above over the evidence for drug courts from randomised trials were echoed in a review conducted by British experts for the Swedish Council for Crime Prevention. It was able to synthesise crime-reduction results from just two high quality trials. Together these registered an advantage for drug courts versus comparison judicial options, but not one which was statistically significant. According to this analysis, treatment in general had been shown to reduce drug-related crime, but the same could not yet be said of treatment delivered via a drug court.
Mandated by US law, in 2011 the US Government Accountability Office investigated how well US adult drug courts have reduced crime and substance use and their associated costs and benefits. They reported that compared to alternative dispositions, generally studies found drug courts were associated with lower rates of criminal recidivism and relapse to drug use, but few studies were free of possible bias arising from non-random selection of drug court versus comparison offenders. Due mainly to reduced future victimisation and justice system expenditures, benefits to society expressed in financial terms usually but not always outweighed costs. This balance was partly dependent on the expense of the alternative disposal; if community sentences supervised by a drug court replaced prison, the cost savings were likely to be positive and substantial.
See a related Findings analysis for a summary of British policy and experience in relation to drug courts. For all Findings drug court analyses run this search. In particular see these background notes with a detailed consideration of one of the most methodologically rigorous studies to date, conducted in Baltimore with a caseload unusually relevant to the UK because it consisted mainly of heroin addicts with extensive criminal records. Though methodological concerns remained, it found that over the three years after offenders had been allocated to the court or to normal proceedings, the average numbers of new arrests and charges were significantly fewer among drug court offenders and drug use was lower.
Last revised 09 April 2012. First uploaded
HOT TOPIC 2016 Don’t treat, just test and sanction
DOCUMENT 2012 Will intensive testing and sanctions displace treatment?
REVIEW 2011 A new paradigm for long-term recovery
STUDY 2009 Review of the Glasgow & Fife drug courts