Crime and Consequences has a guest post entitled “the Drug Court Fraud” here. Usually when I see/hear “fraud” I perk up – pretty strong language. Indeed, the language is far too strong, and much of the author’s case does not hold water. Yes, once again we’re back to evidence-based policy, but in this case the author is closing his eyes to much evidence in support of drug courts.
Steven Erickson argues in the brief posting that there are serious questions about drug courts, Effectiveness questions and Policy questions. Let’s look at effectiveness today.
Erickson argues that many studies are fatally flawed by:
- Intent-to-Treat analysis is not performed, meaning only program completers were part of the treatment group, not drop outs, thus inflating the success rates.
- Comparison Groups are missing or not truly comparable.
- Follow-up periods are too short.
- Sweeping Conclusions of success are belied by low completion rates.
The key to Erickson’s argument is the use of the terms “many studies” and “most studies” when citing fatal flaws. But the Campbell Crime & Justice Coordinating Group here discusses the scientific standards more completely than Erickson, and the means to weed out flawed studies. Systematic reviews are the key to screening out the flawed studies that Erickson rejects:
“Unlike traditional ways of reviewing and summarizing research on ‘what works’, systematic reviews use scientific and explicit methods to identify, screen, appraise and analyze evaluation studies. This kind of rigorous review produces the most reliable evidence on what the science says about a particular question.”
Meta-analysis is then used to survey the cumulative results of the studies that pass muster, often carefully controlled experiments.
Such a meta-analysis was conducted by David B. Wilson, OJ Mitchell, and Doris MacKenzie here. Here are there conclusions:
- Drug courts are an effective policy for handling drug offenders
- Best estimate: offending reduced approximately 24%
- Drug courts generally cheaper than alternatives (e.g., prison)
- This meta-analysis supports a policy of continued expansion of drug courts
(By the way, if a new drug protocol improved survival rates by "24%", would Erickson call that an absurdly "sweeping conclusion."
As "Drug Courts: The Second Decade," (see NCJRS) concludes recently “a number of randomized and controlled experimental studies published in peer-reviewed journals have found that drug court graduates have significantly lower rearrest rates—lasting more than 2 years beyond graduation—than those who do not participate in the program.”
So “many” studies are not fatally flawed, and in fact display some of the highest standards of social scientific evidence. So much so that one author, John Goldkamp, an acknowledged expert in the field, considers the question whether drug courts work the wrong question. Goldkamp’s presentation at NASC here mentions at least 10 reputable studies in what is not intended to be a comprehensive list. He is so certain of the positive findings of enough reputable studies that he does not ask “Whether” drug courts work, but rather “How” do they work and “Why” do they work, when they work.
I will mention one of the most reputable published studies with positive results:
“Long-term effects of participation in the Baltimore City drug treatment court: Results from an experimental study” in the Journal of Experimental Criminology (April 2006) by Denise C. Gottfredson, et. al. This study exhibits NONE of Erickson’s fatal flaws with a careful comparison group and three year follow-up.
Erickson, citing Judge Hoffman from North Carolina, argues that drug courts are bad public policy:
- Drug courts diminish the advocacy role of defense counsel.
- Judges are called to be treatment leaders, yet lack the training and credentials to oversee treatment administration.
This will be the subject of a later post – maybe one of you will volunteer to take this on? Suffice it to say that he may have a stronger case here.