Effectiveness Evaluation of the "Expanded Addiction Treatment" Policy Should Compare Existing Dispositions

Following the conclusions of the National Conference on Judicial Reform, the Executive Yuan is planning the "Expanded Addiction Treatment" policy. In addition to increasing the proportion of drug offenders receiving treatment, it is expected to include drunk-driving offenders. While international criminological research supports prioritizing treatment for drug use and drunk driving, there is currently a lack of sufficient domestic research to support whether the implementation of addiction treatment in Taiwan since its legalization in 2008 has been more effective than other dispositions in reducing drug recidivism.

This article first calls upon policy planners to allocate resources for conducting more and larger-scale effectiveness evaluation studies comparing addiction treatment with other dispositions. The importance of these studies lies in the fact that, according to existing statistical data, reducing recidivism rates for drug use and drunk driving may help lower the overall recidivism and crime rates. Therefore, if it can be proven that addiction treatment is indeed more effective than other dispositions in preventing these two types of recidivism, not only will the "Expanded Addiction Treatment" policy have a solid foundation, but overall crime and recidivism phenomena are also expected to be mitigated.

Next, this article will explain that:

1. Such research requires long-term tracking and continuous execution;
2. We should seize the time to begin research before existing systems disappear or are substantially replaced.

The Need for Long-Term Research

Collecting empirical data requires patience, and observations regarding behavioral changes often take years of monitoring. Criminological observations of recidivism generally require more than two years to be relatively reliable, while drug offenders require even longer observation because the physical and psychological dependence caused by drugs is harder to correct than other behaviors.

According to Taiwan's existing data, the annual marginal recidivism rate for individuals undergoing compulsory drug abstention and treatment does not drop below 5% until the fourth year after leaving the facility, and it remains at 3% in the fifth year. From this, it can be inferred that drug patients undergoing addiction treatment may need to be observed for 3 to 5 years for the figures to be reliable. If the study ends too early, it may fail to observe the full picture of relapse (recidivism) phenomena, thereby leading to incorrect inferences about policy effectiveness. The study by Ch'en Ch'üan-hsi et al. (2012) cited in the first part of this series had a follow-up period of only 18 months, which is somewhat insufficient based on empirical research experience.

Regarding alcohol cessation, the follow-up period in the study by K'o Hui-chen et al. (2001) cited in the first part of this series only lasted until 3 months post-treatment, which is also unfortunately too short (though on page 11 of that paper, a proposal to extend observation by 12 months can be found). Policy planners should consult medical experts on how to determine successful alcohol cessation and how long observation should last—whether 3 months, 6 months, or longer is sufficient. This article focuses on the issue of drunk-driving recidivism: regarding whether alcohol cessation treatment reduces drunk-driving recidivism rates, it is similarly recommended to track each offender for 3 to 5 years.

Since drug use and drunk driving do not appear at a single point in time after the new policy takes effect, but rather emerge continuously and will never completely disappear, the timing and subjects of effectiveness evaluations should not be limited to offenders within the first year or two of the new policy's implementation. Instead, they should become ongoing studies. This will allow for the assessment of whether the effectiveness of measures increases, decreases, or remains stable over time, and makes it possible to more accurately identify the key factors influencing effectiveness.

The Need to Immediately Study the Effects of Existing Systems

The foregoing points out the shortcomings in existing research on addiction treatment, which can serve as a reference for improving new research. What is frequently overlooked—and thus deserves special emphasis—is that one must thoroughly understand the effectiveness of old policies in order to objectively evaluate new ones. If we wait until new policies completely replace old ones before starting research, the circumstances of the old policies can no longer be replicated, and new policies will lose their comparison baseline, leaving us to speculate on the relative effectiveness of old versus new policies based solely on subjective impressions of "reminiscing about the good old days." For example, if we wait until compulsory drug abstention and treatment disappears before studying it, the recidivism risk and treatment details of those subjected to it can no longer be measured.

If the old system is substantially replaced, a sample size that is too small will undermine the statistical power of the research. For instance, since the legalization of addiction treatment in 2008, the annual number of admissions for compulsory drug abstention and treatment has drastically decreased by five-sixths, with only 1,100 to 1,200 individuals per year in the last three years, and the annual number of discharges has dwindled to just 600 to 700. If this declining trend continues, we will soon face the problem of insufficient sample size.

Whether the old system is completely or substantially replaced, both scenarios make it difficult to accurately compare the relative effectiveness of compulsory drug abstention and treatment versus addiction treatment. Similarly, in order to objectively compare the effectiveness of short-term detention/sentencing versus addiction treatment—that is, to prove that addiction treatment is truly better—effectiveness evaluations of existing systems must be conducted. Furthermore, given that such studies require a 3- to 5-year follow-up period, and that the new policy may not be implemented that far in the future, these studies must be conducted now to avoid missing the opportunity to obtain critical data.