Evidence-based alternative models of incarceration, systematic review

Restorative justice (RJ) approaches to criminal behavior provide a holistic alternative to the punirtive models that exist in the United State. RJ "views offenses in terms of relationships and pays particular attention to victim and community needs. These alternative RJ approaches have been shown to decrease recidivism and align with the values of those who have been most impacted by mass incarceration, including Indigenous populations. The purpose of this systematic review is to provide an overview of alternative models of incarceration utilizing RJ principles that could be adapted for a largely Indigenous population." 

"We searched three databases, PubMed, PsycNet, and Google Scholar, resulting in a total of 1904 records after duplicates were removed. We used Cadima, an online tool for systematic reviews. Two research team members independently reviewed titles and abstracts for relevancy. The review of records began with a consistency check of the criteria on 10% of the records (n = 190). The results of the consistency check were a Kappa value of 0.7216796875 or a “good” strength of agreement. The next two rounds of the review included the application of criteria, first to title/abstracts, followed by full-text records. A portion of the records in each round underwent parallel assessment or were assessed by both reviewers, one round with 40% of records in the title/abstract (n = 761) and one with 20% (n = 171) in the full text. Conflicts between reviewers were discussed and resolved through consensus."

"Our inclusion criteria for eligible papers included (1) a focus on Indigenous, restorative, or therapeutic justice, (2) sufficient information on an alternative incarceration model to be recreated, (3) written in English, and (4) published between 2012 and 2022. In our review, we included both peer-reviewed and “gray” literature (i.e., governmental reports, research outcomes, investigative journalism, etc.). Studies that did not meet basic scientific rigor (e.g., studies that were not replicable due to a lack of information or where the evaluation criteria was susceptible to bias, especially in participant selection) were not included. After several rounds of review, a total of 120 articles remained for data extraction."

"Data were extracted from 120 articles in order to screen for the final inclusion criteria, requiring sufficient information on an alternative incarceration model for potential replication. To be considered sufficient, data were needed in the following areas: intervention components, restorative justice components (i.e., as outlined by the aforementioned definitions of RJ), physical design elements, partnering agencies, and information on findings. A total of 41 articles met this criteria and were then combined into eight categories of analysis."