Recidivism trends in California: New CDCR report

Recidivism rates are down, according to a new report by the California Department of Corrections and Rehabilitation (CDCR). In a press release accompanying the report, they state that “the report marks the second year of data showing the effects of the passage of Proposition 57, and the findings point to lower recidivism rates for those who earned credits from participation and completion of rehabilitative programming.”  But this statement is misleading. While the data showed a slight decrease in recidivism rates, correlation does not equal causation, and this would be an overly simplistic interpretation of the data. There are other factors that could have contributed to recidivism rates, such as the COVID-19 pandemic, which influenced crime rates and caused many court closures and temporary suspensions of intakes and transfers to CDCR, which likely influenced recidivism measures in this report. The report did not rigorously evaluate the impacts of Prop 57, and therefore, the findings are not sufficient to demonstrate a casual relationship between Prop 57 and reduced recidivism rates.

The report examines 36,086 people released during Fiscal Year 2018–19. The primary measure of recidivism is the three-year conviction rate, with arrests and returns to prison provided as supplemental measures. The report presents overall recidivism rates and attempts to make claims about the impact of recent policies and initiatives, though the latter seems misguided.

The three-year conviction rate for this cohort is 41.9%, with 58.1% having no convictions, 20.9% convicted of felonies, and 21.1% convicted of misdemeanors. The rate decreased by 2.7 percentage points from the previous year.  Additionally, supplemental measures of recidivism, such as the three-year arrest rate and the three-year return-to-prison rate, also saw declines between the 2017-18 and 2018-19 release cohorts. The three-year arrest rate decreased from 68.4 percent to 66.7 percent, and the three-year return-to-prison rate decreased from 19.8 percent to 16.8 percent.

The authors claim that this drop is due to impacts related to Proposition 57 and associated credit-bearing programs. However, this is a bold claim to make, because the report did not rigorously evaluate Prop. 57’s effectiveness and did not adequately eliminate other potential explanations for the decrease in recidivism. For example, the COVID-19 pandemic likely impacted recidivism rates by causing interruptions in arrests and convictions, court closures, and temporary suspensions of intakes to CDCR.

Rather, the recidivism report is a routine statistical report that the CDCR produces annually. Its main purpose is to present findings on the recidivism rates of released offenders, not to assess the impacts of policies or programs. While these reports provide valuable insights, it is essential to exercise caution and consider the limitations of the data before drawing conclusive statements based on them.

The report also examines recidivism rates based on whether inmates participated in credit-bearing programs. Those who earned any type of programming credit had slightly lower recidivism rates than people with no enhanced credit earnings (39.2 percent versus 45.6 percent). While this may seem like evidence of program effectiveness, this conclusion may be premature. The data may suffer from selection bias, as inmates self-select into these programs. When people self-select into groups, there are likely differences between those who choose to participate and those who don’t. For example, those who chose to participate in programs may already be better candidates for rehabilitation and “going straight” than those who chose not to. Thus, the reduced recidivism rates among those receiving enhanced credits cannot be attributed to program exposure. Future reports should explore these differences more thoroughly.

Individuals who spent more time in prison had a lower likelihood of recidivism. Overall, individuals in the FY 2018–19 cohort who spent less than one year in prison had a three-year conviction rate of 49.5 percent, compared to 7.5 percent for those who spent 15 years or more in prison. This was especially apparent for people with violent offenses. This finding suggests that longer lengths of stay were associated with a lower likelihood of recidivism, particularly for violent offenses. 

Conclusion

The impact of Proposition 57 on recidivism remains a complex and evolving question that requires a more rigorous investigation. The recent CDCR recidivism report indicates a small reduction in recidivism rates, but it did not demonstrate any causal relationship with Prop 57 or with any one factor, for that matter. The exact reasons for the drop in recidivism are still somewhat unclear, but it was likely due to many different factors, including disruptions caused by the COVID-19 pandemic, among others. At this time, the drop in recidivism cannot be attributed to Prop 57, nor can it be attributed to CDCR’s rehabilitative programs.

Drawing broad conclusions from this data could lead to ineffective or misguided policies in the realm of criminal justice, potentially jeopardizing public safety. There is currently a need for a more rigorous evaluation of Prop. 57 as well as the various CDCR rehabilitation programs. It is essential to conduct rigorous studies that systematically evaluate policies and programs.