Decoding Prop 47: What recent audit findings reveal about the impact in San Bernardino and Riverside counties

If you’ve been following California’s criminal justice reforms, you’re likely familiar with Proposition 47. Passed in 2014, Prop. 47 reclassified certain nonviolent offenses from felonies to misdemeanors, with the main goal of alleviating pressures on the state’s prison system by decreasing incarceration rates for nonviolent crimes. Reclassified offenses include drug possession, forgery, as well as instances of burglary, theft, and shoplifting wherein the stolen property amounts to less than $950.

Proponents believe that reducing incarceration rates for nonviolent offenses can lead to better rehabilitation outcomes and more efficient use of public resources, while critics argue that it has caused an increase in theft and drug-related crimes as well as an increase in reoffending. A recent audit by the Joint Legislative Audit Committee delves into how this measure has played out in San Bernardino and Riverside counties. In this post, I’ll break down the key takeaways from this extensive audit.

The audit analyzed the impact of Prop. 47 in Riverside and San Bernardino using data from county superior courts and three law enforcement agencies: the San Bernardino and Riverside local police departments, as well as the Riverside County Sheriff. Unfortunately, the San Bernardino County Sheriff’s Department was excluded from the audit due to a ransomware attack in the spring of 2023, which rendered much of its data unavailable. Together, the three agencies serve a population of nearly three million individuals (2,418,000 in Riverside County, 315,000 in the city of Riverside, and 222,000 in the city of San Bernardino).

To assess the impact of Prop. 47, the researchers compared two groups of people arrested and convicted for crimes that Prop. 47 reclassified. The first group included 107 randomly selected individuals convicted pre-legislation (in 2011), while the second group included 114 randomly selected individuals convicted post-legislation (in 2016).  The main outcomes included average sentence lengths, recidivism rates, and specific trends among those who recidivated. Additionally, the study examined police arrests and public calls for service.

Sentence Length

Researchers used court records to determine sentence lengths for individuals in the observed sample. They found that average sentence lengths for Prop. 47 offenses decreased by 43% (from an average of 349 days to 199 days), which is in line with the law’s intended purpose of reducing sentence lengths for nonviolent crimes. The drop was largely due to reductions in sentences for theft and drug possession. The average sentence length for drug possession dropped by 78% (429 to 96 days), while the average sentence length for theft declined by 38% (335 to 207 days). In contrast, the average sentence length for burglary offenses increased by 125% (268 to 603 days).

Recidivism

Researchers used the Bureau of State and Community Corrections (BSCC) definition for recidivism, which was measured as any re-conviction (whether felony or misdemeanor) within three years of release. The researchers used public records from the respective county courts to determine whether the individuals in each group reoffended within three years, and, if so, how many subsequent convictions they had within that period. The overall recidivism rate decreased in the observed sample, with three-year re-conviction rates dropping from 60% to 55%. This decrease was not statistically significant, though it generally aligns with statewide recidivism estimates (54% for the earlier cohort and 48% for the later cohort).

However, for those with extensive criminal histories, there was a notable, statistically significant increase in recidivism. When looking at the subset of people with four or more prior convictions, the recidivism rate increased by 12 percentage points, from 31% to 43%. Theft convictions among these repeat offenders saw a particularly sharp increase, rising from 25% to 39%. The researchers estimated that this rise likely equates to about 232 additional individuals who reoffended after Prop. 47’s implementation.

This trend highlights challenges in the law’s approach to managing repeat offenders. Law enforcement personnel agreed with this sentiment, commenting that Prop. 47 created an incentive for individuals to commit multiple offenses by reducing the consequences for repeat offenses. For example, when it comes to retail theft, the reduced penalties under Prop. 47 could make people more inclined to engage in such activity due to the perceived lower consequences.

Police Calls for Service and Arrests

To understand Prop. 47’s impact on law enforcement operations, the researchers examined data on calls for service and arrests. Calls for service that were related to Prop. 47 offenses comprised about 7% of total calls during the observed period, but this varied slightly between agencies. From 2014 to 2022, Riverside County’s calls for service related to Prop. 47 offenses decreased by 4%, while unrelated calls for service decreased by 2%. The Riverside Police saw a 7% increase in Prop. 47-related calls for service, while unrelated calls for service increased by 35%. The San Bernardino police department noted a 50% decrease in Prop. 47-related calls for service and a 30% decrease in unrelated calls for service. According to this data, it’s unclear if Prop. 47 had any effect on calls for service.

The data also confirmed that all three law enforcement agencies arrested fewer individuals for Prop. 47 crimes following the law’s implementation. Interestingly, a similar trend was also observed for non-Prop. 47 arrests. In Riverside County, Prop. 47-related arrests dropped by 30% since 2014, while unrelated arrests dropped by 21%. In the city of Riverside, arrests for Prop. 47 crimes decreased by 67%, while unrelated arrests dropped by 15%. In the city of San Bernardino, Prop. 47-related arrests declined by 60%, while unrelated arrests declined by 29%.

Recommendations

In the report, California state auditor Grant Parks expressed concerns regarding the number of individuals with multiple convictions who continue to engage in theft. Based on the study’s findings, the auditor had two key recommendations for the California Legislature. Specifically, they recommended the adoption of the following policy mechanisms:

1. “Appropriating funds for programs and services to address what it determines to be the root causes of recidivism. In particular, it may want to identify and provide funding for the types of programs and services that recent studies have identified as effective in reducing recidivism.”

2. “Either by enacting legislation that is consistent with and furthers the intent of Prop. 47, or by enacting a proposal for approval by the voters that would amend Prop. 47, to allow more severe punishment for those convicted multiple times of theft, likely by identifying such offenses as potential felonies instead of automatically classifying them as misdemeanors.”

Discussion

Some stakeholders advocated for even more drastic changes. For example, Riverside County Sheriff Chad Bianco felt that the report should have included a recommendation to repeal Prop. 47. In a written letter, he objected to the audit, stating that he felt the report had mischaracterized some of the findings and conclusions related to Riverside County. Specifically, he questioned the validity of the audit’s measurement of “organized retail theft” and was concerned that they may have mischaracterized certain types of theft. The agency does not have a specific category denoting whether an offense consists of “organized retail theft.” Rather, all thefts are classified as petty, grand, fraud, and/or burglary. To  accurately extrapolate organized retail theft from the data, the researchers would have had to analyze each individual case, which they did not do.

Bianco also felt that the report minimized the experience of deputies and business owners who have felt the impact of Prop. 47 on their communities: “Nowhere in your report are statements from local business owners who have been burglarized, robbed, or stolen from. Nowhere in your report does it account for businesses that have amended their policies to no longer call the police. For example, large corporations like Target generally do not call law enforcement for petty thefts anymore due to Prop. 47’s theft reclassifications. Without these accounts and data, this report paints a rosier picture for the public, and it is deceiving.”

In the letter’s closing, Bianco reiterated the idea that Prop. 47 should be repealed. Scathingly, he claimed that the audit was subjective and biased to align with Governor Newsom’s proposals: “More disturbing is that this report caters directly to Governor Newsom’s recent announcement and mandate to the legislature to build upon his ‘Real Public Safety Plan’ by recommending stronger penalties and adjustments to theft crimes. His proposal provides a framework to enhance current laws but falls short of recommending a complete repeal of Prop. 47. Interestingly, this report recommends those same proposals, but does not provide an alternate recommendation or suggestion to the legislative body to repeal Prop. 47. Without this recommendation to the legislature, this report is nothing other than Governor Newsom’s proposal in different clothing. California demands objective audit reports that contain accurate and unbiased data. This report fails to meet those demands; it is highly subjective, not accurate, and it is swayed by Governor Newsom‘s political oversight of your office.”

Meanwhile, advocates of Prop. 47 argue that the associated cost savings have funded programs that have effectively reduced recidivism, homelessness, and unemployment among participants. In a statewide assessment conducted by the BSCC, advocates praised the program as successful because recidivism among participants enrolled in services was lower (15%) than other reported statewide recidivism estimates (which ranged from 35% to 45%). The BSCC also noted a 60% decrease in homelessness and a 50% decrease in unemployment among a subset of participants who received services and identified housing or employment as goals of their participation.

However, there are some flaws in that reasoning. First, comparing the recidivism of participants in Prop. 47 programs to statewide recidivism estimates is not an adequate, apples-to-apples comparison. The baseline conditions of individuals in the Prop. 47 programs could be different from the wider population of offenders. For example, participants might be individuals who were already less likely to reoffend due to other factors, such as improved personal circumstances or pre-existing support networks. Relatedly, there is self-selection bias. In other words, participants in the Prop. 47-funded programs are likely to be a self-selected group, meaning they may already be more motivated or have different characteristics compared to the general population of individuals who do not have access to or choose not to participate in such programs. This self-selection can lead to an overestimation of the program’s effectiveness since those who choose to participate may inherently have better outcomes due to their own characteristics or motivations. Further, the statements about housing and employment are somewhat misleading from a research standpoint. First, this is only a very specific subset of the larger group, and second, there is no real comparison group; the authors are simply measuring the percentage change within one group. For more about the outcomes of the various Prop. 47 grantees, check out my previous blog post for a thorough review of the data.

Advocates also like to cite a 2020 study published in Criminology as evidence of Prop. 47’s success while clearly ignoring important parts of that study’s findings. Researchers Bird, Nguyen, and Grattet compared people released pre- and post-Prop. 47 in terms of recidivism rates. They found that individuals in the Prop. 47 drug possession group had statistically significant reductions in rearrests and reconvictions for drug offenses at the one-year follow-up, relative to those in the comparison group. However, very unfortunately, the Prop. 47 group also had a statistically significant increase in rearrests for crimes against persons (specifically, assaults and domestic violence), as well as higher reconviction rates for assault with a deadly weapon, compared with the pre-Prop. 47 group. There were no statistically significant differences between groups in rearrests or reconvictions for property offenses.

Conclusion

Prop. 47 has sparked significant debate since it was passed in 2014. The recent audit in Riverside and San Bernardino counties reveals a mixed picture: while overall recidivism rates showed a slight decline, there were notable increases in recidivism among individuals with multiple prior offenses.

Critics, such as Riverside County Sheriff Chad Bianco, argue that the report is heavily biased towards Governor Newsom’s political agenda rather than providing an objective assessment. In contrast, advocates argue that Prop. 47 has been successful in reducing recidivism. However, the studies that have attempted to evaluate this claim have been methodologically weak and do not always have straightforward results. For example, one research study cited above found that, while overall recidivism rates decreased among Prop. 47 offenders, crimes against persons (specifically assaults and domestic violence) actually increased by a substantial amount.

As California continues to navigate the complexities of criminal justice reform, it is clear that Prop. 47 has generated challenges and controversies that demand further scrutiny and balanced solutions.