Proposition 47 in Riverside
and San Bernardino Counties
The Number of Individuals With Four or More
Convictions Increased, as Did Theft Convictions
Among the Same Group
July 
REPORT 
For questions regarding the contents of this report, please contact our Communications Office at ..
This report is also available online at www.auditor.ca.gov | Alternative format reports available upon request | Permission is granted to reproduce reports
 Capitol Mall, Suite  | Sacramento | CA | 
CALIFORNIA STATE AUDITOR
.. | TTY ..
...
For complaints of state employee misconduct,
contact us through the Whistleblower Hotline:
Don’t want to miss any of our reports? Subscribe to our email list at
auditor.ca.gov
621 Capitol Mall, Suite 1200 | Sacramento, CA 95814 | 916.445.0255 | 916.327.0019 fax | www.auditor.ca.gov
Mike Tilden Chief Deputy
Grant Parks State Auditor
July , 
2023‑107
e Governor of California
President pro Tempore of the Senate
Speaker of the Assembly
State Capitol
Sacramento, California 
Dear Governor and Legislative Leaders:
As directed by the Joint Legislative Audit Committee, my office conducted an audit of the
effects of Proposition  in San Bernardino and Riverside counties. Our assessment focused
on a selection of three law enforcement agencies from the two counties and a review of
twostatisticallyvalid samples of individuals convicted of offenses that were later reclassified by
Proposition  offenses—one from  and one from . Our statistical analyses were limited
to the jurisdictions of these three law enforcement agencies.
We reviewed three indicators—public calls for service, law enforcement response times, and
arrests—to assess the impact that Proposition  had on San Bernardino and Riverside counties.
We did not identify any conclusive trends in calls for service or response times. However, we
found that all three agencies arrested fewer individuals for Proposition  offenses—between
percent and  percent fewer—in the period following the implementation of the proposition.
An analysis of our samples indicates that rates of recidivism—conviction of a new felony or
misdemeanor committed within three years of a previous conviction—in the two counties
generally decreased after the implementation of Proposition . However, the number of
individualswho were convicted four or more times increased by  percentage points. In addition,
theft convictions among those who reoffended four or more times increased by  percentage
points after the implementation of Proposition . e law enforcement agencies we reviewed
asserted that reduced sentences are driving the increase in individuals committing multiple
offenses in general and multiple theft offenses in particular.
We also found that Proposition  is likely fulfilling one of its intended purposes by contributing
toreducing the length of incarceration for those who have committed certain nonviolent crimes.
An analysis of our samples shows that the average sentence length for Proposition  offenses
decreased by more than  percent after the proposition’s implementation. We recommend
that the Legislature consider the adoption of either or both of the following policy mechanisms:
appropriating additional funds for services that eliminate the root causes of recidivism or
increasing penalties for individuals who are convicted of four or more repeat offenses.
Respectfully submitted,
GRANT PARKS
California State Auditor
Selected Abbreviations Used in This Report
ACLU American Civil Liberties Union
BSCC Board of State and Community Corrections
CAD computer-aided dispatch
CDCR California Department of Corrections and Rehabilitation
CDI California Department of Insurance
DOJ Department of Justice
Riverside DA Riverside County District Attorney’s Office
RMS record management system
iv CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Contents
Summary 1
Introduction 3
Audit Results
The Law Enforcement Data We Reviewed Do Not Clearly
DemonstrateWhetherProposition 47 Significantly Affected
ThreeCrimeIndicators 9
Following the Implementation of Proposition 47, the Numberof
IndividualsWhoReoffended With Four or More Convictions
IncreasedattheThree Law Enforcement Agencies We Reviewed 15
Our Review Indicates That Proposition 47 Has Achieved Some
ofItsIntendedPurposes 19
California Has Several Options for Addressing the Increases in
MultipleReoffenses and Theft 21
Recommendations 22
Other Areas Reviewed 23
Appendix A
Volume of Service Calls for Alleged Proposition 47 Offenses
by Crime Category 25
Appendix B
Number of Arrest Charges at Each Law Enforcement Agency for
AllegedProposition 47 Offenses by Crime Category 27
Appendix C
Recidivism Methodology 29
Appendix D
Scope and Methodology 31
Response to the Audit
Riverside Sheriff 33
California State Auditor’s Comments on the Response
From Riverside Sheriff 37
vCALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Blank page inserted for reproduction purposes only.
vi CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
SUMMARY
Key Findings and Recommendations
In  California voters approved the Safe Neighborhoods and Schools Act
(Proposition) to ensure that the State would focus its prison spending on
violent and serious offenses; to maximize alternatives other than incarceration
for nonviolent crime;and to create savings that the State could invest in victim
services, prevention and support programs, and mental health and drug treatment.
To achieve these goals, Proposition reclassified certain theft, burglary, forgery,
and drug possession offenses (referred to in this report as Proposition offenses),
from potential felonies to misdemeanors. e Joint Legislative Audit Committee
(AuditCommittee) requested the California State Auditor to conduct an audit
of the effects of Proposition in SanBernardino and Riverside counties. e
audit, as requested, was limited in scope to the two counties, and the results of
our analysis should not be projected statewide. To address our audit objectives,
we reviewed a selection of three law enforcement agencies from SanBernardino
and Riverside counties and reviewed two statistically valid samples of individuals
convicted of offenses that were later reclassified by Proposition—one from 
and one from .
Our review found the following:
• We reviewed three indicators—public calls for service, law enforcement
response times, and arrests—to assess the impact that Proposition had on
SanBernardino and Riverside counties. For the three law enforcement agencies we
reviewed, we did not identify any conclusive trends in calls for service or response
times either before or after the implementation of Proposition. However, we
found that all three agencies arrested fewer individuals for Proposition offenses
in the period following the implementation of the proposition. Specifically, our
analysis of data from the three law enforcement agencies found that following the
implementation of the proposition, arrests for Proposition offenses declined
between  and  percent, and arrests for all other offenses declined between
and  percent depending on the agency.
• An analysis of two sample groups of individuals, one group convicted in 
(prior to Proposition ) and another group in  (after Proposition ),
indicates that rates of recidivism—conviction of a new felony or misdemeanor
committed within three years of a previous conviction—inthe two counties
generally decreased slightly after the implementation of Proposition. We refer to
recidivism throughout the report as reoffense. However, the number of individuals
who were convicted four or more times increased by  percentage points. In
addition, theft convictions among those who reoffended four or more times
increased by  percentage points after the implementation of Proposition. e
law enforcement agencies we reviewed asserted that reduced sentences are driving
the increase in individuals committing multiple offenses in general and multiple
theft offenses in particular.
1CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
• Proposition is likely fulfilling one of its intended purposes by contributing
to reducing the length of incarceration for certain nonviolent crimes. An
analysis of our two sample groups shows that the average sentence length for
Proposition offenses decreased by more than  percent after the proposition’s
implementation. In addition, according to stakeholder groups advocating for
reforms to the criminal justice system, programs funded by Proposition savings
have decreased recidivism, homelessness, and unemployment among participants.
erefore, we recommend that the Legislature consider the adoption of either
or both of the following policy mechanisms: appropriating additional funds for
services that eliminate the root causes of recidivism and/or increasing penalties for
individuals who are convicted of four or more repeat offenses.
Agency Comments
Because we did not make recommendations to the three law enforcement agencies
we reviewed, we did not expect to receive written responses from them; however,
we did receive a response from the Riverside County Sheriff’s Office. It disagreed
with and mischaracterized some of our findings and conclusions. We excluded the
SanBernardino County Sheriff’s Department—the county’s largest law enforcement
agency—from our audit because the department experienced a ransomware attack in
the spring of  that rendered much of its data unavailable.
2 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
INTRODUCTION
Background
In  California voters approved the Safe
Neighborhoods and Schools Act (Proposition) to
ensure that the State would focus its prison spending
on violent and serious offenses; to maximize the use of
alternatives other than prison for nonserious, nonviolent
crime; and to invest the savings realized as a result of the
proposition into victim services, prevention and support
programs in K– schools, and mental health and drug
treatment. e text box describes the main provisions of
Proposition.
Proposition47’s Reclassification of Offenses
Proposition reclassified certain theft, burglary, forgery
and drug possession offenses, which we refer to as
Proposition offenses, from felonies and wobblers—
a term commonly used to describe a crime that can
be charged as either a felony or a misdemeanor–to
misdemeanors. Felonies are crimes that are punishable
by death, imprisonment in state prison, or imprisonment
in a county jail for more than a year. Misdemeanors are
all other crimes except infractions.
1
epenalties for
misdemeanors are generally imprisonment in county
jail, fines, or both. Asentence of imprisonment for
misdemeanors is generally no more than  months in
county jail but can sometimes be up to a year.
Before Proposition, individuals who committed
offenses that were later reclassified by Proposition 
could receive felony punishments for those offenses.
Such individuals now generally receive misdemeanor
punishments. roughout this report, we separate
Proposition offenses into four categories: burglary,
theft, drug possession, and forgery.
2
We categorized
burglary and theft separately for the reasons we describe
in the text box. Figure  provides examples of offenses
that Proposition reclassified.
1
Prosecutors and courts can choose to designate certain criminal offenses,
such as willfully violating promises to appear in court, disturbing the peace,
or driving without a valid license, as infractions. An infraction is not punishable
byimprisonment.
2
State law divides theft into two degrees: grand and petty. Grand theft can incur
more severe punishment than petty theft.
Proposition 47 Made Three Major Changes
1. It reclassified certain nonviolent offenses as misdemeanors
instead of felonies or wobblers (a term commonly used to
describe a crime that can be charged as either a felony or a
misdemeanor). These reclassified offenses include certain
types of theft and drug possession unless the defendant
has prior convictions for certain violent or serious crimes.
2. It authorized courts to resentence with misdemeanors those
individuals who were serving felony sentences for certain
nonviolent offenses. For these individuals, Proposition 47
required a thorough review of criminal history and a risk
assessment to ensure that these individuals did not pose a
risk to public safety. Proposition47 also authorized courts
to resentence individuals who had already served a felony
sentence for a Proposition 47 offense to designate that
offense as amisdemeanor.
3. It created the Safe Neighborhoods and Schools Fund
that allocates funds saved from the implementation
of Proposition 47 to mental health and substance
abuse treatment programs, support programs aimed
at improving outcomes for public school students, and
recovery services for crime victims.
Source: State law.
The Definition of Burglary
Versus the Definition of Theft
Burglary is generally defined as:
• Any person who enters certain enclosures such as a
house or commercial establishment with the intent to
commit theft or certain other offenses, such as arson or
assault with a deadly weapon, is guilty of burglary.
Prior to Proposition 47, burglary included:
• Entering a commercial establishment with the intent
to commit theft of property whose value does not
exceed$950.
Proposition 47 defined this type of burglary as shoplifting
and established it as a new misdemeanor crime.
Because shoplifting as it became defined is not a type of
burglary, and is similar to theft, we classify shoplifting as
theft in our Proposition 47 analysis. All other burglary crimes
remained categorized as burglary.
Source: State law.
3CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Figure 1
Examples of Offenses That Proposition 47 Reclassified
Burglary
e
Forgery
Possession
Jonathon Doe $$$
Category
*
Offense Example
Pre-Proposition 47
Classification
Post-Proposition 47
Classification
Theft of property
worth $950 or less
from a store or shop.
Felony or
misdemeanor
burglary
Grand theft of
personal property
worth $950 or less.
Felony or
misdemeanor
theft
Misdemeanor
shoplifting
(which we
categorize as
theft for purposes
of our audit)
Misdemeanor
petty theft
Signing the name of
another individual
on a check worth
$950 or less with the
intent to defraud.
Felony or
misdemeanor
Misdemeanor
Possession of certain
controlled substances
for personal use
without a valid
prescription.
Felony or
misdemeanor
Misdemeanor
Source: State law.
Note: The figure provides examples of reclassified offenses, not a complete list of all reclassified offenses. In addition, if an
individual who has previously been convicted of certain severe crimes, such as murder, commits a Proposition 47 offense,
that new offense may still be charged as a felony.
* Refer to the text box on page 3 titled The Definition of Burglary Versus the Definition of Theft for an explanation of why we
separate the two types of offenses.
4 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
The Safe Neighborhoods and Schools Fund
In addition to reclassifying certain crimes, Proposition created the Safe Neighborhoods
and Schools Fund. is fund contains the savings that accrue to the State from the
implementation of Proposition.
3
e California state controller disburses percent of
the fund to the California Board of State and Community Corrections (BSCC),  percent
to the California Department of Education, and  percent to the Victim Compensation
and Government Claims Board. e Department of Finance calculated that from fiscal
years – through –, the State saved an average of million per year
from the implementation of Proposition for transfer to the Safe Neighborhoods and
Schools Fund. All three of the agencies have reported using these funds to facilitate grant
programs; the grants are meant to help fund services, such as the management of trauma
recovery centers and the provision of mental health resources to people in the criminal
justice systemand survivors of violent crime.
Criminal Justice Processes Following the Implementation of Proposition47
State law allows a peace officer (officer) to arrest and hold an individual suspected of
a felony if the officer has probable cause. In contrast, officers generally must release
individuals arrested for misdemeanors after citing—writing a notice for them to appear
in court—or booking them. us, before the implementation of Proposition, a law
enforcement agency could choose to hold individuals suspected of Proposition
offenses after booking—fingerprinting and photographing a person to record an
arrest—because these offenses could be charged as felonies. Consequently, since the
implementation of Proposition, officers have generally been required to release
individuals suspected solely of committing Proposition offenses. However, the
officers can hold such individuals if a separate statutory reason exists to do so. For
example, a statute enacted in  authorizes officers to hold individuals who have been
arrested or convicted of theft from a store in the previous six months or when there is
probable cause to believe individuals are guilty of organized retail theft.
4
Changes to State Law Have Reduced Pressures on California’s Prison and Jail Systems
Proposition is one of several changes to state law that have contributed to a reduction
in California’s incarcerated population. In  the U.S. Supreme Court ruled that
California’s prison overcrowding violated the constitutional rights of incarcerated
individuals and required remedial action. Since that time, California voters have passed
several propositions that have contributed to reductions to incarcerated populations,
as the text box shows. For example, during fiscal year –, the California
Department of Corrections and Rehabilitation(CDCR) released over , individuals
who received resentencing from a court and were required to be released under
Proposition’sprovisions.
3
The director of Finance is responsible for calculating these savings and bases that calculation on the amount saved from
reduced prison use.
4
Organized retail theft became a criminal offense effective January , , several years after Proposition 
wasimplemented.
5CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
In addition, to address overcrowding, state
prisons and some county jails have at times had to
release incarcerated individuals who committed
nonviolent offenses. For example, since the s,
Riverside County has operated under a federal
court order that requires the county to release
individuals from its jails or refuse to accept
individuals who have been recently arrested
whenever all the beds in its jail system are filled,
and allows the county to start refusing to accept
individuals once any jail or specific housing unit
has reached  percent of its capacity. Riverside
County and other large counties in California have
released incarcerated individuals because of a lack
of bed capacity. Moreover, in April  during
the COVID- pandemic, the Judicial Council
of California established emergency rules that
included setting bail for misdemeanor and felony
offenses at , with some exceptions. However, the
Judicial Council of California rescinded these rules
inJune .
Since  the numbers of individuals in state
prisons and county jails have declined. We reviewed
data from the CDCR and the BSCC to analyze
changes in thepopulations of state prisons and
county jails from  through . e factors discussed above and in the text box
likely contributed to a percent decline in the state prison population and a percent
decline in county jail populations, as Figure illustrates.
Our Review of Law Enforcement Agencies in Riverside and SanBernardino Counties
e Joint Legislative Audit Committee (Audit Committee) requested that we review
aselection of law enforcement agencies in SanBernardino and Riverside counties.
Weinitially selected two sheriff’s departments and two police departments based on
the number of sworn officers they employed. However, we removed the SanBernardino
County Sheriff’s Department—the county’s largest law enforcement agency—from
our selection during the course of the audit because the department experienced a
ransomware attack in the spring of  that rendered much of its data unavailable. As
of May , the SanBernardino County Sheriff’s Department still did not have access
to some of the data that it lost because of this attack. We evaluated the data potentially
available at the statewide and local law enforcement level and found that local law
enforcement data contained additional information not captured in the statewide data.
erefore, we analyzed data from the three local law enforcement agencies we selected
for review. e audit was limited to the three law enforcement agencies in two counties,
and the results of our analysis should not be projected statewide. Table  shows the
number of sworn officers employed by the remaining three law enforcement agencies
andthe populations of the communities they serve.
Other Propositions Contributed to
Reductions in the Number of Individuals
IncarceratedinCalifornia
• Proposition 36 (2012): Removed certain increases in
punishment from prison sentences for certain individuals
subject to the State’s existing three‑strikes law. It also
allowed certain individuals serving life sentences to apply
for reduced sentences.
• Proposition 57 (2016): Expanded eligibility for parole
consideration; increased the State’s authority to award
early release due to good behavior and rehabilitative or
educational achievements; and mandated that judges,
instead of prosecutors, determine whether youth should
be transferred from juvenile court to criminal court for
certain offenses.
• Proposition 64 (2016): Legalized specified personal use
and cultivation of marijuana for adults 21 years of age or
older, reduced criminal penalties for specified marijuana‑
related offenses for adults and youth, and authorized
resentencing or dismissal and sealing of eligible
marijuana‑related convictions.
Source: State law.
6 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Figure 2
State Prison and County Jail Populations Have Declined From January 2011 Through September 2023
COVID-19 pandemic
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023*
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
Total prison in-custody
population as of December 31
in the given year
16%
36%
Total average daily
population for county jails
Source: CDCR; BSCC.
Note: The sharp decline in prison and jail populations from 2019 to 2020 coincided with the onset of the COVID-19 pandemic and relatedefforts,
such as reducing the number of individuals in prisons and jails.
* Jail Profile Survey for 2023 data consists of data from January 2023 through September 2023.
Table 1
Riverside and San Bernardino County Law Enforcement Agencies We Reviewed
SWORN
OFFICERS
AS OF
JULY2022
JURISDICTION
POPULATION
*
NEAREST
THOUSAND
OFFICERTO
RESIDENT
RATIO
Riverside County Sheriff’s Office
†
2,802 Riverside County 2,418,000 1:863
Riverside Police Department 349 City of Riverside 315,000 1:903
San Bernardino Police Department 253 City of San Bernardino 222,000 1:877
Source: U.S. Census, the Commission on Peace Officer Standards and Training, and Riverside County Sheriff’s Office.
* Population amounts are from the 2020 decennial census.
†
According to a chief deputy at Riverside County Sheriff’s Office, sworn officers can include officers of various job designations, such as peace
officers, deputy coroners, and court deputies. The chief deputy further explained that as of May 2024, the office has 2,910 officers, of whom
1,680 are peace officers.
7CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Blank page inserted for reproduction purposes only.
8 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
AUDIT RESULTS
The Law Enforcement Data We Reviewed Do Not Clearly Demonstrate Whether
Proposition47 Significantly Affected Three Crime Indicators
To assess the impact of the implementation of Proposition on SanBernardino and
Riverside counties, we reviewed three indicators: calls for service, response times,
and arrests, as Figure  shows. We did not identify any conclusive trends in calls for
service or response times across the law enforcement agencies that demonstrate that
the implementation of Proposition  had a significant impact on these indicators.
However, we found that all three agencies arrested fewer individuals for Proposition
offenses in the period following the implementation of the proposition.
Figure 3
Proposition 47 Offenses Generally Followed Overall Trends When Compared to All Other
Offenses but Showed Steeper Changes for Certain Indicators
Police
Calls for Service
Trends in the volume
of service calls for
Proposition 47 offenses
generally aligned
with trends for other
service calls.
Response Times
The law enforcement
agencies’ response times
have increased for all
offenses, including those
related to Proposition 47.
Arrests
Arrests for all offenses
decreased, but arrests for
Proposition 47 offenses
decreased more after the
implementation of
Proposition 47.
Source: Analysis of computer-aided dispatch system and record management system data from Riverside County Sheriff’s
Office, Riverside Police Department, and SanBernardinoPolice Department.
Trends in the Volume of Service Calls for Proposition 47 Offenses Have Generally Aligned
With Trends for Other Service Calls
e three law enforcement agencies use computer-aided dispatch (CAD) systems
toreceive and log calls for service. ese calls include those that the public makes to
 or to nonemergency lines for alleged criminal activities. We have included some
examples of calls for service in the textbox. In addition, the CAD systems log certain
other incidents, such as when an officer observes a crime in progress and reports
it from the field. We compared and analyzed the three agencies’ CAD-reported
activity, which we refer to as calls for service, for Proposition offenses and other
9CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
incidents from  to  for the Riverside Police
Department (RiversidePolice) and the Riverside
County Sheriff’s Office (Riverside Sheriff) and from
 to  for SanBernardino Police Department
(SanBernardino Police).
5
Calls for service for Proposition offenses—
including burglary, drug possession, forgery, and
theft—totaled about  percent of all calls for service
at the three law enforcement agencies during the
periods we reviewed. However, theother trends
we identified differed amongthe three agencies, as
Figure  shows. For example, calls for service for both call types decreased slightly at
Riverside Sheriff from  to . Meanwhile, Riverside Police experienced increases
in both Proposition-related service calls and other service calls from  to .
SanBernardino Police experienced an overall decrease incalls for service for both
call types from  through ; however, there was a steeper percentage drop in
Proposition-related service calls compared to other service calls. Appendix A shows
the total number of Proposition-related service calls for each of four categories:
burglary, drug possession, forgery, and theft.
It is unclear from the data what effect, if any, Proposition has had on calls for
service. e three law enforcement agencies stated that the calls for service might not
reflect the actual number of Proposition offenses occurring because the community
may be underreporting those offenses. For example, a captain at SanBernardino
Police, a lieutenant at Riverside Police, and a deputy at Riverside Sheriff asserted that
businesses are less likely to call the police to report thefts because even when they do
so, the same individuals often return the next day to reoffend. Nevertheless, all three
acknowledged that their perspective was largely based on anecdotal evidence and did
not provide evidence to substantiate that it was occurring. We contacted the California
Retailer’s Association to determine whether it compiles information on theft from
its members. However, the association informed us it did not have any data on how
frequently retailers reported thefts to law enforcement.
The Law Enforcement Agencies’ Response Times Have Increased for All Offenses, Including
Those Related to Proposition 47
e three law enforcement agencies’ response times generally increased after the
implementation of Proposition. e CAD system for each law enforcement agency
logs the time it receives a call and the time that a responding officer arrives at the scene
of the alleged incident. When we analyzed the three agencies’ average response times
from  to , we found that they increased for all service calls, including those for
Proposition offenses, as Figure  shows. It is unclear from the data what effect, if any,
Proposition has had on response times because of other contributingfactors, such as
law enforcement agency staffing levels and the number of calls for service.
5
San Bernardino Police’s CAD data are only complete dating back to .
Examples of Service Calls
• A welfare check
• A hit‑and‑run
• A robbery
Source: Riverside Police, Riverside Sheriff, and San Bernardino
Police CAD systems.
10 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Figure 4
Calls for Service Trends Varied Across the Law Enforcement Agencies
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Number of Calls for ServiceNumber of Calls for ServiceNumber of Calls for Service
0
100,000
200,000
300,000
400,000
500,000
600,000
700,000
800,000
900,000
1,000,000
0
50,000
100,000
150,000
200,000
250,000
300,000
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
400,000
Riverside Sheriff
Riverside Police
San Bernardino Police
*
4%
35%
15%
16%
5%
4%
Since 2014
2%
4%
Since 2014
Since 2014
Since 2014
6%
7%
50%
30%
Since 2014
Since 2014
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Non-Proposition 47-Related Calls for Service
Proposition 47-Related Calls for Service
Source: Analysis of Riverside Sheriff, Riverside Police, and San Bernardino Police CAD systems.
Note: We calculated the percentage change in total calls for service before and after 2014.
* San Bernardino Police’s CAD data are only complete dating back to 2012.
11CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Figure 5
Law Enforcement Agency Response Times Generally Increased for All Crimes Since 2009
Riverside Police
San Bernardino Police
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
0
50
100
150
200
250
Average Number of MinutesAverage Number of MinutesAverage Number of Minutes
*
Non-Proposition 47-Related Response Times
Proposition 47-Related Response Times
53%
43%
0%
84%
32%
18%
6%
43%
56%
39%
50%
50%
Since 2014
Since 2014
Since 2014
Since 2014
Since 2014
Since 2014
Riverside Sheriff
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Source: Analysis of Riverside Sheriff, Riverside Police, and San Bernardino Police CAD systems.
Note: We calculated the percentage change in average response times before and after 2014.
* San Bernardino Police’s CAD data are only complete dating back to 2012.
12 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Riverside Police and SanBernardino Police explained that the increases in response
times were the result of reductions in staffing levels. For example, Riverside Police cited
challenges related to the recruitment and retention of officers. Research published in
 by the Public Policy Institute of California shows that from  to , Riverside
and SanBernardino counties had decreases in patrol officers per , residents
of  percent and  percent, respectively, which appears to corroborate the agencies’
assertions.
6
In addition, as Figure  shows, the number of calls for service increased
at Riverside Police and generally remained stable at Riverside Sheriff at the same time
as the decrease in staffing, which may have affected the timeliness of their response.
Riverside Sheriff explained that response time results for its department could be skewed
by reclassifying crimes after a deputy arrives on the scene. Riverside Sheriff uses a
tieredpriority system to rank calls for service, with more serious crimes receiving highest
priority and lower-level crimes receiving lower priority. A deputy may respond to a call
for a higher-level crime, but then determine that a less serious crime took place and
reclassify the call after investigating. Nevertheless, response times generally increased
for both Proposition and non-Proposition related service calls and the explanation
given by Riverside Sheriff would not seem to affect the trends we identified.
The Number of Arrests for Proposition 47 Offenses Has Declined
e law enforcement agencies use a record management system (RMS) to document
arrests for criminal offenses. We analyzed their RMS data to determine their number
of arrests, which include citations and bookings, for Proposition offenses and
non-Proposition offenses from  to  for Riverside Police and Riverside
Sheriff and from  to  for SanBernardino Police.
7
We found that arrests for
Proposition offenses declined at all three agencies from  through  following
the implementation of Proposition, as Figure  shows. For example, in 
Riverside Sheriff made , arrests for Proposition offenses, while in  it made
,arrests for Proposition offenses, a decline of  percent. However, prior to the
implementation of Proposition, arrests for Proposition offenses slightly increased
for Riverside Police and SanBernardino Police. For Riverside Sheriff, the largest agency,
arrests for Proposition offenses had decreased prior to the implementation of
Proposition. Appendix B shows the percentage change in arrest charges at the three
agencies for each Proposition offense category.
ese decreases align with California Department of Justice (DOJ) reported trends in
total arrests statewide and in Riverside and SanBernardino counties over the same
period. Specifically, from  through , Riverside and SanBernardino counties had
a decrease in all arrests of  percent and  percent respectively, and statewide levels
also fell by  percent. Although we found a decrease in arrests that coincides with the
implementation of Proposition, our analysis does not provide a definite causative
explanation for the decrease. In  the Public Policy Institute of California reported
that arrest trends can be affected by factors such as economic conditions, demographics,
crime rates, and the implementation of Proposition, among others.
6
Magnus Lofstrom, Brandon Martin, Andrew Skelton, “California’s Notable Declines in Law Enforcement Staffing,” February, .
7
San Bernardino Police’s RMS data are only complete dating back to .
13CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Figure 6
Law Enforcement Agency Arrests for Proposition 47 Offenses Decreased Since the Implementation
ofProposition 47
Number of ArrestsNumber of ArrestsNumber of Arrests
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
11,000
12,000
Riverside Sheriff
49%
59%
50%
30%
14%
4%
1%
Riverside Police
San Bernardino Police
*
Since 2014
21%
Since 2014
Since 2014
15%
Since 2014
67%
60%
Since 2014
29%
Since 2014
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
2011 20122009 2010 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Non-Proposition 47-Related Arrests
Proposition 47-Related Arrests
Source: Analysis of Riverside Sheriff, Riverside Police, and San Bernardino Police RMS data.
Note: We calculated the percentage change in total arrests before and after 2014.
* San Bernardino Police’s RMS data are only complete dating back to 2012.
14 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
According to the agencies, there were likely a number of causes for the decline in
arrests for Proposition offenses. For example, a lieutenant with Riverside Police
explained that businesses now often report crimes days or weeks after they occur,
usually online. ese delays can hinder officers’ ability to make an arrest. e
assistant chief at SanBernardino Police provided a similar explanation. He also
explained that many retail businesses report online after the fact for insurance
purposes only. In addition to issues surrounding online reporting, the assistant
chief for the SanBernardino Police stated that generally, law enforcement agencies
dedicate available resources to consequential crimes that affect the community
and that the implementation of Proposition had a cooling effect on agencies’
devotion of scarce resources to Proposition offenses. A deputy with Riverside
Sheriff explained that calls for property crimes generally lack the evidence needed to
conduct a thorough investigation, and that contributes to less arrests.
Following the Implementation of Proposition47, the Number of Individuals Who
Reoffended With Four or More Convictions Increased at the Three Law Enforcement
Agencies We Reviewed
When we reviewed publicly available case histories and court records for a statistical
sample of individuals convicted of a Proposition offense in  and  at the three
law enforcement agencies we reviewed, we found that overall rates of recidivism—
conviction of a new felony or misdemeanor committed within threeyears of a previous
conviction—declined slightly, as Figure  shows. We refer to recidivism throughout
the report as reoffense. is decline generally aligns with CDCR’s measure of statewide
recidivism for its - group of  percent and its – group of  percent.
However, among those who reoffended in our sample at the three law enforcement
agencies we reviewed, our analysis found that there were more individuals with four or
more convictions, an increase of  percentage points, following the implementation
of Proposition. Moreover, among this group of individuals, the frequency of theft
convictions increased after the implementation of Proposition.
The Number of Individuals Who Reoffended With Four or More Convictions Increased at the
Three Law Enforcement Agencies We Reviewed After the Implementation of Proposition 47
BSCC defines recidivism as conviction for a
new felony or misdemeanor committed within
three years of release for a previous criminal
conviction, as the text box describes. To
assess the rates of reoffense before and after
Proposition took effect, we reviewed court
records and publicly available case histories of
individuals convicted of Proposition offenses
at the three law enforcement agencies we
reviewed before and after its implementation.
Wethen consulted with a statistician to interpret
the results of our analysis.
Definition of Recidivism
Recidivism: Conviction for a new felony or misdemeanor
committed within three years of release from custody
orcommitted within three years of placement on
supervision for a previous criminal conviction.
This definition excludes cases dropped or dismissed,
noncriminal traffic offenses, and violations of supervision
(parole, probation, mandatory supervision, etc.).
Source: BSCC.
15CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Figure 7
Although Overall Reoffense Rates Decreased at the Three Law Enforcement Agencies We Reviewed, the
Number of Individuals Who Reoffended With Four or More Convictions Increased
Individuals with
+ Convictions
31%
Individuals with
+ Convictions
43%
2011
Group
201
Group
Percentage
Point Change
The percentage of individuals
with four or more convictions
who reoffended increased
after Proposition 47.
Reoffense
60%
Reoffense
55%
The overall rate of
reoffense for individuals
convicted of Proposition 47
offenses decreased slightly
after Proposition 47.
Rate of Theft
Convictions
Rate of Theft
Convictions
The frequency of theft
convictions among those
with four or more convictions
increased after Proposition 47.
5%
12%
14%
25%25% 39%39%
Source: DOJ; analysis of Riverside Superior Court and San Bernardino Superior Court public documents and court records.
Note: Refer to Appendix C for a description of the methodology used to select our two sample groups of individuals convicted of
offenses that were later reclassified by Proposition 47 in 2011 and another group in 2016.
We compared a group of randomly selected individuals arrested by one of the three
law enforcement agencies we selected for review and subsequently convicted of a
Proposition offense in Riverside and SanBernardino counties in , before
the implementation of Proposition, to a group of randomly selected individuals
who were convictedof a Proposition offense in , after the proposition’s
implementation. Using public records from the respective county courts, we
determined whether the individuals in each group reoffended within three years of
their first conviction for a Proposition offense and, if so, how many subsequent
convictions they had within that period. We included non-Proposition offenses
when determining subsequent convictions. For example, someone who was convicted
of a Proposition offense involving drug possession might have reoffended with
16 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
a conviction for a non-Proposition offense, such as driving under the influence or
battery. Overall, we analyzed the criminal records of  individuals over a three-year
period. We describe the methodology we used for this analysis in detail in Appendix C.
e overall rates of reoffense in our samples before and after the implementation
of Proposition declined from  percent to  percent.
8
However, the number of
individuals in our sample with four or more convictions who reoffended increased, as
Table  shows.
9
ese results suggest that although individuals who were convicted of
Proposition offenses generally reoffended less frequently in the later sample, a subset
of this group committed more offenses. Although we identified a significant difference
of  percentage points between the two groups among individuals with four or more
convictions, we only identified slight variations— percentage points or less—among
individuals in the two groups with less than four convictions. If we project these sample
results to the population of the three law enforcement agencies’ jurisdictions, we estimate
that about  more individuals may have reoffended with four or more convictions after
the implementation of Proposition than did so before its implementation.
10
Table 2
The Overall Rate of Reoffenses Decreased, but the Percentage of Those Who Reoffended With Four or
More Convictions Increased
Overall Reoense Rate
GROUP SAMPLE SIZE
NUMBER OF
INDIVIDUALS WHO
REOFFENDED
PERCENTAGE OF
INDIVIDUALS WHO
REOFFENDED
PERCENTAGE
POINT
CHANGE
2011 sample (pre-Proposition 47) 107* 64 60%
5% ↓
2016 sample (post-Proposition 47) 114 63 55%
Individuals Who Reoended With Four or More Convictions
GROUP
NUMBER OF
INDIVIDUALS
IN SAMPLE
WHO
REOFFENDED
NUMBER OF
INDIVIDUALS IN
SAMPLE WITH
FOUR OR MORE
CONVICTIONS
PERCENTAGE OF
INDIVIDUALS WHO
REOFFENDED WITH
FOUR OR MORE
CONVICTIONS
PERCENTAGE
POINT
CHANGE
2011 group (pre-Proposition 47) 64 20 31%
12% ↑
2016 group (post-Proposition 47) 63 27 43%
Source: Analysis of court records and publicly available case histories from San Bernardino and Riverside County.
* The sample size for this group was 113, but Riverside Superior Court attested it destroyed the records for six individuals in our
sample pursuant to record retention law.
8
Our results for the overall rates of reoffense were not found to be statistically significant.
9
Our results for those with four or more convictions were determined to be statistically significant at an  percent confidence
level. However, studies by the U.S. Department of Justice Bureau of Justice Statistics use a  percent confidence level for
determining significance. For more information regarding statistical significance, refer to Appendix C.
10
For our  cohort, the lower limit is  and the upper limit is , with a best estimate of . For our  cohort, the lower
limit is  and the upper limit is ,, with a best estimate of .
17CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
In an effort to substantiate our recidivism analysis, we analyzed Riverside Sheriff’s
RMS to determine how many individuals it arrested four or more times in each year
during the audit period. We found that the average number of individuals it arrested
four or more times for Proposition offenses in a year more than doubled following
the implementation of Proposition, from  before Proposition to  after
Proposition. Although not all arrests lead to a conviction, these data support our
recidivism analysis finding that a subset of individuals were convicted ofcommitting
more offenses after the implementation of Proposition.
The Frequency of Theft Convictions at the Three Law Enforcement Agencies We Reviewed
Increased After the Implementation of Proposition 47, Particularly for Those Who Reoffended
With Four or More Convictions
e frequency of theft convictions and—more specifically—of theft convictions
among those who reoffended with four or more convictions increased at the three
law enforcement agencies we reviewed after the implementation of Proposition.
eft convictions in our  sample were  percentage points higher than theft
convictions in our  sample. More significantly, theft convictions among
the subset of individuals who reoffended with four or more convictions rose by
percentage points. is increase was higher than the increases for any other
Proposition offense categories, as Table  shows.
11
For example, we identified an
individual in our sample who was convicted for seven theft-related offenses after the
implementation of Proposition. is individual was convicted three times within
three months and then four more times within the next three years. All of these were
misdemeanor convictions.
e law enforcement agencies asserted that the reduced consequences for
Proposition offenses may have caused the increase in the frequency of theft cases
in our samples. For example, before Proposition, petty theft could be punished as
a felony if a person had three prior theft convictions; however, after Proposition,
petty theft with a prior theft conviction is a misdemeanor except in very limited
circumstances. A deputy with Riverside Sheriff and an assistant chief with
SanBernardino Police stated that they believe Proposition created an incentive
for individuals to commit multiple offenses by reducing the consequences for repeat
offenses. To assess whether the data indicated a decrease in sentencing length, we
analyzed the average sentence length for individuals in our sample groups convicted
of a Proposition offense, which we discuss in the next section.
11
Our results were determined to be statistically significant at an  percent confidence level. However, studies by the
U.S.Department of Justice Bureau of Justice Statistics use a  percent confidence level for determining significance.
Formore information regarding statistical significance, refer to Appendix C.
18 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Table 3
Theft Convictions for Individuals Who Reoffended With Four or More Convictions Increased
MoreThan Convictions in Other Proposition 47 Categories
Frequency of Proposition 47 Convictions for ose With Four or More Convictions
OFFENSE 2011 GROUP PERCENTAGE* 2016 GROUP PERCENTAGE
PERCENTAGE POINT
CHANGE
Theft
25% 39% 14% ↑
Drug possession
60 51 9% ↓
Burglary
15 6 9% ↓
Forgery
0 4 4% ↑
Source: Analysis of court records and publicly available case histories from San Bernardino and Riverside counties.
Note: Our recidivism analysis captured all convictions an individual had, as we explained in the previous section. However,
the results above exclude non-Proposition 47 convictions.
* For our 2011 sample group, convictions for what is now known as shoplifting may fall into either our Theft or Burglary
category. Prior to Proposition 47, the theft of property worth $950 or less from a commercial establishment could be
charged as a felony. After the passage of Proposition 47, this crime became misdemeanor shoplifting. Also, after the
passage of Proposition 47, individuals previously convicted of the theft of property worth $950 or less could petition to
have their felony sentences retroactively converted to misdemeanors. This led to instances of shoplifting appearing for
individuals in our 2011 sample group which we then included in our Theft category, even though the shoplifting penal
code did not exist before 2014. If an individual from the 2011 sample group did not receive a change in sentence to
misdemeanor shoplifting, their felony burglary conviction would still be captured under our Burglary category.
Our Review Indicates That Proposition47 Has Achieved Some of Its Intended Purposes
When we analyzed court records to determine the average sentence length for
Proposition offenses in our sample population, we found that the length of
sentences had decreased at the three law enforcement agencies we reviewed after the
implementation of Proposition, in line with the law’s intended purpose. Specifically,
following the law’s implementation, the average sentences individuals in our samples
received for Proposition offenses decreased by more than percent. is drop was
likely the result of reductions in sentences for theft and drug possession offenses, as
Table shows. Proposition reclassified certain theft and drug possession offenses from
felonies or wobblers to misdemeanors, the latter of which calls for shorter jailsentences.
12
e Riverside County District Attorney’s Office (Riverside DA) agreed that Proposition
has definitely achieved its intended goal of reducing prison sentences for Proposition
offenses. Nevertheless, the Riverside DA attested that it believes there have been
several unintended negative consequences, including an increase in organized retail
theft offenses. We discuss arrests for organized retail theft later in this report.
13
12
The average length of time served was  days in jail for our  group. Of the records we reviewed for Proposition 
offenses in which an individual was sentenced to jail time, the main reasons individuals did not spend the full sentenced
time in jail included court orders to alleviate overcrowding, work release programs, and good behavior credits. However,
Proposition  did not amend or alter these reasons.
13
Organized retail theft became a criminal offense effective January , , several years after Proposition  was implemented.
19CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Table 4
The Average Sentence Length for a Proposition 47 Conviction Decreased After Its Implementation
Sentencing for Proposition 47 Oenses
AVERAGE NUMBER OF DAYS SENTENCED
GROUP
ALL PROPOSITION 47
CONVICTIONS
THEFT CONVICTIONS
DRUG POSSESSION
CONVICTIONS
FORGERY
CONVICTIONS
BURGLARY
CONVICTIONS
2011 Sample
(pre-Proposition 47)
349 days 335 days 429 days 90 days 268 days
2016 Sample
(post-Proposition 47)
199 days 207 days 96 days 90 days 603 days
Percentage change 43% ↓ 38% ↓ 78% ↓ 0% 125% ↑
Source: Analysis of court records and publicly available case histories from San Bernardino and Riverside counties.
Organizations that advocate for reforms to the criminal justice system also provided
their perspective on Proposition’s effects. When we spoke to entities such as the
American Civil Liberties Union of Southern California (ACLU SoCal), the Ella Baker
Center for Human Rights, Californians for Safety and Justice, and Legal Services
for Prisoners with Children, they cited studies that have shown the positive effects
Proposition has had on California. For example, the BSCC performed a statewide
evaluation of its Proposition grant program and found that recidivism among
participants enrolled in services funded by the savings associated with Proposition
was lower than other reported statewide recidivism estimates.
14
Specifically,
participants had a  percent rate of recidivism after enrolling in a grant program,
while statewide recidivism estimates were between  and  percent.
15
is BSCC
study also found a  percent decrease in homelessness and a  percent decrease in
unemployment among participants who received the Proposition-funded services
and identified housing or employment as goals of their participation. In another
example, the National Institute of Justice gave a Promising rating to the policy reform
enacted by Proposition, which implies that the policy reform is achieving its
intended outcomes.
16
Given these and other studies, the organizations to which we spoke largely support
efforts to increase the State’s investment in programs that identify and address
the factors that may cause individuals to commit Proposition offenses. ese
factors include unmet community needs, such as access to housing, employment,
and programs to address mental health concerns. ACLU SoCal staff explained that
14
BSCC, “Proposition  Cohort II: Statewide Evaluation,” February , <https://www.bscc.ca.gov/wp-content/
uploads///H--Proposition--Cohort--Final-Evaluation-Report-FINAL-.pdf>, accessed on April , .
15
The BSCC study indicated that the statewide recidivism rates may not be equitable recidivism timeframe
comparisongroups.
16
National Institute of Justice, “Program Profile: The Impact of California’s Proposition  (The Reduced Penalties for Some
Crimes Initiative) on Recidivism,” November , <https://crimesolutions.ojp.gov/ratedprograms/->, accessed on
April , .
20 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
they believe opportunities to address these community needs have been historically
underfunded in Riverside and SanBernardino counties and constitute some of the
underlying causes of why individuals may commit Proposition offenses.
In addition, the organizations expressed concerns that enhancing sentencing for
Proposition offenses would have negative consequences. For example, they
asserted that this step could remove funding for programs that have been effective
at addressing unmet community needs, as we describe above. Legal Services
forPrisoners with Children also emphasized that efforts to enhance sentencing for
Proposition offenses could actually lead to an increase in recidivism rates because
the collateral consequences associated with individuals receiving felony convictions—
including substantial barriers to accessing housing, employment, and education—
could negatively reinforce the underlying causes of why people commit crimes.
California Has Several Options for Addressing the Increases in Multiple Reoffenses
andTheft
As we previously explain, our analysis identified an increase among those who
reoffended with four or more convictions after the implementation of Proposition
at the three law enforcement agencies we reviewed. We also found an increase in
the frequency of theft convictions, particularly for individuals with four or more
convictions. To address the issues that we observed, the Legislature may want to
consider changes to the laws impacted by Proposition.
During the – legislative session, state legislators introduced multiple pieces of
legislation related to Proposition offenses. One option these bills have proposed
is increasing penalties for individuals convicted of multiple Proposition offenses.
We provide an example of some code sections that could be considered for such a
legislative change in Table . A second option the bills identify is expanding theft
diversion programs that allow individuals to enter treatment programs in lieu of
receiving jail sentences.
Proposition can be amended by a two-thirds vote of the members of each house
of the Legislature and signature by the Governor but only if the amendments are
consistent with, and further the intent of, the proposition. Alternatively, the Legislature
could also enact a proposal to amend Proposition; however, the proposed
amendment would not become effective until submitted to and approved by the
voters. is alternative process is required if the amendment is not consistent with
or does not further the intent of the proposition. Further, the people may introduce
ballot initiatives that amend or repeal Proposition. For example, ballot initiative
-A is qualified for the statewide November general election. If the voters
adopt it, this initiative would, among other things, make a third conviction for petty
theft and shoplifting a potential felony.
Our audit was limited to the three law enforcement agencies in two counties, and
the results of our analysis are not intended to be projected statewide. Although our
analysis showed an increase in the frequency of theft convictions among those who
reoffended with four or more convictions after the implementation of Proposition,
21CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
our analysis shows a correlation, but does not provide an explanation for the cause
of this increase. Nevertheless, we also considered interviews with law enforcement
agencies and other stakeholder groups, along with the external studies we discussed
previously, to ultimately inform the policy options listed below for the Legislature
toconsider.
Table 5
Examples of Laws That Could Be Amended to Allow Felony Charges if an Individual Has Multiple
Theft Convictions
CRIMINAL ACT CRIME TYPE
PC 459.5 — Shoplifting under $950
Theft
PC 490.2 — Theft under $950
PC 488 — Theft of money, labor, or property under $950
PC 496 — Receiving or buying stolen property under $950
Source: State law.
Note: This table provides offenses in the penal code that could be amended. It is not an exhaustive list of all offenses
affected by Proposition 47.
Recommendations
If the Legislature decides that it wants to take action to reduce the likelihood that
individuals with multiple convictions will continue to engage in theft, the Legislature
could consider the adoption of either or both of the following policy mechanisms:
• 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 to the types of programs and services that recent studies have identified as
effective in reducing recidivism.
• Either by enacting legislation that is consistent with and furthers the intent
ofProposition, or by enacting a proposal for approval by the voters that would
amend Proposition, 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.
22 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Other Areas Reviewed
The Law Enforcement Agencies We Reviewed Did Not Make Significant Changes to
Their Policies as a Result of the COVID-19 Pandemic
e Audit Committee requested that we identify public safety policy changes in
SanBernardino and Riverside counties that resulted from the COVID- pandemic
and that we determine whether those policy changes affected the level of criminal
activity. We found that the three law enforcement agencies generally did not make any
policy or procedural changes to their normal operations as a result of the pandemic,
other than requiring notices for COVID- exposures. According to the agencies,
the pandemic did not result in any formally documented changes to department
policies and enforcement procedures except for some logistical changes to reduce
potential exposure to the virus. us, we did not identify substantive policy changes
at the three agencies that affected criminal activity in Riverside and SanBernardino
counties. Nevertheless, as we discuss in the Introduction, statewide  bail rules
implemented during the pandemic may have affected incarceration levels.
Organized Retail Theft From 2019 to 2022
Over the past six years, the Legislature has
passed several pieces of legislation to combat
organized retail theft, an offense that became
effective in , several years after the
implementation of Proposition in . We
summarize some of the key components of
this legislation in the text box. Data from the
selected law enforcement agencies show that
of the nearly , arrests they made from
 to ,  included arrest charges with
penal code provisions for organized retail theft.
According to officers at SanBernardino Police
and Riverside Sheriff, one explanation for the
seemingly small number of arrests for organized
retail theft is that officers may arrest individuals
who have allegedly committed organized retail
theft offenses under penal code sections that
are not specifically labeled as “organized retail
theft.” Riverside DA offered that a reason there
are relatively few arrests for organized retail theft
could be related to underreporting of theft-related
crimes. Riverside Sheriff also explained that the
California Highway Patrol received state funding
to address organized retail theft. erefore, the
arrest figures from the three law enforcement
agencies may not represent the total number of
organized retail theft arrests made in Riverside
and SanBernardino counties. However, this
Key Components of Recent Legislation to
Address Organized Retail Theft
• Defined organized retail theft to include the commission
of acts in concert with one or more persons to steal,
receive, purchase, or knowingly possess merchandise
taken from a merchant’s premises or online marketplace.
• Enables the prosecution of individuals who commit acts
of organized retail theft. Prosecutors can consider the
aggregated value of stolen merchandise over a 12‑month
period when charging an individual with organized
retailtheft.
• Extended the jurisdiction, and altered the subsequent
prosecution, of a criminal action for theft to include the
county where an offense involving the theft or receipt of
the stolen merchandise occurred, the county in which
the merchandise was recovered, or the county where any
act was done by the defendant in instigating, procuring,
promoting, or aiding in the commission of a theft offense.
• Enabled officers to hold a person following a misdemeanor
arrest if the officers have probable cause to believe the
person is guilty of committing organized retail theft.
• Mandated the California Highway Patrol to convene
a regional property crimes task force to assist local
law enforcement in counties with elevated levels of
propertycrime.
Source: State law.
23CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
was an audit specifically of Proposition’s effects in the two counties and the
penal code provisions for organized retail theft we analyzed were enacted years
after the implementation of Proposition. Ultimately, as we note on page , our
analysis found the frequency of theft convictions and—more specifically—of theft
convictions among those who reoffended with four or more convictions increased
after the implementation of Proposition, and we provide recommendations to the
Legislature to address thisissue.
We conducted this performance audit in accordance with generally accepted
government auditing standards and under the authority vested in the California
State Auditor by Government Code section  et seq. ose standards require that
we plan and perform the audit to obtain sufficient, appropriate evidence to provide
a reasonable basis for our findings and conclusions, based on the audit objectives.
Webelieve that the evidence obtained provides a reasonable basis for our findings
and conclusions, based on our audit objectives.
Respectfully submitted,
GRANT PARKS
California State Auditor
July , 
Staff:
Nicholas Kolitsos, CPA, Audit Principal
Aaron Fellner, Senior Auditor
Eduardo Moncada
Kaleb Knoblauch
Myra Farooqi
Ashley Yan
Data Analytics: R. Wade Fry, MPA
Shauna M. Pellman, MPPA, CIA
Sarah Rachael Black, MBA, CISA
Legal Counsel: Joe Porche
24 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Appendix A
VOLUME OF SERVICE CALLS FOR ALLEGED PROPOSITION 47 OFFENSES
BY CRIME CATEGORY
Table A shows the volume of alleged Proposition -related service calls made
to each of the three law enforcement agencies we selected for review. e table
separates the calls by the categories of the alleged offenses: burglary, drug
possession, forgery, and theft.
Table A
Volume of Alleged Proposition 47-Related Service Calls by Crime Category
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Riverside Police
Burglary 5,884 6,066 5,681 5,828 5,613 4,475 5,031 5,723 5,077 4,881 4,585 4,881 5,118 4,937
Drug possession 549 419 433 458 519 467 207 225 178 207 287 143 111 88
Forgery 306 296 277 359 228 281 257 271 307 306 335 206 125 171
Theft 8,454 8,346 8,069 9,237 9,217 9,237 9,838 10,824 9,886 10,093 9,609 9,182 9,028 10,341
Riverside Sheri
Burglary 11,020 10,216 11,107 12,952 11, 294 10,285 9,166 8,334 8,523 7,567 7,302 6,011 5,940 6,037
Drug possession 9,289 9,000 8,898 10,088 11,439 11,6 47 9,888 10,132 9,353 8,992 9,988 9,385 8,770 8,876
Forgery 7,811 7,350 6,897 8,544 9,228 10,275 10,405 10,608 9,881 8,869 8,879 7, 861 8,941 10,346
Theft 32,948 31,059 33,809 39,194 39,339 38,690 45,126 46,775 45,675 44,119 42,878 38,240 38,412 42,619
San Bernardino Police
*
Burglary 1,562 1,493 1,236 1,712 1,298 1,447 1,662 1,169 537 324 335
Drug possession 7 5 2 2 2 1 4 0 0 0 0
Forgery 638 429 480 566 438 408 425 421 137 147 102
Theft 6,215 6,665 6,388 7,004 5,434 4,878 5,044 4,772 3,520 4,084 3,635
Source: Analysis of Riverside Police, Riverside Sheriff, and San Bernardino Police CAD systems.
* San Bernardino Police’s CAD data are only complete dating back to 2012.
25CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Blank page inserted for reproduction purposes only.
26 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Appendix B
NUMBER OF ARREST CHARGES AT EACH LAW ENFORCEMENT AGENCY FOR
ALLEGED PROPOSITION 47 OFFENSES BY CRIME CATEGORY
Table B shows the number of arrest charges for alleged Proposition offenses at the three
law enforcement agencies we selected for review. e table separates the arrest charges by
category. An individual might be arrested on multiple charges. e decreases in arrests,
which we discuss in the Audit Results, correspond with the decreases in arrest charges,
but because multiple arrest charges may be combined in a single arrest, the table will not
reconcile to Figure .
Table B
Number of Arrest Charges at Each Law Enforcement Agency for Alleged Proposition 47 Offenses by Crime Category
2009 2010 2011 2012 2013 2014
PERCENT
CHANGE
THROUGH
2014
2015 2016 2017 2018 2019 2020 2021 2022
PERCENT
CHANGE
FROM 2014
THROUGH
2022
OVERALL
PERCENT
CHANGE
Riverside Police
Burglary 542 500 497 648 657 618
↑ 14%
200 213 194 129 169 167 136 163
↓ 74% ↓ 70%
Drug
possession
1,346 1,306 1,018 1,163 1,381 1,557
↑ 16%
945 1,152 928 1,070 1,169 999 867 537
↓ 66% ↓ 60%
Forgery 144 104 115 108 98 110
↓ 24%
74 51 74 70 36 23 28 8
↓ 93% ↓ 94%
Theft 2,097 1,866 1,700 1,784 1,766 1,936
↓ 8%
1,894 1,843 1,543 1,240 1,182 813 624 600
↓ 69% ↓ 71%
Riverside Sheri
Burglary 2,536 2,211 1,992 2,124 1,862 1,632
↓ 36%
663 600 614 560 488 498 419 610
↓ 63% ↓ 76%
Drug
possession
4,832 4,390 4,041 4,046 4,923 4,719
↓ 2%
2,778 2,976 2,626 2,581 3,142 3,080 3,384 3,757
↓ 20% ↓ 22%
Forgery 393 308 287 229 274 210
↓ 47%
196 185 177 126 130 82 84 119
↓ 43% ↓ 70%
Theft 4,801 4,179 3,909 4,022 3,915 3,840
↓ 20%
4,262 3,671 3,013 2,790 2,597 1,994 1,910 2,986
↓ 22% ↓ 38%
San Bernardino Police
*
Burglary 297 409 340
↑ 14%
269 210 180 178 183 163 123 125
↓ 63% ↓ 58%
Drug
possession
462 498 604
↑ 31%
205 316 383 498 424 390 315 294
↓ 51% ↓ 36%
Forgery 56 39 43
↓ 23%
16 19 22 15 15 4 7 3
↓ 93% ↓ 95%
Theft 1,116 1,050 1,047
↓ 6%
721 634 610 614 530 419 406 367
↓ 65% ↓ 67%
Source: Analysis of Riverside Police, Riverside Sheriff, and San Bernardino Police RMS data.
* San Bernardino Police’s RMS data are only complete dating back to 2012.
27CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Blank page inserted for reproduction purposes only.
28 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Appendix C
RECIDIVISM METHODOLOGY
e Audit Committee directed us to assess the rates of recidivism in SanBernardino
and Riverside counties for individuals following their convictions for Proposition
offenses before and after the proposition took effect. We consulted with a statistician
to use data from the DOJ to select statistically valid sample for two cohorts—one in
 and one in —across the three law enforcement agencies. is allowed us to
project to the population of individuals convicted of a Proposition offense for the
jurisdictions of the three law enforcement agencies in the two counties.
Our samples consisted of  total offenders divided into the two cohorts, as FigureC
describes. To measure reoffense rates before the implementation of Proposition
in November , we randomly selected from the three law enforcement agencies a
proportional sample of  individuals who were convicted of a Proposition offense
in .
17
To measure reoffense rates after the implementation of Proposition and
to allow the changes to law some time to take effect, we also randomly selected from
the three law enforcement agencies a proportional sample of  individuals who were
convicted of a Proposition offense in . e availability of data affected the size
of our samples. In some cases, the smaller sample sizes had minimal impact on the
precision of the results.
We developed our analysis based on the BSCC’s definition of recidivism: a conviction
of a new felony or misdemeanor committed within three years of release from
custody or within three years of placement on supervision for a previous criminal
conviction. According to this definition, “a new felony or misdemeanor” excludes
cases dropped or dismissed; noncriminal traffic offenses, regardless of guilt;
new jail admissions; and violations of supervision (parole, probation, mandatory
supervision, or other terms of release). We reviewed publicly available case histories
and court records provided by Riverside and SanBernardino county superior
courts to determine whether individuals in the samples were convicted of offenses
within three years of their first conviction. Additionally, we reviewed DOJ records
for offenses so that we could include any offenses committed outside of the county
where the individual was first convicted. We analyzed these results to determine the
rates of reoffense before and after Proposition.
We determined whether the sampled individuals reoffended within three years
of their first conviction. We consulted with a statistician to determine whether
our results were statistically significant. e statistician ran tests at  percent,
percent, and  percent confidence levels for our results to determine statistical
significance. For much of our analysis, we assessed our results in terms of statistical
significance using preset confidence levels. Statistical significance is a test commonly
used to determine whether the results derived from analysis of a sample could be
caused by chance and the extent to which we can draw conclusions on the total
17
The sample size for this group was , but Riverside Superior Court attested it destroyed the records for six individuals in
our sample pursuant to record retention law.
29CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
population. For example, for a sample with an  percent confidence level, we would
generally have an  percent certainty that the population’s true results would be
reflected in the sample’s results.
Figure C
Our Selection Methodology for Determining Recidivism at the Three Law Enforcement Agencies
We Reviewed
The first group consisted of
113 individuals convicted of a
Proposition 47 offense in 2011.*
We determined whether
individuals in the 2011 group
were convicted of any crimes
within three years of their first
conviction date (2011–2014).
The second group consisted of
114 individuals convicted of a
Proposition 47 offense in 2016.
We determined whether
individuals in the 2016 group
were convicted of any crimes
within three years of their first
conviction date (2016–2019).
Using DOJ data, we created
two randomized sample groups.
Using court documents and case histories,
we determined the rates of recidivism.
Source: DOJ; analysis of court records and publicly available case histories from San Bernardino and Riverside counties.
* The sample size for this group was 113, but Riverside Superior Court attested it destroyed the records for six individuals in
our sample pursuant to record retention law.
30 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Appendix D
SCOPE AND METHODOLOGY
e Audit Committee requested the California State Auditor to conduct an audit
ofthe effects of Proposition in SanBernardino and Riverside counties. Table D lists
theobjectives that the Audit Committee approved and the methods we used to address
them. Unless otherwise stated in Table D or elsewhere in the report, statements and
conclusions about items selected for review should not be projected to the population.
Table D
Audit Objectives and the Methods Used to Address Them
AUDIT OBJECTIVE METHOD
1
Review and evaluate the laws, rules,and
regulations significant to the audit objectives.
Reviewed state laws, rules, and regulations related to the audit objectives.
2
Review available crime data specific to
SanBernardino and Riverside counties for the
period before and after the implementation
of Proposition47 to identify any trends in
applicable crime categories. Determine
the effect that Proposition 47 had on rates
of reported crime in San Bernardino and
Riversidecounties.
• We evaluated the data potentially available at the statewide and local law enforcement
level, and found that local law enforcement data contained additional information
not captured in the statewide data. Therefore, we analyzed data from the three law
enforcement agencies we selected for review.
• Obtained and analyzed RMS data from the three law enforcement agencies we reviewed
to determine trends from 2009 to 2022 in arrests for Proposition 47 offenses and to
compare rates of Proposition 47 arrests with DOJ reports containing arrests for all crimes
statewide and in SanBernardino and Riverside counties. The analysis for SanBernardino
Police covers only the period from 2012 to 2022 because it implemented a new RMS
system in December 2011 and did not retain the data that was stored in its prior system.
3
For the selected law enforcement agencies
in each county, assess the impacts of
Proposition47by:
We selected four law enforcement agencies for review: Riverside Sheriff, Riverside Police,
San Bernardino County Sheriff’s Department, and San Bernardino Police. Because of data
availability concerns, we ultimately did not review data from San Bernardino County
Sheriff’sDepartment.
a. Reviewing trends of suspected crimes being
reported to law enforcement before and after
the implementation of Proposition 47 and
assess whether the Proposition has changed
how often the public calls for assistance, or
altered law enforcement’s response.
Obtained and analyzed CAD data from the three law enforcement agencies we reviewed
to determine trends from 2009 to 2022 in calls for service for Proposition 47 offenses and in
all other service calls. We used these same data to assess changes in the agencies’ response
times to these calls for service. The analysis for San Bernardino Police covers the period from
2012 to 2022 because it implemented a new CAD system in December 2011 and did not
retain the data that was stored in its prior system.
b. Identifying public safety policy changes that
resulted from the COVID-19 pandemic and
determine whether those policy changes had
an effect on the level of criminal activity.
Interviewed staff and obtained policies from the three law enforcement agencies regarding
COVID-19 pandemic policies to assess the impact of pandemic-related policy changes on
the levels of criminal activity.
4
Review available criminal history information to
assess the rates of recidivism before and after
Proposition 47 took effect on San Bernardino
and Riverside Counties.
• Consulted with a statistician to use DOJ data to select statistically valid samples
for two cohorts—one in 2011 and one in 2016—across the three law enforcement
agencies. Both cohorts consisted of individuals convicted of a Proposition 47 offense.
We calculated reoffense rates for the individuals in the cohorts based on each of the
Proposition 47 offenses and all other crimes by obtaining criminal history information
from the Riverside and San Bernardino county superior courts.
• Analyzed and reviewed court documents, case histories, and Jail Information
Management Systems data for our 2011 and 2016 cohorts to assess rates of reoffense
before and after the implementation of Proposition 47.
continued on next page …
31CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
AUDIT OBJECTIVE METHOD
5
Review and assess any other issues that are
significant to the audit.
• Reviewed organized retail theft laws and related arrests in the two counties.
• Analyzed the length of sentences and the time that individuals in our sample spent in jail
for convictions for Proposition 47 offenses.
• Requested perspective from additional organizations on the effects of Proposition 47.
• Interviewed California Department of Insurance (CDI) staff to determine whether
insurance company rate filings contained information pertaining to commercial
insurance rates changes and the rationale used by companies to justify changes to
California’s commercial insurance rates in an attempt to measure whether theft was
affecting commercial insurance rates. However, according to CDI staff, the department
does not compile or maintain any information on rate filings beyond that on the CDI
website. The information on CDI’s website did not include any data on commercial
insurance rate increases or the specific reasons why insurers modified rates charged to
businesses for property insurance in California. Further, CDI staff stated that it does not
issue any reports on commercial insurance rate increases.
• Contacted the California Retailer’s Association to determine whether it compiles
information on theft from its members. However, the association informed us it did not
have any data on how frequently retailers reported thefts to law enforcement.
• Reviewed the National Retail Federation’s Retail Security Survey to determine whether
retailers reported making changes in how they report theft to law enforcement.
However, the survey did not provide any relevant information.
Source: Audit workpapers.
Assessment of Data Reliability
e U.S. Government Accountability Office, whose standards we are statutorily obligated
to follow, requires us to assess the sufficiency and appropriateness of computer-processed
information that we use to support our findings, conclusions, or recommendations. In
performing this audit, we relied on various data sources. To determine trends of suspected
offenses being reported to law enforcement agencies before and after the implementation of
Proposition, and to assess any changes in law enforcement’s response to these reports, we
obtained calls for service data from Riverside Sheriff, Riverside Police, and SanBernardino
Police. Additionally, we analyzed arrest data from these same law enforcement agencies
before and after the implementation of Proposition. Finally, we relied on criminal history
data from DOJ to identify convictions related to Proposition. To assess these data, we
interviewed staff knowledgeable about the data, reviewed available information about the data,
and performed testing of the data. Although the law enforcement agencies asserted that some
offenses go unreported, we focused our analysis on those offenses that were reported to the law
enforcement agencies. As a result, we determined that these data were sufficiently reliable for
the purposes of our audit.
We also relied on data from Riverside county jail information management system to
identify any trends related to actual time served. To assess these data, we interviewed staff
knowledgeable about the data. We determined that these data were sufficiently reliable for
the purposes of our audit.
Lastly, we relied on data from Riverside and SanBernardino county superior courts to review
and analyze case histories, court records, and other relevant information. To assess these data,
we interviewed staff knowledgeable about the data. We determined that these data were also
sufficiently reliable for the purposes ofouraudit.
32 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
1
1
1
2
*
* California State Auditor’s comments begin on page 37.
33CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
2
3
4
5
5
1 6
7
34 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
8
7
9
35CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
Blank page inserted for reproduction purposes only.
36 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107
Comments
CALIFORNIA STATE AUDITOR’S COMMENTS ON THE RESPONSE FROM
RIVERSIDE SHERIFF
To provide clarity and perspective, we are commenting on Riverside Sheriff’s
response to our audit. e numbers below correspond to the numbers we have placed
in the margin of its response.
We informed Riverside Sheriff throughout the audit that the Government Code
sections  and . make it a misdemeanor to disclose any substantive
information about an audit before it is completed and that this includes information
pertaining to other audited entities. us, we did not provide Riverside Sheriff with
our conclusions related to the other audited entities. Riverside Sheriff is incorrect in
suggesting that we provided them with untimely information on our report findings.
In accordance with our standard practice, we shared our preliminary findings with
Riverside Sheriff throughout the audit, which necessitated that we contact them on
numerous occasions when they did not respond to our requests for information.
We worked closely with Riverside Sheriff’s staff to obtain and analyze RMS and CAD
data. As a result, when we were analyzing arrests, we did not use the CAD call closure
reasons that Riverside Sheriff discusses in its response. Instead, we used the approach
Riverside Sheriff’s staff recommended, which included directly analyzing arrest data
in RMS. is information—shown on pages  through —allowed us to identify the
number of and reasons for arrests, and we stand by our analysis and conclusions.
Finally, on page  we present data on the number of arrest charges for burglary
and theft—including petty theft and grand theft—which show a decrease of nearly
percent at Riverside Sheriff from  through . us, even if Riverside Sheriff
was charging individuals with other types of theft offenses, the data indicates that the
number of those arrest charges has decreased since the proposition’s implementation.
As we state on page , Riverside Sheriff was unable to provide evidence apart
from anecdotes to substantiate that underreporting of Proposition  offenses was
occurring. We provided Riverside Sheriff the opportunity to support their claims, but
Riverside Sheriff’s staff was unable to produce anything other than assertions. We
already provide the Sheriff’s perspective on page  of the report.
Riverside Sheriff improperly assumes we did not attempt to determine whether
retailers underreported theft offenses. As we state on page , we contacted the
California Retailer’s Association to discuss theft reporting by its members. However,
the association informed us that it did not have any data available on how frequently
retailers reported thefts to law enforcement. We also reviewed the National Retail
Federation’s publicly available survey data and the survey did not provide information
on reporting practices by retailers to law enforcement. However, we have no legal
authority to access the records of private entities unless a public entity being audited
has access to the records, or the private entities provide the records voluntarily. us,
we relied on available data and stand by our conclusions.
1
2
3
4
37CALIFORNIA STATE AUDITOR
Report 2023-107
|
July 2024
e Legislature requested an audit of the impact of Proposition  in SanBernardino
and Riverside counties, not organized retail theft, which is outside the scope of our
audit. e penal code sections for organized retail theft referenced on page were
enacted years after the implementation of Proposition . We present information
about organized retail theft for informational purposes only, but do not have any
findings, conclusions or recommendations about that information. Further, in
response to feedback from Riverside Sheriff, we explain on page that the number
from the three law enforcement agencies may not represent the total number of
arrest charges for organized retail theft in Riverside and SanBernardino counties.
However, we did find an increase in repeat theft offenses, and recommended on
page that if the Legislature decides that it wants to take action to reduce the
likelihood that individuals with four or more convictions will continue to engage in
theft, the Legislature may consider the adoption of several policymechanisms.
As we indicate on page , we removed the SanBernardino County Sheriff’s
Department from our review during the course of the audit due to the ransomware
attack that rendered much of its data unavailable.
State law mandates that our office be independent of the executive branch and
legislative control. Further, state law also requires the California State Auditor’s
Office to conduct all of its audits in accordance with auditing standards issued by
the U.S. Comptroller General’s Government Accountability Office. e auditing
standards we are statutorily required to follow also have provisions which guide our
independence from other entities. In addition, we base our conclusions on sufficient
and appropriate evidence in conformance with auditing standards.
e Riverside Sheriff believes that the report should include a recommendation to
repeal Proposition . However, the Legislature’s ability to repeal Proposition  is
limited by both Proposition ’s terms and the State Constitution. Nevertheless, we do
acknowledge on page  the steps that the Legislature could take to allow the voters
to repeal it. Moreover, our report presents detailed data and analysis that found the
length of sentences had decreased after the implementation of Proposition, in line
with the law’s intended purpose, but that the number of repeat theft convictions had
also increased. us, in accordance with auditing standards and best practices, our
recommendations relate to the significant issues that we identified and they do not
extend to areas outside the scope of our review andfindings.
Although Riverside Sheriff objects to the public release of our report, we are required by
law to make our audit reports available to the public in accordance with Government
Code Section ..
5
6
7
8
9
38 CALIFORNIA STATE AUDITOR
July 2024
|
Report 2023-107